Digital BW, The Print

I realize this may be of little interest to most but some have expressed interest. So here is some additional information. This is long. You have been warned.

 

Since we are on the Internet a web search seemed like the thing to do. Here is some of what I found plus an additional text reference. Be careful of long web addresses that wrap from one line to the next. You may need to cut and paste the pieces into your browser to get them to work.

 

Dynamic Range, Some Definitions:

 

From Merriam-Webster's Dictionary:

Main Entry: dynamic range
Function: noun
Date: 1949
: the ratio of the strongest to the weakest sound intensity that can be transmitted or reproduced by an audio or broadcasting system

http://www.m-w.com/cgi-bin/dictionary

 

From NASA website:

"The dynamic range of an image is the ratio between the maximum and minimum brightness levels in the image."

http://nasatech.com/Briefs/Aug98/NPO20254.html

 

From the Cornell University Library:

"DYNAMIC RANGE is the range of tonal difference between the lightest light and darkest dark of an image."

http://www.library.cornell.edu/preservation/tutorial/intro/intro-05.html

 

From Digital Photography Review:

"Dynamic range is the ratio between the brightest and darkest recordable parts of an image or scene."

http://www.dpreview.com/learn/Glossary/Digital_Imaging/Dynamic_Range_01.htm

 

From X-rite's "The Color Guide and Glossary"

"Dynamic Range: An instrument's range of measurable values, from the lowest amount it can detect to the highest amount it can handle."

 

From Mix Magazine:

"Dynamic range, generally measured in dB, is the ratio of maximum undistorted signal (full-scale or onset of clipping) to residual noise (noise floor)."

http://industryclick.com/magazinearticle.asp?releaseid=5828&magazinearticleid=72693&siteid=15&magazineid=141

 

From CCD Direct:

"The dynamic range is often represented as a log ratio of well depth to the readout noise in decibels. For example, a system with a well depth of 45,000 electrons and a readout noise of 15 electrons would have a dynamic range = 20 log (45,000/15), or 69dB."

http://www.ccddirect.com/online-store/scstore/dynamic.html

 

From outer space. [Caution, viewing this page may be hazardous to your sanity!] <G>

http://utam.geophys.utah.edu/UTAMtheses/JingChen/latex_2_html/node13.html

 

So what we have is a variety of definitions that seem to vary depending upon what field you are in but the common element is that they are looking for a meaningful way to describe a relationship between the minimum and maximum as a ratio.

 

I have to wander into some math here to clear up a point. Sorry but I just don't see any way around it.

 

Some people say the dynamic range is the difference between the min and max and some say it is the ratio. The interesting thing is that they are both sometimes correct. Some things such as sound volume, CCD response and image density are not direct measurements of a physical property but are calculated values from actual properties given in logarithmic form.

 

The log of a number is the value of that number as a power of a base number which is most commonly 10. The Log10 of 100 is 2, (10 raised to the 2nd power) the log10 of 10,000 is 4, (10 raised to the 4th power) the log10 of 112.3 is 2.0504, (10 raised to the 2.0504th power).

 

Now one of the nice things about logs is that if you add them the result is the same as multiplying the original numbers and if you subtract them the result is the same as dividing the original numbers when you convert the log values back to their original form. This is the reason logs were so widely used because they reduce complex multiplication and division problems to simple addition and subtraction. Very handy before we had pocket calculators and computers.

 

Photographic materials are usually measured in density. If it is a transparency or negative the density is the log of 1 divided by the percentage of light that passed through the film. If it is a print the density is the log of 1 divided by the percentage of light that was reflected from the print.

 


So if a print has a white paper base reflectance of 95% (95% of the light shining on this spot bounces off toward your eye) and the deepest black has a reflectance of 4% (only 4% of the light falling on this dark spot bounces off and the rest is absorbed or scattered away from your eye.) The dynamic range would be the maximum value of 95% divided by the minimum value of 4%, which is 23.75. Here taking the ratio is proper because we are talking about direct measurements of a physical property and not calculated values. The difference would be 91percentile, which is the range of the reflectance.

 

Usually in photography we do not talk about reflectance but instead use density. So for this case the paper base would have a density of 0.0223 (the base 10 log of 4%)and the darkest black 1.3979 (the base 10 log of 95%). Since these are log values calculated from the reflectance you find the dynamic range by subtracting them. 1.3979 minus 0.0223 is 1.3757 the dynamic range and the range of the Density.

 

Now if you take this Density dynamic range and raise 10 to the power of 1.3757 you get 23.75! The same number you got by dividing the maximum reflectance by the minimum reflectance.

 

So if you are talking about image density (a log value) the dynamic range and the range are the same. If you are talking about reflectance or transmission they are not the same. One is the ratio and the other is the difference.

 

It is interesting to note that if the reflectance range is 1% to 92% the range stays the same at 91 but the dynamic range goes up from 23.75 to 92. If the range shifts up to 8% to 99% (still a range of 91) the dynamic range drops to 12.38. This suggests to me that the simple range of reflectance with the min and max values is of more interest than the dynamic range when looking at a print or judging print medium quality.

 

Noise generally does not seem to be used in the calculation of the dynamic range but it may be a factor in determining the minimum or weakest value. The best example would be in audio systems where there is a background hum or hiss. This is what is known as the noise floor and in audio the weakest or minimum value is this noise level since you cannot hear any sounds that have a volume lower than the noise. So in this field taking noise as the minimum value makes sense. In the CCD definition above it seems that this is similar to the audio example and the noise is equal to the weakest signal you could measure or use. (Notice that they take the log of the ratio of max to min which is the same as subtracting the log of the minimum from the log of the maximum.)

 

There is noise in the process of making a print since information is lost or degraded as you move from scene to camera to print, but once the print is finished the concept of noise does not appear to apply unless it is the transfer of the light from the print to my eyes (which have a lot of noise) to my brain (which has even more noise). In any case this would vary from person to person and there would be nothing meaningful to measure.

 

For all practical purposes a print existing as an object separate from the process that produced it seems to be noiseless and the dynamic range is the ratio of the min and max reflectance or the difference of min and max density.

 

Finally the "number of tones" present in a print does not depend upon the dynamic range or the range or the min and max. A print is an analog image and by definition has continuous tones that flow from one to another without any step change or gaps. Just as the sun sets in a smooth motion and not in increments, so the shades of gray flow in a continuous tone print.

 

A silver print or negative is continuous tone by the nature of the chemical processes that produced it and an inkjet print is continuous tone by virtue of the dither pattern, variable droplet size, etc. that are used to produce it from a stepped digital file. By the definition of continuous tone there are no steps and any tone value between the limits of min and max can be created. Even though our eyes and instruments cannot distinguish between two tones, if they are too close together in value, does not change the fact that an infinite number of tones are available to the print maker in either silver or ink to create their image.

 

For the sake of an example let's assume that we can only distinguish a 1% change in reflectance and this is the same at all levels of light and dark. So in my earlier example of a print with a min of 4% and a max of 95% and a range of 91 percentile we would be able to distinguish between the two tones that have 4% and 5% reflectance, or 5% and 6%, or 6% and 7%, and so on. This would tend to make you think that you could only get 92 shades of gray out of this range. What you have to keep in mind is that if you can distinguish between 4% and 5% you can distinguish between 4.1% and 5.1%, and between 4.001% and 5.001%, and between 4.0000001% and 5.0000001%. Which leads to the conclusion that there are an infinite number of distinguishable tone pairs within any tonal range of a continuous tone image. If there are an infinite number of tone pairs then there must be an infinite number of tones.

 

This then comes back to my original assertion that all continuous tone photographic mediums have an infinite number of tones whether they are ink jet or silver based. Which then means that an ink jet print cannot have more tones than a silver print. They may be easier to control since the source is a digital file. It may be possible to achieve better tonal compression and map a wider real life scene into a digital file then onto film or photo paper but the end result is not more print tones.

 

Thank you for your patience if you got this far. <G>

 

Martin Wesley



[Non-text portions of this message have been removed]

Thanks Martin,

Bottom line to me is that when someone here (except for Austin, because he's
made it completely clear what he means by the term) or elsewhere says
'dynamic range' to me in reference to inkjet printing I will (safely) assume
they are talking about the range of paper base to max black, whether it's
technically correct or not. And, furthermore, that usage will be more than
adequate for conveying the thoughts and ideas under discussion the great
majority of the time.

Larry

-----Original Message-----
From: Martin Wesley [mailto:mwesley250@...]
Sent: Wednesday, March 27, 2002 1:00 AM
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Subject: [Digital BW] Dynamic Range Definitions and Print Tones


I realize this may be of little interest to most but some have expressed
interest. So here is some additional information. This is long. You have
been warned.



Since we are on the Internet a web search seemed like the thing to do. Here
is some of what I found plus an additional text reference. Be careful of
long web addresses that wrap from one line to the next. You may need to cut
and paste the pieces into your browser to get them to work.



Dynamic Range, Some Definitions:



From Merriam-Webster's Dictionary:

Main Entry: dynamic range
Function: noun
Date: 1949
: the ratio of the strongest to the weakest sound intensity that can be
transmitted or reproduced by an audio or broadcasting system

http://www.m-w.com/cgi-bin/dictionary



From NASA website:

"The dynamic range of an image is the ratio between the maximum and minimum
brightness levels in the image."

http://nasatech.com/Briefs/Aug98/NPO20254.html



From the Cornell University Library:

"DYNAMIC RANGE is the range of tonal difference between the lightest light
and darkest dark of an image."

http://www.library.cornell.edu/preservation/tutorial/intro/intro-05.html



From Digital Photography Review:

"Dynamic range is the ratio between the brightest and darkest recordable
parts of an image or scene."

http://www.dpreview.com/learn/Glossary/Digital_Imaging/Dynamic_Range_01.htm



From X-rite's "The Color Guide and Glossary"

"Dynamic Range: An instrument's range of measurable values, from the lowest
amount it can detect to the highest amount it can handle."



From Mix Magazine:

"Dynamic range, generally measured in dB, is the ratio of maximum
undistorted signal (full-scale or onset of clipping) to residual noise
(noise floor)."

http://industryclick.com/magazinearticle.asp?releaseid=5828&magazinearticlei
d=72693&siteid=15&magazineid=141



From CCD Direct:

"The dynamic range is often represented as a log ratio of well depth to the
readout noise in decibels. For example, a system with a well depth of 45,000
electrons and a readout noise of 15 electrons would have a dynamic range =
20 log (45,000/15), or 69dB."

http://www.ccddirect.com/online-store/scstore/dynamic.html



From outer space. [Caution, viewing this page may be hazardous to your
sanity!] <G>

http://utam.geophys.utah.edu/UTAMtheses/JingChen/latex_2_html/node13.html



So what we have is a variety of definitions that seem to vary depending upon
what field you are in but the common element is that they are looking for a
meaningful way to describe a relationship between the minimum and maximum as
a ratio.



I have to wander into some math here to clear up a point. Sorry but I just
don't see any way around it.



Some people say the dynamic range is the difference between the min and max
and some say it is the ratio. The interesting thing is that they are both
sometimes correct. Some things such as sound volume, CCD response and image
density are not direct measurements of a physical property but are
calculated values from actual properties given in logarithmic form.



The log of a number is the value of that number as a power of a base number
which is most commonly 10. The Log10 of 100 is 2, (10 raised to the 2nd
power) the log10 of 10,000 is 4, (10 raised to the 4th power) the log10 of
112.3 is 2.0504, (10 raised to the 2.0504th power).



Now one of the nice things about logs is that if you add them the result is
the same as multiplying the original numbers and if you subtract them the
result is the same as dividing the original numbers when you convert the log
values back to their original form. This is the reason logs were so widely
used because they reduce complex multiplication and division problems to
simple addition and subtraction. Very handy before we had pocket calculators
and computers.



Photographic materials are usually measured in density. If it is a
transparency or negative the density is the log of 1 divided by the
percentage of light that passed through the film. If it is a print the
density is the log of 1 divided by the percentage of light that was
reflected from the print.




So if a print has a white paper base reflectance of 95% (95% of the light
shining on this spot bounces off toward your eye) and the deepest black has
a reflectance of 4% (only 4% of the light falling on this dark spot bounces
off and the rest is absorbed or scattered away from your eye.) The dynamic
range would be the maximum value of 95% divided by the minimum value of 4%,
which is 23.75. Here taking the ratio is proper because we are talking about
direct measurements of a physical property and not calculated values. The
difference would be 91percentile, which is the range of the reflectance.



Usually in photography we do not talk about reflectance but instead use
density. So for this case the paper base would have a density of 0.0223 (the
base 10 log of 4%)and the darkest black 1.3979 (the base 10 log of 95%).
Since these are log values calculated from the reflectance you find the
dynamic range by subtracting them. 1.3979 minus 0.0223 is 1.3757 the dynamic
range and the range of the Density.



Now if you take this Density dynamic range and raise 10 to the power of
1.3757 you get 23.75! The same number you got by dividing the maximum
reflectance by the minimum reflectance.



So if you are talking about image density (a log value) the dynamic range
and the range are the same. If you are talking about reflectance or
transmission they are not the same. One is the ratio and the other is the
difference.



It is interesting to note that if the reflectance range is 1% to 92% the
range stays the same at 91 but the dynamic range goes up from 23.75 to 92.
If the range shifts up to 8% to 99% (still a range of 91) the dynamic range
drops to 12.38. This suggests to me that the simple range of reflectance
with the min and max values is of more interest than the dynamic range when
looking at a print or judging print medium quality.



Noise generally does not seem to be used in the calculation of the dynamic
range but it may be a factor in determining the minimum or weakest value.
The best example would be in audio systems where there is a background hum
or hiss. This is what is known as the noise floor and in audio the weakest
or minimum value is this noise level since you cannot hear any sounds that
have a volume lower than the noise. So in this field taking noise as the
minimum value makes sense. In the CCD definition above it seems that this is
similar to the audio example and the noise is equal to the weakest signal
you could measure or use. (Notice that they take the log of the ratio of max
to min which is the same as subtracting the log of the minimum from the log
of the maximum.)



There is noise in the process of making a print since information is lost or
degraded as you move from scene to camera to print, but once the print is
finished the concept of noise does not appear to apply unless it is the
transfer of the light from the print to my eyes (which have a lot of noise)
to my brain (which has even more noise). In any case this would vary from
person to person and there would be nothing meaningful to measure.



For all practical purposes a print existing as an object separate from the
process that produced it seems to be noiseless and the dynamic range is the
ratio of the min and max reflectance or the difference of min and max
density.



Finally the "number of tones" present in a print does not depend upon the
dynamic range or the range or the min and max. A print is an analog image
and by definition has continuous tones that flow from one to another without
any step change or gaps. Just as the sun sets in a smooth motion and not in
increments, so the shades of gray flow in a continuous tone print.



A silver print or negative is continuous tone by the nature of the chemical
processes that produced it and an inkjet print is continuous tone by virtue
of the dither pattern, variable droplet size, etc. that are used to produce
it from a stepped digital file. By the definition of continuous tone there
are no steps and any tone value between the limits of min and max can be
created. Even though our eyes and instruments cannot distinguish between two
tones, if they are too close together in value, does not change the fact
that an infinite number of tones are available to the print maker in either
silver or ink to create their image.



For the sake of an example let's assume that we can only distinguish a 1%
change in reflectance and this is the same at all levels of light and dark.
So in my earlier example of a print with a min of 4% and a max of 95% and a
range of 91 percentile we would be able to distinguish between the two tones
that have 4% and 5% reflectance, or 5% and 6%, or 6% and 7%, and so on. This
would tend to make you think that you could only get 92 shades of gray out
of this range. What you have to keep in mind is that if you can distinguish
between 4% and 5% you can distinguish between 4.1% and 5.1%, and between
4.001% and 5.001%, and between 4.0000001% and 5.0000001%. Which leads to the
conclusion that there are an infinite number of distinguishable tone pairs
within any tonal range of a continuous tone image. If there are an infinite
number of tone pairs then there must be an infinite number of tones.



This then comes back to my original assertion that all continuous tone
photographic mediums have an infinite number of tones whether they are ink
jet or silver based. Which then means that an ink jet print cannot have more
tones than a silver print. They may be easier to control since the source is
a digital file. It may be possible to achieve better tonal compression and
map a wider real life scene into a digital file then onto film or photo
paper but the end result is not more print tones.



Thank you for your patience if you got this far. <G>



Martin Wesley



[Non-text portions of this message have been removed]



Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:

http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint

Please follow these basic guidelines:
- Include your full name with your message.
- Include the address of your website, if you have one.
- As threads develop, trim off excess portions of earlier messages to keep
them short.
- As the topic of a thread changes remember to change the subject header.
- Good manners are required at all time. No personal attacks or "flames."
- Complete your Yahoo profile.
- Before posting a question, search the message archives and the various
resources on the homepage.




Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

> Thanks Martin,
>
> Bottom line to me is that when someone here (except for Austin,
> because he's
> made it completely clear what he means by the term) or elsewhere says
> 'dynamic range' to me in reference to inkjet printing I will
> (safely) assume
> they are talking about the range of paper base to max black, whether it's
> technically correct or not. And, furthermore, that usage will be more than
> adequate for conveying the thoughts and ideas under discussion the great
> majority of the time.
>
> Larry

Hi Larry,

Yes, but Martin's explanation and understanding are incorrect.  Taking two
measurements and subtracting them gives you a STATIC range, NOT a DYNAMIC
range.  Using the term dynamic range as Martin has, and you state above is
simply incorrect.  Why NOT use the correct term "density range"?

As has been shown, you can have a very high density range (a high black
value on a very white paper, with no intermediate tones) and that has a very
LOW dynamic range, because there are no tones in-between.  Don't let this
simple concept slip by...it's important.

Austin

Hi Austin,

Your right about all this of course, but I dont think 
it's going to change what people mean when they use the 
term. I guess I'm used to it, in my industry there are 
several definitions of snr, all different from the 
textbook version. Most of the time when someone uses 
that term with me it's to tell me that something is good 
or bad, and I dont need to know more than that. When the 
situation warrants it I'll ask very specific questions 
to be sure I'm getting the right info. And, yes, over 
the years I've been burned a few times because of the 
ambiguity when I wasnt careful enough. Same thing 
applies here, except for right now, because you've 
diligently put out all this info and explanation, but in 
a week or so all that will be forgotten. Just my 2 cents.

Larry

> > Thanks Martin,
> >
> > Bottom line to me is that when someone here (except for Austin,
> > because he's
> > made it completely clear what he means by the term) or elsewhere says
> > 'dynamic range' to me in reference to inkjet printing I will
> > (safely) assume
> > they are talking about the range of paper base to max black, whether it's
> > technically correct or not. And, furthermore, that usage will be more than
> > adequate for conveying the thoughts and ideas under discussion the great
> > majority of the time.
> >
> > Larry
> 
> Hi Larry,
> 
> Yes, but Martin's explanation and understanding are incorrect.  Taking two
> measurements and subtracting them gives you a STATIC range, NOT a DYNAMIC
> range.  Using the term dynamic range as Martin has, and you state above is
> simply incorrect.  Why NOT use the correct term "density range"?
> 
> As has been shown, you can have a very high density range (a high black
> value on a very white paper, with no intermediate tones) and that has a very
> LOW dynamic range, because there are no tones in-between.  Don't let this
> simple concept slip by...it's important.
> 
> Austin
> 
> 
> 
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and other 
> resources as they are often being updated. The page is at:
> 
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
> 
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep them 
> short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or "flames."
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various 
> resources on the homepage. 
> 
> 
>  
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 
> 
>

Hi Martin,

> Dynamic Range, Some Definitions:

The definition/concept of dynamic range is taught in any good EE program,
and is very basic to most electrical engineers (NOT computer engineers, they
are not the same).  I have probably measured the dynamic range, and written
the specs for, over 100 different systems.  I am an EE, and, as you know,
have been designing both audio and video equipment for over 25 years.  I
don't mean to sound smug, but it is unquestionable that my
definition/understanding of dynamic range is correct.  If it was not, since
I do this for a living, and am subject to direct peer review by hundreds of
different engineers over my career, as well as thousands of customers and
people who read my specs...and if I had it wrong, someone would have pointed
it out by now...and no one has.

Most of the definitions you cited are "pedestrian" definitions written by
non-technical people who are, for the most part, mis-explaining a concept
that they apparently do not understand.  What you are calling "dynamic
range" is merely static range.  Measuring the max and min value of something
merely gives you the STATIC range, NOT the DYNAMIC range.  Comparing them by
division or subtraction still give you a STATIC result, not a DYNAMIC
result.

> So what we have is a variety of definitions that seem to vary
> depending upon what field you are in

They vary simply because the people who wrote the incorrect definitions did
not understand what it was they were saying.  Dynamic range IS dynamic
range, no matter what the application.  The X-Rite definition you supplied
is the best one as far as I am concerned.

> but the common element is
> that they are looking for a meaningful way to describe a
> relationship between the minimum and maximum as a ratio.

That's the issue, they do not describe what they mean by minimum and
maximum...  They are ambiguous terms.  These terms can mean either amplitude
or static values.  Per the diagram I posted in another post, they clearly
mean amplitude.

> Some people say the dynamic range is the difference between the
> min and max and some say it is the ratio.

It is always a ratio, as defined by any dynamic range equation.  Forget web
and dictionary definitions, it's the equation that matters here, and only
the equation.  The equation if finite and definitive.  Your understanding
does not follow the equation.

The dynamic range equation from "Digital Signal Processing in VLSI" by
Richard J. Higgins is:

Dynamic Range (dB) = log10 (largest signal/smallest discernable signal)

The diagram I posted previously is from this very book/equation.  It clearly
shows that largest and smallest are amplitude.  Also note DISCERNABLE.  To
discern something means it must change with respect to what it is you are
trying to discern it from, which in this case is it self.  You need two data
points for "discernability".  You can not discern something that is static
(shows no change).

> Some things such as
> sound volume, CCD response and image density are not direct
> measurements of a physical property but are calculated values
> from actual properties given in logarithmic form.

Absolutely not true.  You absolutely can measure the dynamic range of audio
systems and CCDs and image density.  You CAN run an FFT on the data to
arrive at the dynamic range, but you still have to do direct measurements to
arrive at the values you give to the FFT...you don't just make them up!

> So if a print has a white paper base reflectance of 95% (95% of
> the light shining on this spot bounces off toward your eye) and
> the deepest black has a reflectance of 4% (only 4% of the light
> falling on this dark spot bounces off and the rest is absorbed or
> scattered away from your eye.) The dynamic range would be the
> maximum value of 95% divided by the minimum value of 4%, which is
> 23.75.

Because density is represented in a log form, and dynamic range is
represented in a log form does not make them both the same.  Dynamic range
and density range are NOT the same.

How do you know you really measured 95% and not 95.876593254?

> Usually in photography we do not talk about reflectance but
> instead use density. So for this case the paper base would have a
> density of 0.0223 (the base 10 log of 4%)and the darkest black
> 1.3979 (the base 10 log of 95%). Since these are log values
> calculated from the reflectance you find the dynamic range by
> subtracting them. 1.3979 minus 0.0223 is 1.3757 the dynamic range
> and the range of the Density.

You have NOT measured any dynamic range, as you have NOT measured the noise
in the system.  What you have is the DENSITY range NOT the DYNAMIC range.
Again, you can NOT relate dynamic range to density range simply because they
are expressed in log form.

> So if you are talking about image density (a log value) the
> dynamic range and the range are the same.

No, absolutely wrong.  Because two things use the same numeric
representation it does not mean they ARE the same.  Dynamic range is
measured in DECIBELS, density is measured NOT in DB, but on a calibrated
relative density scale.  NO density measurement equation will contain DB.
That fact alone negates your entire premise that they are the same.

OK, here is the Density measurement equation from "Introduction to
Densitometry" published by the Graphics Communication Association:

Density = log10 1/R (where R is the reflectance)

and to quote the book:

"This is a pure definition, which means that it defines the relationship
between light reflectance and density.  This definition numerically
interprets density, and, more importantly, describes density in a manner
that approximates the way in which the human eye sees objects."

No where in this entire book do they talk about dynamic range.  Also,
density measurements REQUIRE they be compared to a known calibrated sample,
which is why you have to calibrate densitometers.  Measuring dynamic range
does NOT require it be calibrated to any known sample, it is simply relative
in and of it self.  Density measurements aren't, they are relative to an
absolute measurement...hence, they are STATIC.

> Noise generally does not seem to be used in the calculation of
> the dynamic range

Absolutely wrong.  The "smallest discernable signal" in any analog systems
IS the noise in the system.  In a digital system, as in number of bits let's
say, the minimum discernable signal is 1, so 4 bits has a dynamic range of
log10(16/1) or 1.2dB.  The key is "smallest DISCERNABLE" signal, what ever
that might be in that particular system.

> The best example would be in audio
> systems where there is a background hum or hiss. This is what is
> known as the noise floor and in audio the weakest or minimum
> value is this noise level since you cannot hear any sounds that
> have a volume lower than the noise.

Again, you are using ambiguous terms here (weakest/minimum).  The weakest
value is NOT necessarily the noise floor.  You can turn the volume down, and
decrease the noise to almost nothing, and that also doesn't take into
consideration the noise before the volume control.  Also, the noise will be
different depending on the volume, as when you add gain to a system, you
also amplify the noise.

> There is noise in the process of making a print since information
> is lost or degraded as you move from scene to camera to print,
> but once the print is finished the concept of noise does not
> appear to apply

Of course noise applies.  Anything that as distorted the actual image
information has caused noise, anywhere in the system.  It is not relevant if
the print is "finished" or not.

> Finally the "number of tones" present in a print does not depend
> upon the dynamic range or the range

Absolutely wrong...as I've said time and time again, and for some reasons
you don't understand my examples.  The number of tones is SOLELY based on
the dynamic range.  It's the definition OF dynamic range (as defined by any
competent source).

How many tones does 8 bits represent?  256.  How many tones does 9 bits
represent? 512.  Now, you can represent ANY density range with only two
bits, as I have explained in another post.  You can represent any density
range with ANY number of bits.  It is merely a representation.  BUT the
dynamic range will be entirely different, but the density range will be the
same.

> A print
> is an analog image and by definition has continuous tones that
> flow from one to another without any step change or gaps.

Absolutely wrong, again.  NO analog system is completely continuous.  There
is NO such thing.  ALL systems have a resolution analog or not!  It's a
fact.  Because you don't understand that it doesn't doesn't make it not
true, it just means you don't understand/get that.

> Just as
> the sun sets in a smooth motion and not in increments, so the
> shades of gray flow in a continuous tone print.

Your ability to SEE the sun setting absolutely is in increments.  It is well
knows that your vision has a resolution, and you can only discern things
with the accuracy of so much of an arc.  It's a fact.

Quoted from "Modern Optical Engineering" by Warren Smith:

"The resolution of the eye was AT BEST about one minute of arc."

Given that you can NOT discern any movement of the sun/earth smaller than
one minute of arc.

One thing you are missing that everything has resolution...everything, even
all analog systems.

> This then comes back to my original assertion that all continuous
> tone photographic mediums have an infinite number of tones
> whether they are ink jet or silver based.

And that assertion is flawed.  ALL analog systems have resolution, and
therefore have a number of discernable tones, and are NOT continuous.

> Thank you for your patience if you got this far. <G>

I have tried, and don't believe I can not explain anything further to you.
It is not a matter of disagreeing, it is a matter of what you are saying is
just wrong, and is based in ambiguous definitions and flawed assertions, and
you don't want to understand that.

I don't know what you do for a living, and what your background is.  I am
guessing software?

If you would like to ask another expert in the field, I am very happy to
suggest someone to you.  Perhaps they can explain it better to you than I
have been able to.

Regards,

Austin

Hi Larry,

> in my industry there are
> several definitions of snr, all different from the
> textbook version.

You do know that dynamic range and MAXIMUM SNR end up being the same ;-)

> Most of the time when someone uses
> that term with me it's to tell me that something is good
> or bad, and I dont need to know more than that.

Understood.  But, Martin is arguing that my definition of dynamic range is
wrong, and I know it isn't.  Kind of like "resolution" of a monitor.  It is
an incorrectly used term, but as you say, everyone uses it, and understands
what it means...though they may not understand it's misused, and that 1024 x
768 is NOT a resolution.

Regards,

Austin

Austin,

I'm trying hard to follow along with you guys, but these equations loose me.
Is it that the main point over which you and Martin disagree is whether or
not noise is part of the equation?

You make some excellent points toward the end of your post, maybe because
you use words to make them rather than math <g>, but then I thought Martin
did too, so, so much for what I think. ;-)

One thing I request though: this is turning into a classic debate, one I'd
like to see continue to some kind of resolution if possible, or at least
continue in a friendly way, on both sides, and you do get a bit burley
toward the end. I understand your frustration when people don't "want to
understand the truth" in what you say, but please understand this isn't just
Martin you are trying to convince, there is an entire establishment of photo
and computer gurus who represent the opinions Martin is representing. Don't
expect to turn that establishment on it's ear overnight.

In fact, the extent to which there is an establishment that seems in
contradiction to your definition makes me wonder if this isn't simply a case
where each field uses it's own jargon, and in the parlance of these
different fields the term dynamic range doesn't have two different meanings?

> Of course noise applies.  Anything that as distorted the actual image
> information has caused noise, anywhere in the system.  It is not relevant if
> the print is "finished" or not.

This confuses me. What exactly is the noise in a finished print? What I'm
getting at is where in the process is the noise accounted for? Is the noise
you speak of the noise that is introduced when making a print from the
generation that proceeds it (file for digital, film for traditional), or the
noise of the instrument that is measuring the density range?

I guess what I'm wondering is when discussing the dynamic range of a
finished print one even needs to consider what the dynamic range of the
original was, and therefor the noise incurred between the two. If one
doesn't, and I don't believe they should (meaning one should be able to
assess the Dynamic range of something for what it is without regard for
where it came from), then what IS the noise you speak of, and how do you
measure it? If the noise is that of the measuring instrument, who cares
about that, that's a little too "dynamic", if you know what I mean...

So, long story short, please clarify the noise issue as it relates to inkjet
and traditional prints, which seems to be so relevant to your definition of
dynamic range. 

Thanks,
Todd

----- Original Message ----- 

From: "Larry Roohr" <lrryr@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, March 27, 2002 6:45 AM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones

(snip)


> And, furthermore, that usage will be more than
> adequate for conveying the thoughts and ideas under discussion the great
> majority of the time.
> 
 Larry,

Exactly. It tells you what you need to know about a print or print media.

Martin
> 
> 
> 
(snip)

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, March 27, 2002 7:18 AM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


(snip)
>
> Hi Larry,
>
> Yes, but Martin's explanation and understanding are incorrect.  Taking two
> measurements and subtracting them gives you a STATIC range, NOT a DYNAMIC
> range.  Using the term dynamic range as Martin has, and you state above is
> simply incorrect.  Why NOT use the correct term "density range"?

Austin,

They are not my explanations they are the commonly accepted definitions by a
wide range of people in many fields. I do not see the need to reinvent these
meanings. Dynamic range is a ratio but subtracting log values is
mathematically the same as taking the ratio. I don't see the need to
reinvent the math either.
>
> As has been shown, you can have a very high density range (a high black
> value on a very white paper, with no intermediate tones) and that has a
very
> LOW dynamic range, because there are no tones in-between.  Don't let this
> simple concept slip by...it's important.

Where in the definitions was this stated? Dynamic range tells you nothing
about the number of intermediate values.

Martin

----- Original Message -----

From: Austin Franklin
To: DigitalBlackandWhiteThePrint@yahoogroups.com
Sent: Wednesday, March 27, 2002 8:51 AM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


Hi Martin,

>> Dynamic Range, Some Definitions:

>The definition/concept of dynamic range is taught in any good EE program,
>and is very basic to most electrical engineers (NOT computer engineers,
they
>are not the same).  I have probably measured the dynamic range, and written
>the specs for, over 100 different systems.  I am an EE, and, as you know,
>have been designing both audio and video equipment for over 25 years.  I
>don't mean to sound smug, but it is unquestionable that my
>definition/understanding of dynamic range is correct.  If it was not, since
>I do this for a living, and am subject to direct peer review by hundreds of
>different engineers over my career, as well as thousands of customers and
>people who read my specs...and if I had it wrong, someone would have
pointed
>it out by now...and no one has.

>Most of the definitions you cited are "pedestrian" definitions written by
>non-technical people who are, for the most part, mis-explaining a concept
>that they apparently do not understand.  What you are calling "dynamic
>range" is merely static range.  Measuring the max and min value of
something
>merely gives you the STATIC range, NOT the DYNAMIC range.  Comparing them
by
>division or subtraction still give you a STATIC result, not a DYNAMIC
>result.

I guess they probably have a shortage of technical people at Cornell and
NASA to write up these definitions. <G>

You ask me to believe that all these other sources are wrong and you are
right. I don't mean to offend but I just can't do that. As I have said
before, you seem to be taking mathematical concepts from EE and assuming
that they directly apply to non-EE systems.

>> So what we have is a variety of definitions that seem to vary
>> depending upon what field you are in

>They vary simply because the people who wrote the incorrect definitions did
>not understand what it was they were saying.  Dynamic range IS dynamic
>range, no matter what the application.  The X-Rite definition you supplied
>is the best one as far as I am concerned.

All these people are wrong and you are correct? Austin, please. The X-Rite
one is good for a measuring instrument.

>> but the common element is
>> that they are looking for a meaningful way to describe a
>> relationship between the minimum and maximum as a ratio.

>That's the issue, they do not describe what they mean by minimum and
>maximum...  They are ambiguous terms.  These terms can mean either
amplitude
>or static values.  Per the diagram I posted in another post, they clearly
>mean amplitude.

The nature and the determination of the min and max will vary with what
physical property is under examination. The words "minimum" and "maximum"
are perfectly clear.

>> Some people say the dynamic range is the difference between the
>> min and max and some say it is the ratio.

>It is always a ratio, as defined by any dynamic range equation.

Austin, I cannot believe that you do not understand the difference of the
log of the max and the log of the min is the ratio of the max and min.

>  Forget web
>and dictionary definitions,

I should throw out the dictionary definitions? I don't see any reason to do
that. If that is what is required to prove your point you would seem to be
on thin ice.

> it's the equation that matters here, and only
>the equation.  The equation if finite and definitive.  Your understanding
>does not follow the equation.

>The dynamic range equation from "Digital Signal Processing in VLSI" by
>Richard J. Higgins is:

>Dynamic Range (dB) = log10 (largest signal/smallest discernable signal)

If I say Dynamic Range (dB) = log10(largest meaningful signal) -
log10(smallest meaningful signal) this is mathematically the same and if you
plug in the numbers you get the same answer.

>The diagram I posted previously is from this very book/equation.  It
clearly
>shows that largest and smallest are amplitude.  Also note DISCERNABLE.  To
>discern something means it must change with respect to what it is you are
>trying to discern it from, which in this case is it self.  You need two
data
>points for "discernability".  You can not discern something that is static
>(shows no change).

The criteria for determining or discerning the min and max values will vary
from field to field and may be different at opposite ends of the scale. In
audio the minimum is determined in relation to the background noise floor
but the maximum is chosen based on a certain degree of distortion. I am sure
what you say and your diagram are excellent relationships to use in
evaluating electrical systems but are not a necessarily a good way to
determine the min and max in all systems. For instance "noise floor" and
"third harmonic distortion" would not be helpful in deciding what the usable
min and max tones in a photographic print.

It is more important to talk about meaningful values rather than simply
discernable values. You have to look at the system and decide what is of
interest and produces usable information. This may be different from what is
discernable.

>> Some things such as
>> sound volume, CCD response and image density are not direct
>> measurements of a physical property but are calculated values
>> from actual properties given in logarithmic form.

>Absolutely not true.  You absolutely can measure the dynamic range of audio
>systems and CCDs and image density.  You CAN run an FFT on the data to
>arrive at the dynamic range, but you still have to do direct measurements
to
>arrive at the values you give to the FFT...you don't just make them up!

I am sorry but dynamic range is not measured, it is a calculated value.
Didn't you just give us an equation to do this? dB in audio are calculated
values derived from a direct measurement of pressure. This is where you are
missing my point. Some values are so common we have lost sight of the fact
that they do no represent actual physical properties. Print density is not a
physical property. It is a value calculated from a measurement of light flux
(a real physical property) by an instrument which does an internal
calculation and gives you a calculated density value.

In order for us to make a decision about the significance of the min and max
values we need to understand how they were derived.

>> So if a print has a white paper base reflectance of 95% (95% of
>> the light shining on this spot bounces off toward your eye) and
>> the deepest black has a reflectance of 4% (only 4% of the light
>> falling on this dark spot bounces off and the rest is absorbed or
>> scattered away from your eye.) The dynamic range would be the
>> maximum value of 95% divided by the minimum value of 4%, which is
>> 23.75.

>Because density is represented in a log form, and dynamic range is
>represented in a log form does not make them both the same.  Dynamic range
>and density range are NOT the same.

They are because of the math. The difference between two log values is a
ratio of the root numbers.

> How do you know you really measured 95% and not 95.876593254?

I did not measure anything. I asked that you assume actual values not
measurements. There is a difference between the real value and the
instrument readings.

>> Usually in photography we do not talk about reflectance but
>> instead use density. So for this case the paper base would have a
>> density of 0.0223 (the base 10 log of 4%)and the darkest black
>> 1.3979 (the base 10 log of 95%). Since these are log values
>> calculated from the reflectance you find the dynamic range by
>> subtracting them. 1.3979 minus 0.0223 is 1.3757 the dynamic range
>> and the range of the Density.

>You have NOT measured any dynamic range, as you have NOT measured the noise
>in the system.  What you have is the DENSITY range NOT the DYNAMIC range.
>Again, you can NOT relate dynamic range to density range simply because
they
>are expressed in log form.

Once again you make up your own definitions and ignore the published ones.
Noise is not required to calculate the dynamic range, only the meaningful
min and max.

You have repeatedly refused to answer my question as to what you think the
noise is in a photographic print.

>> So if you are talking about image density (a log value) the
>> dynamic range and the range are the same.

>No, absolutely wrong.  Because two things use the same numeric
>representation it does not mean they ARE the same.  Dynamic range is
>measured in DECIBELS, density is measured NOT in DB, but on a calibrated
>relative density scale.  NO density measurement equation will contain DB.
>That fact alone negates your entire premise that they are the same.

Density is calculated not measured. A densitometer measures the light flux
reflecting from or passing through the material. It then takes the
difference between this flux and its standard flux. It then converts the
flux to a percentage, divides it into 1, takes the log10 of the result and
displays it.

>OK, here is the Density measurement equation from "Introduction to
>Densitometry" published by the Graphics Communication Association:

>Density = log10 1/R (where R is the reflectance)

I believe I have stated that several times now.

>and to quote the book:

>"This is a pure definition, which means that it defines the relationship
>between light reflectance and density.  This definition numerically
>interprets density, and, more importantly, describes density in a manner
>that approximates the way in which the human eye sees objects."

Yes because our eyes do not respond to light in a linear fashion. Density is
a convenient way to look at reflectance but it is a calculated value not a
direct measurement.

>No where in this entire book do they talk about dynamic range.  Also,
>density measurements REQUIRE they be compared to a known calibrated sample,
>which is why you have to calibrate densitometers.  Measuring dynamic range
>does NOT require it be calibrated to any known sample, it is simply
relative
>in and of it self.  Density measurements aren't, they are relative to an
>absolute measurement...hence, they are STATIC.

Of course density is static. (Unless your prints are fading of course.<G>)
You would then seem to be saying that a photographic print does not have a
dynamic range.

>> Noise generally does not seem to be used in the calculation of
>> the dynamic range

>Absolutely wrong.  The "smallest discernable signal" in any analog systems
>IS the noise in the system.  In a digital system, as in number of bits
let's
>say, the minimum discernable signal is 1, so 4 bits has a dynamic range of
>log10(16/1) or 1.2dB.  The key is "smallest DISCERNABLE" signal, what ever
>that might be in that particular system.

Austin, this is just a semantic game. It just so happens that the smallest
meaningful signal in an analog system is equal to the noise floor. Again
what is the noise in a static photographic print and how does it relate to
either reflectance or density?

>> The best example would be in audio
>> systems where there is a background hum or hiss. This is what is
>> known as the noise floor and in audio the weakest or minimum
>> value is this noise level since you cannot hear any sounds that
>> have a volume lower than the noise.

>Again, you are using ambiguous terms here (weakest/minimum).  The weakest
>value is NOT necessarily the noise floor.  You can turn the volume down,
and
>decrease the noise to almost nothing, and that also doesn't take into
>consideration the noise before the volume control.  Also, the noise will be
>different depending on the volume, as when you add gain to a system, you
>also amplify the noise.

The entire idea, as I clearly stated, is to use the smallest meaningful
value which is at the noise floor in this example. I see no ambiguity in
that.

>> There is noise in the process of making a print since information
>> is lost or degraded as you move from scene to camera to print,
>> but once the print is finished the concept of noise does not
>> appear to apply

>Of course noise applies.  Anything that as distorted the actual image
>information has caused noise, anywhere in the system.  It is not relevant
if
>the print is "finished" or not.

Noise is only relevant in describing the output to the input in this case.
Once we are left with just the output hanging on the wall what is the noise?

>> Finally the "number of tones" present in a print does not depend
>> upon the dynamic range or the range

>Absolutely wrong...as I've said time and time again, and for some reasons
>you don't understand my examples.  The number of tones is SOLELY based on
>the dynamic range.  It's the definition OF dynamic range (as defined by any
>competent source).

The number of tones is only related to the dynamic range in digital systems
which have finite steps. Analog systems are stepless and do not have a
finite number of tones.

I chose my sources and I stick with them.

>How many tones does 8 bits represent?  256.  How many tones does 9 bits
>represent? 512.  Now, you can represent ANY density range with only two
>bits, as I have explained in another post.  You can represent any density
>range with ANY number of bits.  It is merely a representation.  BUT the
>dynamic range will be entirely different, but the density range will be the
>same.

 This is true in digital but does not relate to analog which has no bits.

>> A print
>> is an analog image and by definition has continuous tones that
>> flow from one to another without any step change or gaps.

>Absolutely wrong, again.  NO analog system is completely continuous.  There
>is NO such thing.  ALL systems have a resolution analog or not!  It's a
>fact.  Because you don't understand that it doesn't doesn't make it not
>true, it just means you don't understand/get that.

Austin, you choose not to acknowledge that the word continuous means
unbroken or that analog is defined as "a mechanism in which data is
represented by continuously variable physical quantities."

Once again if you want to reinvent the definition of words we have no basis
for a meaningful dialog.

>> Just as
>> the sun sets in a smooth motion and not in increments, so the
>> shades of gray flow in a continuous tone print.

>Your ability to SEE the sun setting absolutely is in increments.  It is
well
>knows that your vision has a resolution, and you can only discern things
>with the accuracy of so much of an arc.  It's a fact.

How I see it and what it does are hardly the same thing. As I have said
before, you keep wanting to equate the measurement of a property to the
property itself and this is not correct.

>Quoted from "Modern Optical Engineering" by Warren Smith:

>"The resolution of the eye was AT BEST about one minute of arc."

>Given that you can NOT discern any movement of the sun/earth smaller than
>one minute of arc.

>One thing you are missing that everything has resolution...everything, even
>all analog systems.

NO! NO! NO! (sorry for shouting) Measurements of things have resolutions.
The things being measured DO NOT have resolution.

>> This then comes back to my original assertion that all continuous
>> tone photographic mediums have an infinite number of tones
>> whether they are ink jet or silver based.

>And that assertion is flawed.  ALL analog systems have resolution, and
>therefore have a number of discernable tones, and are NOT continuous.

No, analog systems do not have resolution because by definition they are
continuous. The measurements of continuous analog systems do have resolution
because they are not made with perfect instruments and the readings are
expressed in discrete mathematical values.

>> Thank you for your patience if you got this far. <G>

>I have tried, and don't believe I can not explain anything further to you.
>It is not a matter of disagreeing, it is a matter of what you are saying is
>just wrong, and is based in ambiguous definitions and flawed assertions,
and
>you don't want to understand that.

I suppose that if you cannot explain your position within the framework of
standard definitions and must resort to rewriting or refuting them, you may
have no recourse but to simply state, "You are wrong!" which is seldom a
compelling argument.

>I don't know what you do for a living, and what your background is.  I am
>guessing software?

Should have gone into computer science but my degree is in Chemical
Engineering. Five years teaching high school math and science. 17 years in
chemical process and project engineering.

>If you would like to ask another expert in the field, I am very happy to
>suggest someone to you.  Perhaps they can explain it better to you than I
>have been able to.

Unfortunately, I feel I understand the matter reasonably well for my
purposes. I had hoped that providing other sources would help you understand
what I was saying. The dialogue has helped me to clarify and confirm my
thoughts.

Thank you,
Martin

Austin,

Congratulations on what seems to be an excellent explanation of this
subject. I say 'seems to be' just to cover the fact that I'm not sure that I
understand every nuance of your explanation.

However, you missed one thing off the end - "QED".

Bob Frost.


----- Original Message -----

From: "Austin Franklin" <darkroom@...>


>
> And that assertion is flawed.  ALL analog systems have resolution, and
> therefore have a number of discernable tones, and are NOT continuous.

Oh dear,

I thought I'd just understood this argument!

Back to color printing I think.

Bob Frost

----- Original Message -----
From: "Martin Wesley" <mwesley250@...>


>
> Unfortunately, I feel I understand the matter reasonably well for my
> purposes. I had hoped that providing other sources would help you
understand
> what I was saying. The dialogue has helped me to clarify and confirm my
> thoughts.

On 3/27/02 mwesley250@... wrote:

>For all practical purposes a print existing as an object separate from
>the process that produced it seems to be noiseless and the dynamic range
>is the ratio of the min and max reflectance or the difference of min and
>max density.

no, by the definitions you cite, the variation in the reflectance of the
substrate (paper) is the noise floor. Since the paper is not perfectly
white, there is a limit to how delicate the highlights can be before they
are more delicate than the variations of the paper itself. Therefore the
choice of paper may have a large effect on the dynamic range of the
printing system.

-- 
John Brownlow

http://www.pinkheadedbug.com

> Austin,
>
> They are not my explanations they are the commonly accepted
> definitions by a
> wide range of people in many fields.

Martin,

Commonly accepted by whom?  Merriam Webster, magazines, etc. are hardly good
sources for technical definitions.  As I've said, the definitions you cited,
with the exception of a couple of them, require interpretation of ambiguous
terms.  They are NOT wrong, just open to (mis)interpretation.  Specifically
two terms used, smallest and largest (or what ever words were used).  I have
shown VERY CLEARLY what is meant by smallest and largest when used with
respect to dynamic range, as defined by a very definitive source.

> I do not see the need to
> reinvent these
> meanings.

Me either, and clarifying ambiguous terminology is hardly reinventing
anything.  I have merely cited the clarifications from reliable sources.
You have not attempted at all to cite any clarification to the terms that
are in question.

> Dynamic range is a ratio but subtracting log values is
> mathematically the same as taking the ratio. I don't see the need to
> reinvent the math either.

Me either, and I certainly am not "reinventing" math.  Just because two
"measurements" are expressed in log form does NOT mean they are the same.
Dynamic range stands in and of it self as ONE log10 number.  It is expressed
in dB.  Density range requires TWO log10 values, dMin and dMax.  Density
values are NOT expressed in any "scale" simply because they are relative
unto themselves.

> > As has been shown, you can have a very high density range (a high black
> > value on a very white paper, with no intermediate tones) and that has a
> very
> > LOW dynamic range, because there are no tones in-between.
> Don't let this
> > simple concept slip by...it's important.
>
> Where in the definitions was this stated?

What's "this" that you are questioning?  The measurement for density range
is clear, dMax - dMin.  The measurement for dynamic range is clearly defined
by the dynamic range equation.  Here, I'll apply some simple numbers to
this:

dMax = 1.8
dMin = .2

Density Range (dRange) = 1.6

Easy, right?

What's the dynamic range?  Well, we don't know what the smallest discernable
signal is, now do we?  Nor do we really know the largest that the
paper/system can attain either.  This makes for a quandary.

The dynamic range is based on the largest ATTAINABLE signal for the medium,
not the largest for a single print.  You certainly could just print the
blackest black you could, and get it from that, not too tough.  Let's say it
was 1.8...for sake of argument.

The smallest discernable signal isn't dMin, since we could possibly be able
to "discern" in .01 density value steps...  and without knowing what the
minimum discernable signal is, we don't know what the dynamic range is, by
definition.

Let's say our smallest discernable signal was 0.01 density value (which is
hardly far fetched, and actually a reasonable number).  The dynamic range
would be 10log10((1.8-0.2)/.01) or 22dB.  Note that even in Bells, 2.2 is
not the same as 1.6.  Of course there happens to be a value for the smallest
discernable signal that would make dynamic range and density range the
same...but that's the same as saying a clock is right twice a day...it's
purely by happenstance.

> Dynamic range tells you nothing
> about the number of intermediate values.

That's what you claim, and I keep telling you, you are mistaken.

It really would help if YOU provide an EQUATION, with appropriate
definitions of terms used in the equation (as I have done), NOT some
ambiguous verbiage that is open to interpretation.


Austin

On 3/27/02 mwesley250@... wrote:

>Finally the "number of tones" present in a print does not depend upon the
>dynamic range or the range or the min and max. A print is an analog image
>and by definition has continuous tones that flow from one to another
>without any step change or gaps. Just as the sun sets in a smooth motion
>and not in increments, so the shades of gray flow in a continuous tone print.
>
>A silver print or negative is continuous tone by the nature of the
>chemical processes that produced it and an inkjet print is continuous
>tone by virtue of the dither pattern, variable droplet size, etc. that
>are used to produce it from a stepped digital file. By the definition of
>continuous tone there are no steps and any tone value between the limits
>of min and max can be created. Even though our eyes and instruments
>cannot distinguish between two tones, if they are too close together in
>value, does not change the fact that an infinite number of tones are
>available to the print maker in either silver or ink to create their image.

Both of these statements are true and not true. In a sort of general
hand-wavy practical way they are true, but they are inaccurate when you
get close to them.

1. No analog systems are continuous at the quantum level. Not even the
sun crossing the sky. Quantum noise is a limiting factor in many micro-
electronic systems.

2. A silver print or negative is not continuous at the granular level in
terms of its representation of the scene.

3. An inkjet print is certainly not continuous since each channel is
(currently) only capable of representing 256 shades of gray.

All of these things appear continuous if you squint hard enough but we
are not squinting in this discussion.

It is *absolutely not true* that an inky printmaker has an infinite
number of tones available. All bw output processes currently only output
8 bits or 256 tones. A quadtone system is *theoretically* capable of
representing many more tones... 256 to the power 4 in fact, which is a
hell of a lot, but no drivers that I am aware of address this ability.




-- 
John Brownlow

http://www.pinkheadedbug.com

> It is *absolutely not true* that an inky printmaker has an infinite
> number of tones available. All bw output processes currently only output
> 8 bits or 256 tones. A quadtone system is *theoretically* capable of
> representing many more tones... 256 to the power 4 in fact, which is a
> hell of a lot, but no drivers that I am aware of address this ability.

Hi Johnny,

The PiezoPro supposedly takes 16 bit data, as does Piezo V.6...and it also
works with 6 tone printers...  People have reported on the Piezo list that
they notice a difference in using 16 bit files.

The number of tones in a halftone system is limited by the "cell" size of
the halftone algorithm (stochastic halftone algorithms aside for simplicity
of discussion), as well as the number of inks.  The cell size is a practical
limit dependant on viewability and dot size.  If you have a halftone cell
256 x 256, with a quad ink system, you are capable of 256k tones...but, you
would have a very funny print up close!  Stochastics would do a much better
job of this.

Now, whether there is any advantage to having more tones, since the human
eye can only distinguish ~100 or so tones in any given static light...is
another issue.  Perhaps a variable light source to enhance tonality ;-)

Austin

BTW, your innate understanding of things completely out of your realm never
ceases to amaze me!

Hi Bob,

Well, Martin is a very bright guy, and I respect his views, experience and
opinion on most every subject...but on this issue, we obviously disagree...

The basic equation for dynamic range stands in and of it self, no matter
what pedestrian definitions people want to "interpret" to it.

Dynamic Range (dB) = 10log10 (largest signal/smallest discernable signal)

The diagram I provided:

http://www.darkroom.com/Images/DynamicRange01.jpg

provided unambiguous definitions for largest signal and smallest discernable
signal, which are the only two terms that need definition, since the rest of
the equation is solely based on these two definitions.

Therefore, dynamic range is basically a measurement of the number of
different discernable "signals" within a particular range, and in the case
of print images, discernable tones.

Density range and dynamic range are not the same.  Density range is purely a
static measurement of the dMax - dMin.  It has nothing to do with dynamic
range (or they would have called it dynamic range ;-), as dynamic range, by
definition, takes into account the "smallest discernable signal", and
density range does not.

dMin is not the same as the "smallest discernable signal", as dMin can be
way above the noise floor.  This is obvious in negative film, since the
density of the film base is substantially larger than the density values
you'll get between "discernable" tones in the image.

Basically, I believe that's what I claim in a nut shell, and the basic
points Martin "disagrees" with.

Regards,

Austin

> Oh dear,
>
> I thought I'd just understood this argument!
>
> Back to color printing I think.
>
> Bob Frost
>
>
> ----- Original Message -----
> From: "Martin Wesley" <mwesley250@...>
>
>
> >
> > Unfortunately, I feel I understand the matter reasonably well for my
> > purposes. I had hoped that providing other sources would help you
> understand
> > what I was saying. The dialogue has helped me to clarify and confirm my
> > thoughts.

On 3/27/02 darkroom@... wrote:

>> It is *absolutely not true* that an inky printmaker has an infinite
>> number of tones available. All bw output processes currently only output
>> 8 bits or 256 tones. A quadtone system is *theoretically* capable of
>> representing many more tones... 256 to the power 4 in fact, which is a
>> hell of a lot, but no drivers that I am aware of address this ability.
>
>Hi Johnny,
>
>The PiezoPro supposedly takes 16 bit data, as does Piezo V.6...and it also
>works with 6 tone printers...  People have reported on the Piezo list that
>they notice a difference in using 16 bit files.

I knew that they could print from 16-bit files but I don't know that they
use 16-bit data to calculate the dither. If they did so I would be very
impressed. But I am often impressed by the Coneheads so who knows?
Perhaps someone here.

-- 
John Brownlow

http://www.pinkheadedbug.com

While the heavy weights take a well deserved breather I thought I'd throw in
my two cents.

<Geek mode on>
Dynamic range has a very specific meaning to people involved in signal
processing and designers of optical, audio and many other types of
equipment.  That definition is just as Austin has stated.  Dynamic Range
(DR) = 10log10 (smallest signal/largest signal).  Some of the confusion
comes from the definition of the numerator.  The "smallest signal" is not
the least dense or the least grey, it is the difference in density (or
greyness or whatever you are measuring) between one sample and another which
is the smallest possible amount greyer (or more dense or whatever you are
measuring).

If you accept the definition above, then Dynamic Range does describe the
number of intermediate values.  The number of intermediate values is the
range of values divided by the "smallest signal".

<Geek mode off>

Having said that, it seems to me that a perfectly meaningful and unambiguous
(in any ordinary sense of those words) conversation can be had between two
printers using "dynamic range" and "density range" interchangeably to
describe the range of tones produced by an ink/paper/workflow combination.


Good Printing,



Kevin Gulstene

Hi Kevin,

> While the heavy weights take a well deserved breather I thought
> I'd throw in
> my two cents.

One minor correction, I promise ;-)

> <Geek mode on>
> Dynamic range has a very specific meaning to people involved in signal
> processing and designers of optical, audio and many other types of
> equipment.  That definition is just as Austin has stated.  Dynamic Range
> (DR) = 10log10 (smallest signal/largest signal).

Er, you mean largest over smallest...

> Having said that, it seems to me that a perfectly meaningful and
> unambiguous
> (in any ordinary sense of those words) conversation can be had between two
> printers using "dynamic range" and "density range" interchangeably to
> describe the range of tones produced by an ink/paper/workflow combination.

OK, I changed my mind, two...  density range does NOT describe any RANGE of
tones, it describes the two endpoints only...

Regards,

Austin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 1:39 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and >
> OK, I changed my mind, two...  density range does NOT describe any RANGE
of
> tones, it describes the two endpoints only...

Oh no it does not. What it indicates (as against describe) is the difference
between two end density points.

> Subject: RE: [Digital BW] Dynamic Range Definitions and >
> > OK, I changed my mind, two...  density range does NOT describe any RANGE
> of
> > tones, it describes the two endpoints only...
>
> Oh no it does not. What it indicates (as against describe) is the
> difference
> between two end density points.

Dickbo,

You missed the point.  I know what density range and dynamic range are
perfectly well, and how they are arrived at and what they "describe".  My
response was about number of data points.  Saying that density range
"describes" the two end points was merely to point out that only the two end
points are used to arrive at density range, and no intermediate data points
are at all relevant to density range.  As has been discussed, dynamic range
requires addition information beyond the two end points.

Austin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 4:00 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


> > Subject: RE: [Digital BW] Dynamic Range Definitions and >
> > > OK, I changed my mind, two...  density range does NOT describe any
RANGE
> > of
> > > tones, it describes the two endpoints only...
> >
> > Oh no it does not. What it indicates (as against describe) is the
> > difference
> > between two end density points.
>
> Dickbo,
>
> You missed the point.  I know what density range and dynamic range are
> perfectly well, and how they are arrived at and what they "describe".  My
> response was about number of data points.  Saying that density range
> "describes" the two end points was merely to point out that only the two
end
> points are used to arrive at density range, and no intermediate data
points
> are at all relevant to density range.

But that is not what you said

On 3/28/02 5:39 AM, "Austin Franklin" <darkroom@...> wrote:

> Hi Kevin,
> 
>> While the heavy weights take a well deserved breather I thought
>> I'd throw in
>> my two cents.
> 
> One minor correction, I promise ;-)
> 
>> <Geek mode on>
>> Dynamic range has a very specific meaning to people involved in signal
>> processing and designers of optical, audio and many other types of
>> equipment.  That definition is just as Austin has stated.  Dynamic Range
>> (DR) = 10log10 (smallest signal/largest signal).
> 
> Er, you mean largest over smallest...
> 
I intended to write (DR) = -10log10 (smallest signal/largest signal) but
omitted the leading negative sign.  I'm going to blame it on the time of
day.
>> Having said that, it seems to me that a perfectly meaningful and
>> unambiguous
>> (in any ordinary sense of those words) conversation can be had between two
>> printers using "dynamic range" and "density range" interchangeably to
>> describe the range of tones produced by an ink/paper/workflow combination.
> 
> OK, I changed my mind, two...  density range does NOT describe any RANGE of
> tones, it describes the two endpoints only...
> 
OK.  I had hoped to communicate that, in my experience, the increased
precision of the definition does not enhance the description of a print.
The increased precision, in my view, is most often lost in the "noise' of
our casual use of language.  Sort of like a 16bit ADC vs an 8bit ADC in a
'system' with a DR of 24.  Sorry, I couldn't resist the analogy ;).

> Regards,
> 
> Austin
> 
> 
> 
> 
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and other
> resources as they are often being updated. The page is at:
> 
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
> 
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
> them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or "flames."
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
> resources on the homepage.
> 
> 
> 
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> 
> 
>

----- Original Message -----

From: "John Brownlow" <lists@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, March 27, 2002 5:51 PM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


> On 3/27/02 mwesley250@... wrote:
>
> >Finally the "number of tones" present in a print does not depend upon the
> >dynamic range or the range or the min and max. A print is an analog image
> >and by definition has continuous tones that flow from one to another
> >without any step change or gaps. Just as the sun sets in a smooth motion
> >and not in increments, so the shades of gray flow in a continuous tone
print.
> >
> >A silver print or negative is continuous tone by the nature of the
> >chemical processes that produced it and an inkjet print is continuous
> >tone by virtue of the dither pattern, variable droplet size, etc. that
> >are used to produce it from a stepped digital file. By the definition of
> >continuous tone there are no steps and any tone value between the limits
> >of min and max can be created. Even though our eyes and instruments
> >cannot distinguish between two tones, if they are too close together in
> >value, does not change the fact that an infinite number of tones are
> >available to the print maker in either silver or ink to create their
image.
>
> Both of these statements are true and not true. In a sort of general
> hand-wavy practical way they are true, but they are inaccurate when you
> get close to them.
>
> 1. No analog systems are continuous at the quantum level.

Hmmm. That is the same as saying there are no analog systems since the very
concept of analog means continuous.

> Not even the
> sun crossing the sky.

The rotation of the earth occurs in small steps and not in a continuous
motion? An apple falls to the ground in distinct intervals? I know back to
my rubber example but reality moves form state to state hitting all the
positions in between without breaks or step changes.

> Quantum noise is a limiting factor in many micro-
> electronic systems.

This is true. Noise is always an issue in electronic systems.
>
> 2. A silver print or negative is not continuous at the granular level in
> terms of its representation of the scene.

Only with 400 ASA film in Xtol. (Just kidding your methods match you style
perfectly.)

Not as a representation of the scene perhaps but it is still a continuous
tone that imperfectly recorded the tones of the scene. Getting down to a
microscopic level and examining the crystal growth, bloom and infectious
development things look uneven but they vary continuously and somewhat
unpredictably. The important thing is that you can control the system to hit
any density not a finite number of densities. (Besides if you developed your
negs in Pyro this would not be an issue.<G> )
>
> 3. An inkjet print is certainly not continuous since each channel is
> (currently) only capable of representing 256 shades of gray.

Strictly speaking you have 6 channels so you should have 1536 combinations.
This is increased by the blending provided by dither patterns and bleeding.
>
> All of these things appear continuous if you squint hard enough but we
> are not squinting in this discussion.

I don't have to squint at all to look at the real world and see that is full
of continuous tones and motions. Remember the idea of usable or meaningful
values. Taking things down to the level of quantum mechanics may not be
helpful in establishing values to assess the relative merits of one printing
medium over another.
>
> It is *absolutely not true* that an inky printmaker has an infinite
> number of tones available. All bw output processes currently only output
> 8 bits or 256 tones. A quadtone system is *theoretically* capable of
> representing many more tones... 256 to the power 4 in fact, which is a
> hell of a lot, but no drivers that I am aware of address this ability.

Well you can control where those 256 bits fall within a tonal range and if
you can do that, any tone is available to you even if you can only get a
finite number of them into a particular print. Think of it this way. Within
the range, what density can't I reproduce in an inkjet print?

Actually I think the Epson or Piezo drivers are getting you more than 256
tones out of an ink set. Check it out by doing a black ink only print and a
full quad print. It is interesting that the single ink gets you about 90%
there (100% in some cases if the grain size is a good match to the dither
pattern) and the rest of the quad/hex inks fill in the remaining. For my
work I couldn't live with just 256 shades of gray and would still be
sloshing around in the dark.

Martin Wesley

----- Original Message -----

From: "John Brownlow" <lists@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, March 27, 2002 5:43 PM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


> On 3/27/02 mwesley250@... wrote:
>
> >For all practical purposes a print existing as an object separate from
> >the process that produced it seems to be noiseless and the dynamic range
> >is the ratio of the min and max reflectance or the difference of min and
> >max density.
>
> no, by the definitions you cite, the variation in the reflectance of the
> substrate (paper) is the noise floor. Since the paper is not perfectly
> white, there is a limit to how delicate the highlights can be before they
> are more delicate than the variations of the paper itself. Therefore the
> choice of paper may have a large effect on the dynamic range of the
> printing system.
>
John,

I like that! Finally a concept of noise for a paper print. There is probably
a similar variation or mottle in the deepest black so you have a noise floor
and a noise ceiling.

If you pick up the newspaper you can easily see the base brightness
variation. There would seem to be an obvious difference between an art paper
like H. German Etching or William Turner and Epson Photo Paper just because
of the texture.

There would be two factors to this noise the amplitude and the frequency as
you moved across the paper surface. I know that the spot size measured by
the densitometer I have used would not be small enough to pick this up in
quality papers and I doubt even the really good ones would be able to spot
the variations in brightness. I suspect that the surface variations you are
talking about would be less than the uncertainly of standard instruments.

I have no idea what the magnitude of such noise might be. From the papers
you have worked with do feel that you can see this as limiting property on
the quality of your prints?

Martin

On 3/28/02 1:22 PM, "Martin Wesley" <mwesley250@...> wrote:

>> Both of these statements are true and not true. In a sort of general
>> hand-wavy practical way they are true, but they are inaccurate when you
>> get close to them.
>> 
>> 1. No analog systems are continuous at the quantum level.
> 
> Hmmm. That is the same as saying there are no analog systems since the very
> concept of analog means continuous.

Not at all. An analog system or device is one that represents data variation
by a measurable physical quality. You are arguing that that is continuous,
not me.
> 
>> Not even the
>> sun crossing the sky.
> 
> The rotation of the earth occurs in small steps and not in a continuous
> motion? An apple falls to the ground in distinct intervals? I know back to
> my rubber example but reality moves form state to state hitting all the
> positions in between without breaks or step changes.

I am just being pernickety. At the sub-atomic level it all falls apart. An
electron does not trundle from A to B like a billiard ball. For this reason,
neither does the sun. It is almost infinitely unlikely, but perfectly
possible, that one day the sun will hop from one point of the sky to the
other. That's the nature of quantum unpredictablity.

>> 2. A silver print or negative is not continuous at the granular level in
>> terms of its representation of the scene.

> Not as a representation of the scene perhaps but... [SNIP]

We are talking about it as a representation. That's the whole point of the
discussion.

>> All of these things appear continuous if you squint hard enough but we
>> are not squinting in this discussion.
> 
> I don't have to squint at all to look at the real world and see that is full
> of continuous tones and motions. Remember the idea of usable or meaningful
> values. Taking things down to the level of quantum mechanics may not be
> helpful in establishing values to assess the relative merits of one printing
> medium over another.

Certainly agreed.
>> 
>> It is *absolutely not true* that an inky printmaker has an infinite
>> number of tones available. All bw output processes currently only output
>> 8 bits or 256 tones. A quadtone system is *theoretically* capable of
>> representing many more tones... 256 to the power 4 in fact, which is a
>> hell of a lot, but no drivers that I am aware of address this ability.
> 
> Well you can control where those 256 bits fall within a tonal range and if
> you can do that, any tone is available to you even if you can only get a
> finite number of them into a particular print. Think of it this way. Within
> the range, what density can't I reproduce in an inkjet print?

You can map each of those 256 brightness values to any output value you like
but that doesn't make it a continuous tone image. There are still 256
brightness levels in the output image with discrete steps in reflectance
between them.

-- 
John Brownlow

http://www.pinkheadedbug.com

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, March 27, 2002 5:47 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


> > Austin,
> >
> > They are not my explanations they are the commonly accepted
> > definitions by a
> > wide range of people in many fields.
>
> Martin,
>
> Commonly accepted by whom?  Merriam Webster, magazines, etc. are hardly
good
> sources for technical definitions.  As I've said, the definitions you
cited,
> with the exception of a couple of them, require interpretation of
ambiguous
> terms.  They are NOT wrong, just open to (mis)interpretation.
Specifically
> two terms used, smallest and largest (or what ever words were used).  I
have
> shown VERY CLEARLY what is meant by smallest and largest when used with
> respect to dynamic range, as defined by a very definitive source.

Austin,

I guess I am a generalist and I like to start with a broad definition,
simple if you will, and then work my way to something more specific to the
case in front of us. If we start from the dictionary definition and try to
move to paper prints selecting the meaningful or usable min and max as seems
appropriate to that system.
>
> > I do not see the need to
> > reinvent these
> > meanings.
>
> Me either, and clarifying ambiguous terminology is hardly reinventing
> anything.  I have merely cited the clarifications from reliable sources.
> You have not attempted at all to cite any clarification to the terms that
> are in question.

For the most part I feel you are simply substituting synonyms. The point
where we collide is over "discernable." I agree with this as part of the
process of determining the min and max. You move on into the idea of
"discernable difference" over the entire tonal range and I think that this
is not covered by the general concept of dynamic. "Discernable difference"
is very important when you are talking about systems where the input and
output are changing with time and you want to discuss the ability of the
output to respond to the input. I feel that the concept of "discernable
difference" is not applicable to a paper print since it is not changing with
time but only with linear movement across the surface. With the linear
movement you can see multiple pairs simultaneously and "discernable
difference" does not seem significant.
>
> > Dynamic range is a ratio but subtracting log values is
> > mathematically the same as taking the ratio. I don't see the need to
> > reinvent the math either.
>
> Me either, and I certainly am not "reinventing" math.  Just because two
> "measurements" are expressed in log form does NOT mean they are the same.
> Dynamic range stands in and of it self as ONE log10 number.

The requirement that dynamic range be a log is not part of the general
definition or concept if you will. It seems very cleat that the dynamic
range is an expression of a ratio relationship. You need some values to
calculate that ratio. I do not see how you can directly "measure" the
dynamic range of a print. Perhaps in electronics you can.

> It is expressed
> in dB.  Density range requires TWO log10 values, dMin and dMax.  Density
> values are NOT expressed in any "scale" simply because they are relative
> unto themselves.

Well any range needs endpoints of course. Density is a log of the inverse of
a percentage, the reflectance, so it is all unitless. It would be
interesting to have the reflectance as a direct flux reading to work with.
>
> > > As has been shown, you can have a very high density range (a high
black
> > > value on a very white paper, with no intermediate tones) and that has
a
> > very
> > > LOW dynamic range, because there are no tones in-between.
> > Don't let this
> > > simple concept slip by...it's important.
> >
> > Where in the definitions was this stated?
>
> What's "this" that you are questioning?

Back to the general definition which says the dynamic range is a ratio of
the min and max. This does not seem to account for the values in between or
they assumed a continuously variable system.

> The measurement for density range
> is clear, dMax - dMin.  The measurement for dynamic range is clearly
defined
> by the dynamic range equation.  Here, I'll apply some simple numbers to
> this:
>
> dMax = 1.8
> dMin = .2
>
> Density Range (dRange) = 1.6
>
> Easy, right?
>
> What's the dynamic range?  Well, we don't know what the smallest
discernable
> signal is, now do we?  Nor do we really know the largest that the
> paper/system can attain either.  This makes for a quandary.

Not really because you are interested in the meaningful or useful values
which may simply be the only the values you do know.
>
> The dynamic range is based on the largest ATTAINABLE signal for the
medium,
> not the largest for a single print.  You certainly could just print the
> blackest black you could, and get it from that, not too tough.  Let's say
it
> was 1.8...for sake of argument.

Okay. A little high for inkjet and rather low for silver.
>
> The smallest discernable signal isn't dMin, since we could possibly be
able
> to "discern" in .01 density value steps...  and without knowing what the
> minimum discernable signal is, we don't know what the dynamic range is, by
> definition.

I agree that you may be able to discern a change of 0.01 density near Dmin.
The problem is that as you approach Dmax the density change you can discern
decreases. The discernable difference is not a constant in this system which
is why I object to using this in a dynamic range calculation for this
system.
>
> Let's say our smallest discernable signal was 0.01 density value (which is
> hardly far fetched, and actually a reasonable number).  The dynamic range
> would be 10log10((1.8-0.2)/.01) or 22dB.  Note that even in Bells, 2.2 is
> not the same as 1.6.  Of course there happens to be a value for the
smallest
> discernable signal that would make dynamic range and density range the
> same...but that's the same as saying a clock is right twice a day...it's
> purely by happenstance.

Here is the reason you cannot divide by the density value of 0.01. Assuming
as you say we can discern a density of 0.01, this means that we could
discern the difference between 0.03 and 0.04 densities. If you look at that
in terms of the reflectances that were used to calculate the density they
would be 93.3% and 91.2% so the discernable difference would be a 2.1
percentile difference. If we go to the other end of the scale and say I can
discern between tones 0.01 density values apart then I would say that I can
distinguish between 1.78 and 1.79. In terms of reflectance this would be the
difference between 1.62% and 1.66% so the discernable difference would be
0.4 percentile.

Do you see what is happening here? From your earlier diagram you indicated
that the discernable difference was constant across the entire range. That
is not the case here. The discernable difference gets larger as the tones
get darker until finally you reach a point where you eye cannot tell the
difference between a density of 2.4 and 2.8.

I think all of this would be much more easily understood if it was viewed in
terms of reflectance rather than the density log values.
>
> > Dynamic range tells you nothing
> > about the number of intermediate values.
>
> That's what you claim, and I keep telling you, you are mistaken.
>
> It really would help if YOU provide an EQUATION, with appropriate
> definitions of terms used in the equation (as I have done), NOT some
> ambiguous verbiage that is open to interpretation.

I believe I did. From the definition the dynamic range is the ratio of the
maximum divided by the minimum. Lets look at it this way, for a print:

Dynamic Range of the Reflectance = (maximum reflectance of the brightest
visually discernable white)/(minimum reflectance of the darkest visually
discernable black)

This ratio can also be expressed by raising 10 to the power of the density
range (maximum density minus minimum density).

My thought is that the dynamic range of a print may not be a particularly
useful expression for us. The simple range or knowing the lightest and
darkest shades of gray of a medium are what is of importance to us. We also
tend to overly focus on getting the deepest black where in reality you gain
more by increasing the brightness white.

Artistically range may not matter to everyone and some people will make
great prints using only a portion of the available range.

I understand that you are trying to give us an expression that says
something about the number of discernable tones that a media or system can
create but I don't think that this can be done here because I do not see
that there are a finite number of tones.

Best,
Martin

----- Original Message -----

From: "Kevin Gulstene" <kevin@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Wednesday, March 27, 2002 11:27 PM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


> While the heavy weights take a well deserved breather I thought I'd throw
in
> my two cents.
>
> <Geek mode on>
> Dynamic range has a very specific meaning to people involved in signal
> processing and designers of optical, audio and many other types of
> equipment.  That definition is just as Austin has stated.  Dynamic Range
> (DR) = 10log10 (smallest signal/largest signal).  Some of the confusion
> comes from the definition of the numerator.  The "smallest signal" is not
> the least dense or the least grey, it is the difference in density (or
> greyness or whatever you are measuring) between one sample and another
which
> is the smallest possible amount greyer (or more dense or whatever you are
> measuring).
>
> If you accept the definition above, then Dynamic Range does describe the
> number of intermediate values.  The number of intermediate values is the
> range of values divided by the "smallest signal".

Kevin,

Thank you for the nice concise explaination. My concern with a print is that
the "smallest possible amount grayer" will go to zero if the print or medium
is continuous in tone which renders the expression meaningless for this
application.
>
> <Geek mode off>
>
> Having said that, it seems to me that a perfectly meaningful and
unambiguous
> (in any ordinary sense of those words) conversation can be had between two
> printers using "dynamic range" and "density range" interchangeably to
> describe the range of tones produced by an ink/paper/workflow combination.

Absolutely and this has been a common practice for quite some time.

Martin

On 3/28/02 2:57 PM, "Martin Wesley" <mwesley250@...> wrote:

> I do not see how you can directly "measure" the
> dynamic range of a print. Perhaps in electronics you can.

It seems simple to me.

You get the print driver to output the lightest continuous (yeah I know)
tone it is capable of short of not printing at all. You measure the
reflectance of that patch.

You measure the reflectance of the darkest tone it is capable of laying
down.

The dynamic range is the ratio between them, exactly as Kevin said.

I don't know why you're all making it so damn complicated.

Martin, if you think an inkjet print is a continuous tone, then for any
shade of gray which Kevin prints using his printer and claims to be the
lightest gray possible, you with your magic continuous printer, will be able
to print a lighter tone.

I shall be interested to hear how you intend to do this.

-- 
John Brownlow

http://www.pinkheadedbug.com

on 3/28/02 1:22 PM, Martin Wesley wrote:

> For my
> work I couldn't live with just 256 shades of gray and would still be
> sloshing around in the dark.

Hmm. Next time you're working a grayscale image on your computer change your
monitor from millions of colors to 256 shades of gray and tell me if your
image looks any worse for wear. And unless you are printing from images will
full histograms, end to end, I'd presume your images to have less than 256
shades of gray. 

I suppose one could create a white to black gradient in Photoshop, and
posterize it into as many steps as they want, print it, and see how many
steps they can distinguish. I'd bet somewhere beyond 100 or so either the
printing system or our perception would begin to break down. However, any
system that can reproduce Tyler's Zees successfully can handle the 100 tones
that comprise it, so we know at least that much IS possible today.

Todd

On 3/28/02 12:04 PM, "Martin Wesley" <mwesley250@...> wrote:

> 
> ----- Original Message -----
> From: "Kevin Gulstene" <kevin@...>
> To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> Sent: Wednesday, March 27, 2002 11:27 PM
> Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones
> 
> 
>> While the heavy weights take a well deserved breather I thought I'd throw
> in
>> my two cents.
>> 
>> <Geek mode on>
>> Dynamic range has a very specific meaning to people involved in signal
>> processing and designers of optical, audio and many other types of
>> equipment.  That definition is just as Austin has stated.  Dynamic Range
>> (DR) = 10log10 (smallest signal/largest signal).  Some of the confusion
>> comes from the definition of the numerator.  The "smallest signal" is not
>> the least dense or the least grey, it is the difference in density (or
>> greyness or whatever you are measuring) between one sample and another
> which
>> is the smallest possible amount greyer (or more dense or whatever you are
>> measuring).
>> 
>> If you accept the definition above, then Dynamic Range does describe the
>> number of intermediate values.  The number of intermediate values is the
>> range of values divided by the "smallest signal".
> 
> Kevin,
> 
> Thank you for the nice concise explaination. My concern with a print is that
> the "smallest possible amount grayer" will go to zero if the print or medium
> is continuous in tone which renders the expression meaningless for this
> application.

I think I am with you on this.  While I agree the "smallest possible amount
greyer" will get much smaller as the sensing technology improves, there will
always be some level at which there is a smallest possible amount (ranging
from the practical limits of our eyes to the other extreme of quantum
whatevers). More importantly we should keep in mind the purpose of the
textbook definition which is to characterise a system and not an attribute.

The text book definition of Dynamic Range is appropriate for measuring a
system.  We would have a similar problem discussing the dynamic range of a
negative.  On the other hand it is quite easy to see how the term applies to
a scanner.  In the context of a scanner (a "system" for measuring the
density of a negative) Austin's Diagram makes perfect sense.  There is a
minimum density of the negative (film base + fog) and a maximum density that
can be read by the sensor, and there is a "smallest increment of grey" that
can be measured by the "system".  This smallest increment is determined by
the system (# of bits in the ADC, the 'noise' introduced when amplifying the
signal from the sensors, etc).

In my mind, it is much easier to apply the text book definition (which I
think we agree can differ from the common use definition) to a system for
measuring an attribute of a print, than to the attribute of the print.

Ouch, I have to go find the Tylenol now.

>> 
>> <Geek mode off>
>> 
>> Having said that, it seems to me that a perfectly meaningful and
> unambiguous
>> (in any ordinary sense of those words) conversation can be had between two
>> printers using "dynamic range" and "density range" interchangeably to
>> describe the range of tones produced by an ink/paper/workflow combination.
> 
> Absolutely and this has been a common practice for quite some time.
> 
> Martin
> 
> 
> 
> 
> 
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and other
> resources as they are often being updated. The page is at:
> 
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
> 
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
> them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or "flames."
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
> resources on the homepage.
> 
> 
> 
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> 
> 
>

Austin, et. al.

There are three important scales within the totaal range of exposures that
can be printed. The full range from black to white is represented by Zones 0
to X. Within this range lies the dynanic range, representing the first
useful values above Zone 0 and below Zone X, or Zones I to IX. The range of
zones which convey definite qualities of texture and the recognition of
substance is the textural range from Zones II to VII. (1)

(1) Ansel Adams, The Negative, Little Brown, Nineteenth Printing, 1998.

-Mike


----- Original Message -----
From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 8:39 AM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


> Hi Kevin,
>
> > While the heavy weights take a well deserved breather I thought
> > I'd throw in
> > my two cents.
>
> One minor correction, I promise ;-)
>
> > <Geek mode on>
> > Dynamic range has a very specific meaning to people involved in signal
> > processing and designers of optical, audio and many other types of
> > equipment.  That definition is just as Austin has stated.  Dynamic Range
> > (DR) = 10log10 (smallest signal/largest signal).
>
> Er, you mean largest over smallest...
>
> > Having said that, it seems to me that a perfectly meaningful and
> > unambiguous
> > (in any ordinary sense of those words) conversation can be had between
two
> > printers using "dynamic range" and "density range" interchangeably to
> > describe the range of tones produced by an ink/paper/workflow
combination.
>
> OK, I changed my mind, two...  density range does NOT describe any RANGE
of
> tones, it describes the two endpoints only...
>
> Regards,
>
> Austin
>
>
>
>
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and
other resources as they are often being updated. The page is at:
>
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
>
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or "flames."
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
resources on the homepage.

>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>

> > Having said that, it seems to me that a perfectly meaningful and
> unambiguous
> > (in any ordinary sense of those words) conversation can be had
> between two
> > printers using "dynamic range" and "density range" interchangeably to
> > describe the range of tones produced by an ink/paper/workflow
> combination.
>
> Absolutely and this has been a common practice for quite some time.

Martin,

"common practice" and "quite some time" for whom?  How long have you had
anything to do with digital imaging, seriously?  I've been at this 25 years,
and only in the past few years, because of all the
misunderstanding/misinformation on the web, combined with non-technical (as
in home owners etc.) people getting into digital imaging, and the
proliferation of cheap scanners, have I seen this density range/dynamic
range misinterpretation "crop up" ;-)

Austin

----- Original Message -----

From: "John Brownlow" <lists@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 11:28 AM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


> On 3/28/02 1:22 PM, "Martin Wesley" <mwesley250@...> wrote:
>
(snip)

> >> 1. No analog systems are continuous at the quantum level.
> >
> > Hmmm. That is the same as saying there are no analog systems since the
very
> > concept of analog means continuous.
>
> Not at all. An analog system or device is one that represents data
variation
> by a measurable physical quality. You are arguing that that is continuous,
> not me.

John,

The definition I was working from includes continuous:

analog: 2 a : of, relating to, or being a mechanism in which data is
represented by continuously variable physical quantities
> >
> >> Not even the
> >> sun crossing the sky.
> >
> > The rotation of the earth occurs in small steps and not in a continuous
> > motion? An apple falls to the ground in distinct intervals? I know back
to
> > my rubber example but reality moves form state to state hitting all the
> > positions in between without breaks or step changes.
>
> I am just being pernickety. At the sub-atomic level it all falls apart. An
> electron does not trundle from A to B like a billiard ball. For this
reason,
> neither does the sun.

I don't know if that is a reasonable assumption but we are really getting
far a field.

I would need to call in Dad on this one. He is the theoretical physicist in
the family but he insists I would need to be 2 to 4 years into my physics
doctorate to be able to understand what he is talking about. I definately
agree with him.

> It is almost infinitely unlikely, but perfectly
> possible, that one day the sun will hop from one point of the sky to the
> other. That's the nature of quantum unpredictablity.

I can give you a good rate to insure you against lost you suffer in the
event that should occur. Say for $100 per month? <G>
>
> >> 2. A silver print or negative is not continuous at the granular level
in
> >> terms of its representation of the scene.
>
> > Not as a representation of the scene perhaps but... [SNIP]
>
> We are talking about it as a representation. That's the whole point of the
> discussion.

Well actually no. We got started over the claim that a Piezo print contains
more tones than a silver print without any reference to how the print was
made. I admit that this may be a dubious discussion. From a representational
point of view Photoshop and a good scanner or digital camera may allow you
to achieve a better tonal capture/compression than traditional photography
so that you can represent more real life tones in a print. However, you
could make that print from Photoshop to inkjet or onto silver paper in a
LightJet or by producing an interneg for silver printing. So the claim may
have some validity for digital B&W but I don't think it is true for a silver
vs. inkjet comparison.
>
(snip)
> >
> > Well you can control where those 256 bits fall within a tonal range and
if
> > you can do that, any tone is available to you even if you can only get a
> > finite number of them into a particular print. Think of it this way.
Within
> > the range, what density can't I reproduce in an inkjet print?
>
> You can map each of those 256 brightness values to any output value you
like
> but that doesn't make it a continuous tone image. There are still 256
> brightness levels in the output image with discrete steps in reflectance
> between them.
>
Agreed. But I think that by adding dither and varying droplet size you can
produce a continuous tone, perhaps not with 8 bit but with multiples of
8-bit. Can't back this up with hard fact though. Practically, a well made
inkjet print seems to convince my eye that it is continuous tone although
ones that are not well done have a very non-continuous look to them.

Martin

----- Original Message -----

From: "Todd Flashner" <tflash@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 12:35 PM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


> on 3/28/02 1:22 PM, Martin Wesley wrote:
>
> > For my
> > work I couldn't live with just 256 shades of gray and would still be
> > sloshing around in the dark.
>
> Hmm. Next time you're working a grayscale image on your computer change
your
> monitor from millions of colors to 256 shades of gray and tell me if your
> image looks any worse for wear.

Todd,

I tried it and the on screen image looks terribly degraded and posterized in
comparison to 32-bit color monitor setting.

> And unless you are printing from images will
> full histograms, end to end, I'd presume your images to have less than 256
> shades of gray.

I have done some mono-ink prints and I am guessing that with dither they
probably give you something beyond 256 shades. They are very good for my
TMax 400 35mm negs but I am not happy with this for my 6x7 or 4x5.
>
> I suppose one could create a white to black gradient in Photoshop, and
> posterize it into as many steps as they want, print it, and see how many
> steps they can distinguish. I'd bet somewhere beyond 100 or so either the
> printing system or our perception would begin to break down. However, any
> system that can reproduce Tyler's Zees successfully can handle the 100
tones
> that comprise it, so we know at least that much IS possible today.

I really wish I had some comfort level as to how many shades of gray can be
perceived by the human eye or at least a range since there is bound to be
person to person variation. I have heard 100, 256 up to 1030. The most
thorough study which I neglected to save was done in Sweden and they
reported that most people could differentiate between 800 and 900 shades of
gray but that artists and people in the graphics industry could see 900 to
the 1030 number. Unfortunately they did give any details of how the testing
was done. I hate newspaper articles about scientific tests.

I think that if it is some fixed number of say 256 you can have multiple
sets of 256 tones in an image that contribute to its impact. Say you could
only differentiate 256 but you had 1024 tones. Your eye could see the
difference between 1 and 4, 4 and 8, 8 and 12, etc. You could also
distinguish between 2 and 5, 5 and 9, 9 and 13, etc.

Martin

> If we start from the dictionary definition and try to
> move to paper prints selecting the meaningful or usable min and
> max as seems
> appropriate to that system.

But the equation for dynamic range does not allow YOU to select the
"meaningful" definition of smallest and largest as YOU understand them.
They are defined by the equation.

> > > I do not see the need to
> > > reinvent these
> > > meanings.
> >
> > Me either, and clarifying ambiguous terminology is hardly reinventing
> > anything.  I have merely cited the clarifications from reliable sources.
> > You have not attempted at all to cite any clarification to the
> terms that
> > are in question.
>
> For the most part I feel you are simply substituting synonyms.

They you are missing the entire concept.  What I have shown, quite clearly,
is that there are AMBIGUOUS definitions of both largest and smallest.  I
have also shown what the two ambiguous definitions are.

> The point
> where we collide is over "discernable."

But it's clearly defined by the equation and what dynamic range is there to
measure.  As I've said time and time again, you are arguing over a very
clear concept/definition/equation that I did not come up with.  I am merely
trying to get you to understand what it really is, and how your
understanding of it is flawed.  Unfortunately, for what ever reason I do not
know, it isn't working.

> I agree with this as part of the
> process of determining the min and max.

I don't follow you there.

> You move on into the idea of
> "discernable difference" over the entire tonal range and I think that this
> is not covered by the general concept of dynamic.

Well, then what do YOU believe is "dynamic" in dynamic range?  That is a
very important question.

> "Discernable difference"
> is very important when you are talking about systems where the input and
> output are changing with time

Time has nothing to do with it.  You do not need time to discern change, but
you do need change to discern time (off topic, but it's true anyway).  Two
things can have a discernable difference at the exact same time!

> I do not see how you can directly "measure" the
> dynamic range of a print. Perhaps in electronics you can.

Very simply.  Print a completely solid tone, with as minimum detectable
difference in density in the original as possible.  Print the darkest tone
you can.  Measure the darkest tone, measure the paper, then take multiple
measurements of the "patch" of your "test" tone.  The variance across the
test tone is your noise, and yes, I know, you have to account for the
original variances, as well as possible variances caused by the lense...but
none the less, it does give you a basis for noise.  You get the "largest
signal" in the dynamic range equation by your dMax - dMin, and you get your
"smallest discernable signal" from the variance in your "patch".  There you
have it.

> Density is a log of the
> inverse of
> a percentage, the reflectance, so it is all unitless.

He he, I'll remind you at a later date that you said that ;-)  It's also
unitless since it's only a relative number...and units don't matter.

> Back to the general definition which says the dynamic range is a ratio of
> the min and max.

As I've said countless times, min and max WHAT?  What does min and max mean?
They are incomplete terms.

 This does not seem to account for the values in
> between or
> they assumed a continuously variable system.

Min and Max does if you take max to mean the maximum signal and min to mean
the minimum discernable signal ;-)

> > What's the dynamic range?  Well, we don't know what the smallest
> discernable
> > signal is, now do we?  Nor do we really know the largest that the
> > paper/system can attain either.  This makes for a quandary.
>
> Not really because you are interested in the meaningful or useful values
> which may simply be the only the values you do know.

Sorry, I have no idea what you mean by that...

> I agree that you may be able to discern a change of 0.01 density
> near Dmin.
> The problem is that as you approach Dmax the density change you
> can discern
> decreases.

You're right, but only if you're above 3!!!  It's also not relevant to the
discussion, but is a noted point.

> The discernable difference is not a constant in this
> system which
> is why I object to using this in a dynamic range calculation for this
> system.

That's the first you've said that!  But...as I said, your claim is only good
for density values above 3, below 3, they are quite constant...and since
this IS a B&W list...and as we all know, B&W doesn't go above 2...I think
we're safe ;-)

> Do you see what is happening here?

Yes, and it doesn't matter.

> From your earlier diagram you indicated
> that the discernable difference was constant across the entire range. That
> is not the case here.

Well, it is the case, but even if it weren't, it's not relevant.  Did you
ever take Differential Equations?

> > > Dynamic range tells you nothing
> > > about the number of intermediate values.
> >
> > That's what you claim, and I keep telling you, you are mistaken.
> >
> > It really would help if YOU provide an EQUATION, with appropriate
> > definitions of terms used in the equation (as I have done), NOT some
> > ambiguous verbiage that is open to interpretation.
>
> I believe I did. From the definition the dynamic range is the ratio of the
> maximum divided by the minimum.

That's NOT an equation.  AND maximum and minimum are left undefined.

> Dynamic Range of the Reflectance = (maximum reflectance of the brightest
> visually discernable white)/(minimum reflectance of the darkest visually
> discernable black)

Yeah, but that's wrong.  You have just given an equation for DENSITY range,
not dynamic range.

> My thought is that the dynamic range of a print may not be a particularly
> useful expression for us. The simple range or knowing the lightest and
> darkest shades of gray of a medium are what is of importance to
> us.

But it is so obvious that some images may have NO tones between black and
white, and some may have a LOT of tones between black and white.  That is
why dynamic range IS important.

> We also
> tend to overly focus on getting the deepest black where in
> reality you gain
> more by increasing the brightness white.

I very much so agree with that!

> I understand that you are trying to give us an expression that says
> something about the number of discernable tones that a media or system can
> create but I don't think that this can be done here because I do not see
> that there are a finite number of tones.

And I do see a finite number of tones, or at least a "range" of the number
of tones.  I can certainly say one has more or less than another!  I have
done the same print that has very bad tonality, there aren't many
discernable tones, and I have re-printed the exact same image and managed to
get a heck of a lot more tones out of that same image...via development
time/exposure.  Similarly with film, I have film that has the blacks so
blocked up, and some that has them nice and open...and that was based on
development/exposure too ;-)  Dynamic range is VERY important to me, as I
want full tonality in my specific prints.  Density range isn't as important
to me, for reasons that have been mentioned by you and others.

You seem to be shifting a bit here...more to an argument that dynamic range,
what ever we believe it is, isn't really that important...  Just a note.

Regards,

Austin

> The definition I was working from includes continuous:
> 
> analog: 2 a : of, relating to, or being a mechanism in which data is
> represented by continuously variable physical quantities

CONTINUOUSLY VARIABLE...that doesn't say how much the variation is ;-)

> > Hmm. Next time you're working a grayscale image on your computer change
> your
> > monitor from millions of colors to 256 shades of gray and tell
> me if your
> > image looks any worse for wear.
>
> Todd,
>
> I tried it and the on screen image looks terribly degraded and
> posterized in
> comparison to 32-bit color monitor setting.

Did you really set your monitor to 256 shades of gray?  I do not have that
ability, I only have the ability to set it to 256 COLORS, not to 256 shades
of gray...that is very different...256 colors will give you virtually no
shades of gray.

Austin

----- Original Message -----

From: "Kevin Gulstene" <kevin@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 1:05 PM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


> On 3/28/02 12:04 PM, "Martin Wesley" <mwesley250@...> wrote:
>
> >
> > ----- Original Message -----
> > From: "Kevin Gulstene" <kevin@...>
> > To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
> > Sent: Wednesday, March 27, 2002 11:27 PM
> > Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones
> >
> >
(snip)

> > Thank you for the nice concise explaination. My concern with a print is
that
> > the "smallest possible amount grayer" will go to zero if the print or
medium
> > is continuous in tone which renders the expression meaningless for this
> > application.
>
> I think I am with you on this.  While I agree the "smallest possible
amount
> greyer" will get much smaller as the sensing technology improves, there
will
> always be some level at which there is a smallest possible amount (ranging
> from the practical limits of our eyes to the other extreme of quantum
> whatevers). More importantly we should keep in mind the purpose of the
> textbook definition which is to characterise a system and not an
attribute.

As we get deeper into things, there is the risk of neglecting the actual
reality in favor of our measurements and equations. You always need to step
back and do a general sanity check on your numbers and what they tell you.
>
> The text book definition of Dynamic Range is appropriate for measuring a
> system.  We would have a similar problem discussing the dynamic range of a
> negative.  On the other hand it is quite easy to see how the term applies
to
> a scanner.  In the context of a scanner (a "system" for measuring the
> density of a negative) Austin's Diagram makes perfect sense.  There is a
> minimum density of the negative (film base + fog) and a maximum density
that
> can be read by the sensor, and there is a "smallest increment of grey"
that
> can be measured by the "system".  This smallest increment is determined by
> the system (# of bits in the ADC, the 'noise' introduced when amplifying
the
> signal from the sensors, etc).

Exactly. I think one of the confusions is that you can talk about the
potential dynamic range of a type of film in conjunction with a type of
developement as a system in these terms but not the individual negative once
it is exposed and developed unless you still have the original scene to
compare it to, which you normal don't.  I keep fair records of my view
camera work and will record light meter readings in the form of zone
placements for at least two areas in the scene so I could relate a given
negative back to these values and talk about how well they were or weren't
captured in numerical terms but it is still very crude. But it is enough to
help me adjust my developing conditions in the future.
>
> In my mind, it is much easier to apply the text book definition (which I
> think we agree can differ from the common use definition) to a system for
> measuring an attribute of a print, than to the attribute of the print.

This is one of the points I have been trying to make. The measurements are
not the print and the noise of the measurements is not the noise of the
print.
>
> Ouch, I have to go find the Tylenol now.

Try 4 Advil.

Martin

----- Original Message -----

From: "John Brownlow" <lists@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 12:12 PM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


> On 3/28/02 2:57 PM, "Martin Wesley" <mwesley250@...> wrote:
>
> > I do not see how you can directly "measure" the
> > dynamic range of a print. Perhaps in electronics you can.
>
> It seems simple to me.
>
> You get the print driver to output the lightest continuous (yeah I know)
> tone it is capable of short of not printing at all. You measure the
> reflectance of that patch.
>
> You measure the reflectance of the darkest tone it is capable of laying
> down.
>
> The dynamic range is the ratio between them, exactly as Kevin said.

John,

I completely agree with you and that is what I started out saying. My
response above was to Austin who said that you could measure it directly
without measuring the Dmin and Dmax individually.

>
> I don't know why you're all making it so damn complicated.

Sorry, I really don't mean too.
>
> Martin, if you think an inkjet print is a continuous tone, then for any
> shade of gray which Kevin prints using his printer and claims to be the
> lightest gray possible, you with your magic continuous printer, will be
able
> to print a lighter tone.
>
> I shall be interested to hear how you intend to do this.

Obviously you are limited by the range of Dmin and Dmax. Only within the
limits of Dmin and Dmax could you do that. Practically, don't you think that
if you wanted a certain tone in your print to be lighter by an amount equal
to Kevin's "lightness gray possible", you would have a good chance at doing
it in Photoshop? Back to my guitar string example. With a slide you can hit
any frequency within the range of any two points on the string. You can
slide from C to C# or hit anything in between. (Of course I would be happy
if I could hit one or the other but that's another story.)

Martin

> My
> response above was to Austin who said that you could measure it directly
> without measuring the Dmin and Dmax individually.

Er, Martin...I never said that.  EVER.  I have ALWAYS said you need to
measure the dMin and dMax as (dMax - dMin) IS the "largest signal" part of
the dynamic range equation.

Austin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 6:49 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


> > > Having said that, it seems to me that a perfectly meaningful and
> > unambiguous
> > > (in any ordinary sense of those words) conversation can be had
> > between two
> > > printers using "dynamic range" and "density range" interchangeably to
> > > describe the range of tones produced by an ink/paper/workflow
> > combination.
> >
> > Absolutely and this has been a common practice for quite some time.
>
> Martin,
>
> "common practice" and "quite some time" for whom?  How long have you had
> anything to do with digital imaging, seriously?

Austin,

Perhaps this is the root of our disagreement. I am not making any distiction
between digial photography/imaging and traditional photography. My remark
above would be that this has been common among photographers for quite some
time.

> I've been at this 25 years,
> and only in the past few years, because of all the
> misunderstanding/misinformation on the web, combined with non-technical
(as
> in home owners etc.) people getting into digital imaging, and the
> proliferation of cheap scanners, have I seen this density range/dynamic
> range misinterpretation "crop up" ;-)

The issue of dynamic range is very important in digital cameras and
scanners. You are correct that in these areas the term is not well used,
especially by manufacturers who want to claim bit depth is equal to the
dynamic range of their latest gadget. I do recognize that and your expertise
in this area. I just don't think that you can take those relationships that
apply to digital imaging and apply them to a finished print silver or inkjet
or whatever without stepping back and seeing if this is appropriate or if it
is telling you something meaningful about the print.

Martin

> > I do not see how you can directly "measure" the
> > dynamic range of a print. Perhaps in electronics you can.
>
> It seems simple to me.
>
> You get the print driver to output the lightest continuous (yeah I know)
> tone it is capable of short of not printing at all. You measure the
> reflectance of that patch.
>
> You measure the reflectance of the darkest tone it is capable of laying
> down.
>
> The dynamic range is the ratio between them, exactly as Kevin said.

That's not right.  That is the DENSITY range between the two, and is not the
dynamic range.

Austin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 7:19 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


(snip)

> But the equation for dynamic range does not allow YOU to select the
> "meaningful" definition of smallest and largest as YOU understand them.
> They are defined by the equation.

Austin,

You write the equations to reflect the meaningful relationships in what you
are examining. The dynamic range equation is an engineering expression for
certain systems, not a law of physics.
>
(snip)
> >
> > For the most part I feel you are simply substituting synonyms.
>
> They you are missing the entire concept.  What I have shown, quite
clearly,
> is that there are AMBIGUOUS definitions of both largest and smallest.  I
> have also shown what the two ambiguous definitions are.

In a print you can choose to start at either end of the scale but the two
values are the darkest black and the whitest white. The usual description is
Density calculated from reflectance.
>
> > The point
> > where we collide is over "discernable."
>
> But it's clearly defined by the equation and what dynamic range is there
to
> measure.  As I've said time and time again, you are arguing over a very
> clear concept/definition/equation that I did not come up with.  I am
merely
> trying to get you to understand what it really is, and how your
> understanding of it is flawed.  Unfortunately, for what ever reason I do
not
> know, it isn't working.

Your equation is fine and works very well in electronic and digital imaging
but since you have not defined or quantified the noise of a static print, I
don't see how it can be aplied. If there is no noise or if the noise is so
low it can be ignored then the relationship you are showing would not have
any relevance.
>
> > I agree with this as part of the
> > process of determining the min and max.
>
> I don't follow you there.

Knowing the noise floor and/or ceiling is part of determining the usable min
and max.
>
> > You move on into the idea of
> > "discernable difference" over the entire tonal range and I think that
this
> > is not covered by the general concept of dynamic.
>
> Well, then what do YOU believe is "dynamic" in dynamic range?  That is a
> very important question.

The ability to change tone.
>
> > "Discernable difference"
> > is very important when you are talking about systems where the input and
> > output are changing with time
>
> Time has nothing to do with it.  You do not need time to discern change,
but
> you do need change to discern time (off topic, but it's true anyway).  Two
> things can have a discernable difference at the exact same time!

True but time is not a factor in a print.
>
> > I do not see how you can directly "measure" the
> > dynamic range of a print. Perhaps in electronics you can.
>
> Very simply.  Print a completely solid tone, with as minimum detectable
> difference in density in the original as possible.  Print the darkest tone
> you can.  Measure the darkest tone, measure the paper, then take multiple
> measurements of the "patch" of your "test" tone.  The variance across the
> test tone is your noise, and yes, I know, you have to account for the
> original variances, as well as possible variances caused by the
lense...but
> none the less, it does give you a basis for noise.  You get the "largest
> signal" in the dynamic range equation by your dMax - dMin, and you get
your
> "smallest discernable signal" from the variance in your "patch".  There
you
> have it.

I am sorry. I misunderstood you to mean that you could take a single
measurement to get a value for dynamic range. So you are taking multiple
measurements and getting the DR by calculation. Your idea for determining
the noise is an interestin one and similar to what John suggested earlier
regarding using variations in the paper at Dmin.

What tonal value do you print for your test patch and how do you know that
the variation at this single tone will be representative of all the tones in
the range?

From a practical point of view I think that this will be difficult to
measure, at least with the densitometers I have seen but it certainly could
be done in some fashion. There is also the issue of what is contributing to
this variation and to what degree. Variation in the paper, the ink or the
way the ink is laid down. This could give you multiple noise sources that
contribute differently at each end of the range.

Finally what do you conclude if this turned out to be a very small number?
Would your concept of dynamic range still be of value in comparing two
different print media?
>
(snip)
>
> As I've said countless times, min and max WHAT?  What does min and max
mean?
> They are incomplete terms.

Whitest white and blackest black.
>
>  This does not seem to account for the values in
> > between or
> > they assumed a continuously variable system.
>
> Min and Max does if you take max to mean the maximum signal and min to
mean
> the minimum discernable signal ;-)

I agree, but we have not clarified what the minmum discernable signal might
be or if it is of significant magnitude.
>
> > > What's the dynamic range?  Well, we don't know what the smallest
> > discernable
> > > signal is, now do we?  Nor do we really know the largest that the
> > > paper/system can attain either.  This makes for a quandary.
> >
> > Not really because you are interested in the meaningful or useful values
> > which may simply be the only the values you do know.
>
> Sorry, I have no idea what you mean by that...

That you may have to describe a reationship based upon the values you do
know and neglect those that cannot be determined. In other words if you
cannot determine your "minimum discernable signal" you would need to
describe the density relationship in some other fashion.
>
> > I agree that you may be able to discern a change of 0.01 density
> > near Dmin.
> > The problem is that as you approach Dmax the density change you
> > can discern
> > decreases.
>
> You're right, but only if you're above 3!!!  It's also not relevant to the
> discussion, but is a noted point.

From my own experience with a MacBeth densitometer the ability to discern
tonal differences declined steadily as you increased print density
especially above 2 and stopped around 2.4. I was doing a lot of contact
exposures with a Stouffer step wedge to measure the different brands of
silver paper and the effects of different developers and toning processes.
Why do you say 3?
>
> > The discernable difference is not a constant in this
> > system which
> > is why I object to using this in a dynamic range calculation for this
> > system.
>
> That's the first you've said that!  But...as I said, your claim is only
good
> for density values above 3, below 3, they are quite constant...and since
> this IS a B&W list...and as we all know, B&W doesn't go above 2...I think
> we're safe ;-)

Yes it is the first time I have said that. But remember I do not believe
that our ability to discern tones has anything to do with the number of
tones present. Well I would appreciate where the density of three comes in
and I assure that silver prints do go up to 2.6 from my own measurements and
others have reported as high as 2.8. I am not convinced that it is constant
since I personally can easily detect the difference between a density of
0.02 and 0.04 but not so easily between 2.36 and 2.40. I believe we can
differentiate the highlights easier than the shadows.
>
> > From your earlier diagram you indicated
> > that the discernable difference was constant across the entire range.
That
> > is not the case here.
>
> Well, it is the case, but even if it weren't, it's not relevant.  Did you
> ever take Differential Equations?

Well if you are going to do the division in your equation you need to choose
a value. So if it varies what do you use? Yes but long ago and far away.
>
(snip)
> >
> > I believe I did. From the definition the dynamic range is the ratio of
the
> > maximum divided by the minimum.
>
> That's NOT an equation.  AND maximum and minimum are left undefined.

Why is the lightest and darkest points on a print so vague?
>
> > Dynamic Range of the Reflectance = (maximum reflectance of the brightest
> > visually discernable white)/(minimum reflectance of the darkest visually
> > discernable black)
>
> Yeah, but that's wrong.  You have just given an equation for DENSITY
range,
> not dynamic range.

Not according to most of the definitions I posted earlier. And yes the
dynamic range of the reflectance is the density range. Remeber the
difference of two logs express a ratio.
>
> > My thought is that the dynamic range of a print may not be a
particularly
> > useful expression for us. The simple range or knowing the lightest and
> > darkest shades of gray of a medium are what is of importance to
> > us.
>
> But it is so obvious that some images may have NO tones between black and
> white, and some may have a LOT of tones between black and white.  That is
> why dynamic range IS important.

Well that is an artistic decision by whoever made the print and not a result
of the dynamic range of the materials unless they were specifically designed
that way like litho film.
>
> > We also
> > tend to overly focus on getting the deepest black where in
> > reality you gain
> > more by increasing the brightness white.
>
> I very much so agree with that!

Wow! Austin and I agreed on something! Where's the champaign? <G>
>
> > I understand that you are trying to give us an expression that says
> > something about the number of discernable tones that a media or system
can
> > create but I don't think that this can be done here because I do not see
> > that there are a finite number of tones.
>
> And I do see a finite number of tones, or at least a "range" of the number
> of tones.  I can certainly say one has more or less than another!

Well give me a rough estimate of the number of tones you think there are in
an inkjet and silver prints.

> I have
> done the same print that has very bad tonality, there aren't many
> discernable tones, and I have re-printed the exact same image and managed
to
> get a heck of a lot more tones out of that same image...via development
> time/exposure.

Same here, but is this meaningful information about the materials or about
our skills?

  Similarly with film, I have film that has the blacks so
> blocked up, and some that has them nice and open...and that was based on
> development/exposure too ;-)  Dynamic range is VERY important to me, as I
> want full tonality in my specific prints.  Density range isn't as
important
> to me, for reasons that have been mentioned by you and others.

Here it seems you are talking about zone placement with expanded and
contracted development. This is an issue of how do you map the real world
tones you want to capture onto your film. Whether you do this well or badly
the film will still go from base fog + film density to totaly blocked and
contain every tone in between. You do not add or subtract the number of
tones in the negative by adjusting your exposure or or development,
(assuming you are at least in the ballpark and took your lens cap off and
remembered to put the dark slide back in) you are adjusting which specific
tones in the scene wind up in the usable range of the negative.
>
> You seem to be shifting a bit here...more to an argument that dynamic
range,
> what ever we believe it is, isn't really that important...  Just a note.

In general, I think that knowing the Dmin, Dmax and the contrast grade of a
particular silver paper is adequate and for inkjet just the Dmin and Dmax
since the software seems to give us control over the contrast. I sure do
want to know the DR of my scanners though!

Best,
Martin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 7:23 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


(snip)
> >
> > I tried it and the on screen image looks terribly degraded and
> > posterized in
> > comparison to 32-bit color monitor setting.
>
> Did you really set your monitor to 256 shades of gray?  I do not have that
> ability, I only have the ability to set it to 256 COLORS, not to 256
shades
> of gray...that is very different...256 colors will give you virtually no
> shades of gray.

Good point. You're right I only have three settings available 256 colors,
16-bit and 32-bit.

Martin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Thursday, March 28, 2002 8:23 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


>
> > My
> > response above was to Austin who said that you could measure it directly
> > without measuring the Dmin and Dmax individually.
>
> Er, Martin...I never said that.  EVER.  I have ALWAYS said you need to
> measure the dMin and dMax as (dMax - dMin) IS the "largest signal" part of
> the dynamic range equation.

I am sorry. I misundersood you.

Martin

>>> For my
>>> work I couldn't live with just 256 shades of gray and would still be
>>> sloshing around in the dark.
>> 
>> Hmm. Next time you're working a grayscale image on your computer change
> your
>> monitor from millions of colors to 256 shades of gray and tell me if your
>> image looks any worse for wear.
> 
> Todd,
> 
> I tried it and the on screen image looks terribly degraded and posterized in
> comparison to 32-bit color monitor setting.

Really? You selected grayscale, not 256 shades of color? The difference must
be in our graphics card. No matter.
 
> I really wish I had some comfort level as to how many shades of gray can be
> perceived by the human eye or at least a range since there is bound to be
> person to person variation. I have heard 100, 256 up to 1030. The most
> thorough study which I neglected to save was done in Sweden and they
> reported that most people could differentiate between 800 and 900 shades of
> gray but that artists and people in the graphics industry could see 900 to
> the 1030 number. Unfortunately they did give any details of how the testing
> was done. I hate newspaper articles about scientific tests.

I'm sure it depends on how they test for this stuff. The test methodology,
as well as the environmental conditions, will obviously have a huge impact.
I wonder if every institute even performs the same test, and if so, is it
even a well designed test?
 
> I think that if it is some fixed number of say 256 you can have multiple
> sets of 256 tones in an image that contribute to its impact. Say you could
> only differentiate 256 but you had 1024 tones. Your eye could see the
> difference between 1 and 4, 4 and 8, 8 and 12, etc. You could also
> distinguish between 2 and 5, 5 and 9, 9 and 13, etc.

I have no idea how it all works. I guess the point is how many tones are
needed to trick perception into believing all the tones are present. How
many tones can be deleted before the eye notices them as missing. That sort
of thing.

I mean it doesn't even seem to be cut and dry as to whether staying in a
16-bit workflow and printing 16-bits through the Piezo driver makes a
visible difference in print, and that can be the difference between 256
shades and tens of thousands. (And to those who've heard that printing
16-bits through piezo does make a visible difference, I need to ask, did the
8-bit version it was compared to contain all 256 tones? You have to give the
8-bit version a fighting chance...)

Todd

on 3/28/02 6:38 PM, Michael Kravit wrote:

> Austin, et. al.
> 
> There are three important scales within the totaal range of exposures that
> can be printed. The full range from black to white is represented by Zones 0
> to X. Within this range lies the dynanic range, representing the first
> useful values above Zone 0 and below Zone X, or Zones I to IX. The range of
> zones which convey definite qualities of texture and the recognition of
> substance is the textural range from Zones II to VII. (1)
> 
> (1) Ansel Adams, The Negative, Little Brown, Nineteenth Printing, 1998.

Nobody ever said Ansel was an engineer...

Todd

> You get the print driver to output the lightest continuous (yeah I know)
> tone it is capable of short of not printing at all. You measure the
> reflectance of that patch.
>
> You measure the reflectance of the darkest tone it is capable of laying
> down.
>
> The dynamic range is the ratio between them, exactly as Kevin said.

Hi Johnny,

What I originally though was wrong with the above isn't what's wrong with
it.  What the "equation" is missing is dMin.

Your equation was (dMax / "noise"), instead of ((dMax - dMin) / "noise").

Austin

> (snip)
> > >
> > > I tried it and the on screen image looks terribly degraded and
> > > posterized in
> > > comparison to 32-bit color monitor setting.
> >
> > Did you really set your monitor to 256 shades of gray?  I do
> not have that
> > ability, I only have the ability to set it to 256 COLORS, not to 256
> shades
> > of gray...that is very different...256 colors will give you virtually no
> > shades of gray.
>
> Good point. You're right I only have three settings available 256 colors,
> 16-bit and 32-bit.
>
> Martin

Yeah, me too, as I wanted to try what Todd was suggesting.  It may be a Mac
thing?  Are you on a Mac or a PC Todd?

Austin

on 3/29/02 4:46 AM, Austin Franklin wrote:

>> Good point. You're right I only have three settings available 256 colors,
>> 16-bit and 32-bit.
>> 
>> Martin
> 
> Yeah, me too, as I wanted to try what Todd was suggesting.  It may be a Mac
> thing?  Are you on a Mac or a PC Todd?
> 
> Austin


Morning Austin ;-)

I'm on a Mac. 

Todd

>
> > But the equation for dynamic range does not allow YOU to select the
> > "meaningful" definition of smallest and largest as YOU understand them.
> > They are defined by the equation.
>
> Austin,
>
> You write the equations to reflect the meaningful relationships
> in what you
> are examining. The dynamic range equation is an engineering expression for
> certain systems, not a law of physics.

I never said it was a "law of physics", but it IS a specific measurement,
and as I said, it IS defined by a fixed, non-negotiable (except different
systems have some different multipliers, but the basic equation, and
meaning, is the same) equation with defined relationships.

> > They you are missing the entire concept.  What I have shown, quite
> clearly,
> > is that there are AMBIGUOUS definitions of both largest and smallest.  I
> > have also shown what the two ambiguous definitions are.
>
> In a print you can choose to start at either end of the scale but the two
> values are the darkest black and the whitest white. The usual
> description is
> Density calculated from reflectance.

What does that have to do with dynamic range?  I'm missing your point here.
For "largest", of course you need to measure the darkest black and whitest
white and "largest" is the difference of the two...which is the same as
density range...but to derive dynamic range from that, you divide that by
the "smallest discernable signal".

> Your equation is fine and works very well in electronic and
> digital imaging
> but since you have not defined or quantified the noise of a
> static print, I
> don't see how it can be aplied.

OK.  I'm confused.  At first you said density range was the same as dynamic
range, and that dynamic range didn't require noise...are you now agreeing
that density range and dynamic range ARE different, and that dynamic range
does require noise?

> > > I agree with this as part of the
> > > process of determining the min and max.
> >
> > I don't follow you there.
>
> Knowing the noise floor and/or ceiling is part of determining the
> usable min
> and max.

Not at all.  Noise floor and min are two entirely different things.  I've
given examples of that being true.  Noise "ceiling" has nothing to do with
max either...  I'm not getting your point here.

> > Well, then what do YOU believe is "dynamic" in dynamic range?  That is a
> > very important question.
>
> The ability to change tone.

OK, then how does only using dMax and dMin give you anything dynamic?

> > > "Discernable difference"
> > > is very important when you are talking about systems where
> the input and
> > > output are changing with time
> >
> > Time has nothing to do with it.  You do not need time to discern change,
> but
> > you do need change to discern time (off topic, but it's true
> anyway).  Two
> > things can have a discernable difference at the exact same time!
>
> True but time is not a factor in a print.

I'm happy to take the "true" admission, and work on the application to a
print at another time ;-)

> I am sorry. I misunderstood you to mean that you could take a single
> measurement to get a value for dynamic range. So you are taking multiple
> measurements and getting the DR by calculation. Your idea for determining
> the noise is an interestin one and similar to what John suggested earlier
> regarding using variations in the paper at Dmin.

Oh, oh...is this a breakthrough?

> What tonal value do you print for your test patch and how do you know that
> the variation at this single tone will be representative of all
> the tones in
> the range?

You can do various samples.

> From a practical point of view I think that this will be difficult to
> measure, at least with the densitometers I have seen but it
> certainly could
> be done in some fashion.

OK, now I feel like we are getting somewhere ;-)

> There is also the issue of what is
> contributing to
> this variation and to what degree. Variation in the paper, the ink or the
> way the ink is laid down. This could give you multiple noise sources that
> contribute differently at each end of the range.

Absolutely agree.

> Finally what do you conclude if this turned out to be a very small number?
> Would your concept of dynamic range still be of value in comparing two
> different print media?

I believe so, but I haven't thought about that at all.  I'm just trying to
lay down the foundation for that part of the discussion.

> > As I've said countless times, min and max WHAT?  What does min and max
> mean?
> > They are incomplete terms.
>
> Whitest white and blackest black.

Those definitions only work for density range, not dynamic range.  min and
max mean something different in the dynamic range equation.

> >
> >  This does not seem to account for the values in
> > > between or
> > > they assumed a continuously variable system.
> >
> > Min and Max does if you take max to mean the maximum signal and min to
> mean
> > the minimum discernable signal ;-)
>
> I agree, but we have not clarified what the minmum discernable
> signal might
> be or if it is of significant magnitude.

That's OK for now.

> >
> > > > What's the dynamic range?  Well, we don't know what the smallest
> > > discernable
> > > > signal is, now do we?  Nor do we really know the largest that the
> > > > paper/system can attain either.  This makes for a quandary.
> > >
> > > Not really because you are interested in the meaningful or
> useful values
> > > which may simply be the only the values you do know.
> >
> > Sorry, I have no idea what you mean by that...
>
> That you may have to describe a reationship based upon the values you do
> know and neglect those that cannot be determined. In other words if you
> cannot determine your "minimum discernable signal" you would need to
> describe the density relationship in some other fashion.

But minimum discernable signal only applies to dynamic range, not to density
range/relationship...

> > > I agree that you may be able to discern a change of 0.01 density
> > > near Dmin.
> > > The problem is that as you approach Dmax the density change you
> > > can discern
> > > decreases.
> >
> > You're right, but only if you're above 3!!!  It's also not
> relevant to the
> > discussion, but is a noted point.
>
> From my own experience with a MacBeth densitometer the ability to discern
> tonal differences declined steadily as you increased print density
> especially above 2 and stopped around 2.4. I was doing a lot of contact
> exposures with a Stouffer step wedge to measure the different brands of
> silver paper and the effects of different developers and toning processes.
> Why do you say 3?

X-Rite 810 manual ;-)

> > > The discernable difference is not a constant in this
> > > system which
> > > is why I object to using this in a dynamic range calculation for this
> > > system.
> >
> > That's the first you've said that!  But...as I said, your claim is only
> good
> > for density values above 3, below 3, they are quite constant...and since
> > this IS a B&W list...and as we all know, B&W doesn't go above
> 2...I think
> > we're safe ;-)
>
> Yes it is the first time I have said that. But remember I do not believe
> that our ability to discern tones has anything to do with the number of
> tones present.

No, not OUR ability to discern tones, but some methodology for discerning
tones.

> Well I would appreciate where the density of three comes in
> and I assure that silver prints do go up to 2.6 from my own
> measurements and
> others have reported as high as 2.8. I am not convinced that it
> is constant
> since I personally can easily detect the difference between a density of
> 0.02 and 0.04 but not so easily between 2.36 and 2.40. I believe we can
> differentiate the highlights easier than the shadows.

I completely agree with you on this.

> >
> > > From your earlier diagram you indicated
> > > that the discernable difference was constant across the entire range.
> That
> > > is not the case here.
> >
> > Well, it is the case, but even if it weren't, it's not
> relevant.  Did you
> > ever take Differential Equations?
>
> Well if you are going to do the division in your equation you
> need to choose
> a value. So if it varies what do you use? Yes but long ago and far away.

You can integrate it very easily.  This is not a problem.  Your
concern/point is entirely valid, but it can easily be dealt with.


> > > I believe I did. From the definition the dynamic range is the ratio of
> the
> > > maximum divided by the minimum.
> >
> > That's NOT an equation.  AND maximum and minimum are left undefined.
>
> Why is the lightest and darkest points on a print so vague?

They aren't, but "maximum" and "minimum" are vague...and as I've pointed
out, each can mean one of two things.

> And yes the
> dynamic range of the reflectance is the density range.

Sigh ;-(

> Well give me a rough estimate of the number of tones you think
> there are in
> an inkjet and silver prints.

It depends on the print, of course!  Inkjet, using the Cone system is over
100.  Silver I have not "counted", but I can easily tell when one print has
"more" than another.  The same negative printed in a wet darkroom, that has
a high dynamic range (many tones ;-) shows more tones on the inkjet print
than the wet print.

> > I have
> > done the same print that has very bad tonality, there aren't many
> > discernable tones, and I have re-printed the exact same image
> and managed
> to
> > get a heck of a lot more tones out of that same image...via development
> > time/exposure.
>
> Same here, but is this meaningful information about the materials or about
> our skills?

Understood.

>   Similarly with film, I have film that has the blacks so
> > blocked up, and some that has them nice and open...and that was based on
> > development/exposure too ;-)  Dynamic range is VERY important
> to me, as I
> > want full tonality in my specific prints.  Density range isn't as
> important
> > to me, for reasons that have been mentioned by you and others.
>
> Here it seems you are talking about zone placement with expanded and
> contracted development. This is an issue of how do you map the real world
> tones you want to capture onto your film. Whether you do this
> well or badly
> the film will still go from base fog + film density to totaly blocked and
> contain every tone in between. You do not add or subtract the number of
> tones in the negative by adjusting your exposure or or development,
> (assuming you are at least in the ballpark and took your lens cap off and
> remembered to put the dark slide back in) you are adjusting which specific
> tones in the scene wind up in the usable range of the negative.

Not true!  I have negatives that are completely blocked up, and almost look
like litho film!  The fact that there IS litho film backs up my statement.
It doesn't necessarily have to do with Zone etc, but can be only related to
development...trust me!  I've done some lousy development in the past.

> I sure do
> want to know the DR of my scanners though!

Get a high range target and scan it ;-)

Regards,

Austin

> >> Good point. You're right I only have three settings available
> 256 colors,
> >> 16-bit and 32-bit.
> >>
> >> Martin
> >
> > Yeah, me too, as I wanted to try what Todd was suggesting.  It
> may be a Mac
> > thing?  Are you on a Mac or a PC Todd?
> >
> > Austin
>
>
> Morning Austin ;-)
>
> I'm on a Mac.
>
> Todd

Ha, you new father!  Are you on baby duty ;-)  My 2 year old son woke my up
wanting "Orangina" at about 3:30...and I haven't been able to get back to
sleep...I had a very physically tiring day yesterday (much less this dynamic
range discussion ;-)

I thought the Mac let you set the screen to grayscale!  Good thing.

Austin

> I just don't think that you can take those
> relationships that
> apply to digital imaging and apply them to a finished print
> silver or inkjet
> or whatever without stepping back and seeing if this is
> appropriate or if it
> is telling you something meaningful about the print.

Hi Martin,

And I completely agree with that...but I personally DO believe it's
appropriate and meaningful, hence the discussion and my willingness to try
to muddle through this ;-)

Regards,

Austin

Austin,

This is going to be a noisy/rambling post - I'm tired, I don't think I have
sufficient command of the jargon to express this as precisely as may be
required, and I'm having trouble to get my heed around some of this stuff.
But here goes, please bear with me and see what you can make of it.

Martin asked:
>> I do not see how you can directly "measure" the dynamic range of a print.
>> Perhaps in electronics you can.

Austin replied: 
> Very simply.  Print a completely solid tone, with as minimum detectable
> difference in density in the original as possible.  Print the darkest tone you
> can.  Measure the darkest tone, measure the paper, then take multiple
> measurements of the "patch" of your "test" tone.  The variance across the test
> tone is your noise, and yes, I know, you have to account for the original
> variances, as well as possible variances caused by the lense...but none the
> less, it does give you a basis for noise.  You get the "largest signal" in the
> dynamic range equation by your dMax - dMin, and you get your "smallest
> discernable signal" from the variance in your "patch".  There you have it.


Earlier in the thread you have said (I'm paraphrasing) that the dynamic
range of a print is an expression of how many tones that print contains.

I don't see how the simple test you suggest, which only differs from the
more common test for dynamic range (or what you call density range) is just
the additional step of measuring the variance across your test patch. I
don't see how that information tells you how many tones are in the print.
This would seem to be a case where, as you've said before, only 2 bits of
info is needed to establish dynamic range, no? But this then could suggest
the material is capable of a huge dynamic range, when in fact only three
tones are present.

Furthermore, what does it say about a scenario where ones printer/workflow
happens to produce a certain granularity/banding/mottling/artifacting
precisely in the tonal region of your test patch, but nowhere else; or
worse, the converse, everywhere but your patch?

And what of all this inkjet artifacting? I suppose it's debatable whether or
not it affects tonality, but surely it lowers dynamic range via the
inclusion of greater noise in the formula. Lets assume you can have
microbanding, which does not destroy the tonality of n area, while it does
heighten the noise of it. If dynamic range is a description of the number of
discrete tones in an image, but noise can reduce the dynamic range while at
the same time not reduce the number of tones, something is screwy in the
logic of the formula.

Furthermore, I agree that there needs to be some term to distinguish between
a litho-like substance, and a continuous tone substance, where the two could
conceivably have the same density range, while one is capable of far more
shades in between. But I wonder now if dynamic range IS that descriptor,
when in fact you could measure a three toned medium and presume it to have
the same number of tones and the continuous tone item, based upon that test
and calculation. So, at best dynamic range can provide a predictor of what a
medium is capable of, but it assumes a complete and linear set of tones
between the ends, which may or may not be so; at worst it could mislead one
wildly.

I need to explore this subject of linearity more fully as it pertains to
dynamic range. For instance, a print with an expansion of highlights and
contraction of shadows (= high contrast) will contain fewer tones than the
material is capable of, but your measure of dynamic range will not indicate
that has occurred. So, is dynamic range a measure of the number of tones an
item contains, or a measure of what the material is capable of if perfect
linearity is present?

Now I presume this instance of not exploiting the range of this material in
a linear fashion, in order to utilize the maximum number of tones the
material is capable of, is not a failing of the materials per se, but a
failing of the operator to utilize it in that way.

This has me wonder why you assume a Piezo print to have greater dynamic
range than a silver print given these considerations:

A) We know a well handled glossy fiber print is capable of a greater density
range than Piezo.

B) We know the source material of a silver print (negative) is capapble of
more tones (millions? billions?) than the source material for an inkjet
print (256? 65,000?).

C) Noise can go either way, depending on which papers are used in each
respective system, and other workflow considerations.

What else needs to be accounted for? Linearity. Admittedly this is the main
advantage of Piezo - it's easier (presumably) to acheive linearity in Piezo
than in silver, but that's not to say it's impossible to do it with silver.
Furthermore, there are several inkjet artifacts that can reduce the
linearity of Piezo, so it's not a guarantee.

I'm trying to get my head around my sense that the approach you advocate for
determining the dynamic range of a print is only marginally more informative
than the approach to determining density range (which is fine, you didn't
invent the formula <G>); that measuring the dynamic range of a print still
doesn't tell you how many tones it contains; that I see no logic that a
Piezo print is any more capable of greater dynamic range than silver, other
than the assumption that it is easier to maintain linearity - which with
banding, artifacting, mottling, hardware resolution limits, droplet size,
etc (all of which can raise noise if not tonality too), really isn't a
given.

I told you this would be noisy/rambling. ;-)

Your thoughts please.

Todd

> Earlier in the thread you have said (I'm paraphrasing) that the dynamic
> range of a print is an expression of how many tones that print contains.

Basically correct.

> I don't see how the simple test you suggest, which only differs from the
> more common test for dynamic range (or what you call density
> range) is just
> the additional step of measuring the variance across your test patch.

Exactly.  Since you can only discern by that value, that gives you the
minimum discernable value.  How it differs is quite simple.  Density range
does NOT take this into consideration at all.  It is merely a range between
two static values.  Dynamic range is how many discernable values you get
across that range.  Case in point is you can have a print with a high
density range, pure black on a pure white background, but it has no dynamic
range...no tones in between.

> I
> don't see how that information tells you how many tones are in the print.

You can only discern what you can discern, and anything that can't be
discerned is the same.


> This would seem to be a case where, as you've said before, only 2 bits of
> info is needed to establish dynamic range, no?

No, only two bits are needed to represent the DENSITY range.

> But this then could suggest
> the material is capable of a huge dynamic range, when in fact only three
> tones are present.

No, huge density range, but very low dynamic range.

> Furthermore, what does it say about a scenario where ones printer/workflow
> happens to produce a certain granularity/banding/mottling/artifacting
> precisely in the tonal region of your test patch, but nowhere else; or
> worse, the converse, everywhere but your patch?

It's merely an example.  You can quantify your system any way you want.
Those details are not relevant to the concepts involved that are trying to
be discussed here.

> And what of all this inkjet artifacting?

We aren't talking about inkjet prints in this example...only chemical
prints.

> Furthermore, I agree that there needs to be some term to
> distinguish between
> a litho-like substance, and a continuous tone substance, where
> the two could
> conceivably have the same density range, while one is capable of far more
> shades in between. But I wonder now if dynamic range IS that descriptor,

It is.

> when in fact you could measure a three toned medium and presume it to have
> the same number of tones and the continuous tone item, based upon
> that test
> and calculation. So, at best dynamic range can provide a
> predictor of what a
> medium is capable of, but it assumes a complete and linear set of tones
> between the ends, which may or may not be so; at worst it could
> mislead one
> wildly.

Not at all.  Dynamic range can be measured/calculated as precisely as you
like.  It's like measuring the width of a 4 x 8 sheet of plywood.  You can
just say it's 4' wide, or you can take a measurement every 1' or every 1" or
over any distance you want...  I'm just showing a measurement technique
here...there is nothing about dynamic range, per se, in this plywood
example.

> I need to explore this subject of linearity more fully as it pertains to
> dynamic range. For instance, a print with an expansion of highlights and
> contraction of shadows (= high contrast) will contain fewer tones than the
> material is capable of, but your measure of dynamic range will
> not indicate
> that has occurred. So, is dynamic range a measure of the number
> of tones an
> item contains, or a measure of what the material is capable of if perfect
> linearity is present?

It entirely depends on what you are measuring and how you measure it.

> This has me wonder why you assume a Piezo print to have greater dynamic
> range than a silver print given these considerations:

Because I have printed the exact same negative on both, and the Piezo print
had more tones visible to the eye.

> A) We know a well handled glossy fiber print is capable of a
> greater density
> range than Piezo.

Which is not really relevant to dynamic range...

> B) We know the source material of a silver print (negative) is capapble of
> more tones (millions? billions?) than the source material for an inkjet
> print (256? 65,000?).

I don't believe we know that.

> C) Noise can go either way, depending on which papers are used in each
> respective system, and other workflow considerations.

Absolutely.

> I'm trying to get my head around my sense that the approach you
> advocate for
> determining the dynamic range of a print is only marginally more
> informative
> than the approach to determining density range (which is fine, you didn't
> invent the formula <G>);

But density range only tells you the relative distance between the
endpoints, and as we have discussed in the past, can be represented with
only two bits...and as the simple equation for density range shows, it
merely is (dMax - dMin) and is a static measurement.

> that measuring the dynamic range of a print still
> doesn't tell you how many tones it contains;

It certainly does, as defined by the equation for dynamic range ((dMax -
dMin) / smallest discernable value).  Note, it has an additional variable in
the equation beyond what density range does.  The dynamic range of a
chemical print certainly can be done as a range.  You can add any complexity
to it you want...but it doesn't change the basic concept.

> that I see no logic that a
> Piezo print is any more capable of greater dynamic range than
> silver,

It doesn't have logic associated with it, it's simply observation.

> I told you this would be noisy/rambling. ;-)

You had a respectable dynamic range here ;-)

> Your thoughts please.

Anything else?

Regards,

Austin

P.S.  Please keep it short if you could...

Todd,

BTW, the source of the noise, though it may be interesting to know, isn't
really relevant to what the dynamic range is.  It is THE print that we are
trying to determine the dynamic range of, and it is what it is...no matter
what caused it to be what it is.  So, if the source is the paper, lense,
negative, developer...what ever...won't matter.

Regards,

Austin

----- Original Message -----

From: "Todd Flashner" <tflash@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Friday, March 29, 2002 12:10 AM
Subject: Re: [Digital BW] Dynamic Range Definitions and Print Tones


(snip)
>
> Really? You selected grayscale, not 256 shades of color? The difference
must
> be in our graphics card. No matter.

No I can only get 256 color so I can't do the test.
>
(snip)
> > Unfortunately they did give any details of how the testing
> > was done. I hate newspaper articles about scientific tests.
>
> I'm sure it depends on how they test for this stuff. The test methodology,
> as well as the environmental conditions, will obviously have a huge
impact.
> I wonder if every institute even performs the same test, and if so, is it
> even a well designed test?

I am pretty certain that there is no standard for testing this as we would
be seeing results that were similar
>
> > I think that if it is some fixed number of say 256 you can have multiple
> > sets of 256 tones in an image that contribute to its impact. Say you
could
> > only differentiate 256 but you had 1024 tones. Your eye could see the
> > difference between 1 and 4, 4 and 8, 8 and 12, etc. You could also
> > distinguish between 2 and 5, 5 and 9, 9 and 13, etc.
>
> I have no idea how it all works. I guess the point is how many tones are
> needed to trick perception into believing all the tones are present. How
> many tones can be deleted before the eye notices them as missing. That
sort
> of thing.

Two things on that. You have the question of whose eyes and they do seem to
vary and we really don't want to be right on the edge. I think to fool the
eye you need some multiple of that number.
>
> I mean it doesn't even seem to be cut and dry as to whether staying in a
> 16-bit workflow and printing 16-bits through the Piezo driver makes a
> visible difference in print, and that can be the difference between 256
> shades and tens of thousands. (And to those who've heard that printing
> 16-bits through piezo does make a visible difference, I need to ask, did
the
> 8-bit version it was compared to contain all 256 tones? You have to give
the
> 8-bit version a fighting chance...)

The file may be 8-bit but I don't know what the print is. Since you have 4
or 6 inks and possibly 256 bits for each ink, the effect of the dither to
create additional tones, the ink drops may bleed into each other or the
paper. Once its dry what is there in terms of a number of tones is pretty
vague which is why I am thinking it may be a continuous tone without
discrete values. Which may also be why you can lose so many tones in the
histogram and still get a good print. I keep thinking there must be
technical papers on this by Epson or someone.

Martin

I had some time to think about this last night as I licked 100 pieces of
photorag trying to get them right side up  in the box that my 2 year
"helped" me with.  I has helped me clarify my thinking - hopefully it is not
just catching up with the rest of you.

1.  I now think you do need both density range and dynamic range to
characterise the tones in a print.  Clearly the blackest black is important
and the whitest white is important -- these are measured by their density
and the difference would be density range. Also the ability to render smooth
transitions from one tone to another across the density range is important
and this would be measured by Dynamic Range.

2.  I think dynamic ranges as Austin has described it to us is correct.  The
problem I was having was in assessing the minimum perceptible change value.

3.  The dynamic range is a characterisation of a system, not an attribute.
In the case of looking at a print I would suggest that the system is
comprised of the print, the incident light and, importantly, our eyes.  This
means that the minimum _perceptible_ change is what our eyes can
discriminate not quantum level stuff, or even microscopic changes like the
ink droplets or silver bits on paper.

This means that the dynamic range of a system measuring the density of a
print with a spectrophotometer will likely be different than the dynamic
range of the system using our eyes.

4. The dynamic range of my prints (using the piezo driver v5, and epson
1200, and MIS FS inks) has a maximum of:

-10log10((measured paper black density)-(measured paper white density)/256))

Since the originating image only contains 256 shades of grey this caps the
upper limit (ie smallest value of) on the minimum perceptible change.  If my
eyes can only perceive 100 shades of grey then the dynamic range is lower.
If my eyes can see 1000 shades of grey the dynamic range is still capped by
the 256 shades that the print originated in.

To assess whether a silver print has more or less dynamic range you would
have to use a "system" that can accurately discriminate as many distinct
tones as our eyes and see if either print reaches its discrimination limit
before the limit of our eyes is reached.

For me, this framework works well.  Does it make any sense to others?

Kevin Gulstene

On 3/28/02 8:28 PM, "Austin Franklin" <darkroom@...> wrote:

> 
>>> I do not see how you can directly "measure" the
>>> dynamic range of a print. Perhaps in electronics you can.
>> 
>> It seems simple to me.
>> 
>> You get the print driver to output the lightest continuous (yeah I know)
>> tone it is capable of short of not printing at all. You measure the
>> reflectance of that patch.
>> 
>> You measure the reflectance of the darkest tone it is capable of laying
>> down.
>> 
>> The dynamic range is the ratio between them, exactly as Kevin said.
> 
> That's not right.  That is the DENSITY range between the two, and is not the
> dynamic range.
> 
> Austin
> 
> 
> 
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and other
> resources as they are often being updated. The page is at:
> 
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
> 
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
> them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or "flames."
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
> resources on the homepage.
> 
> 
> 
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> 
> 
>

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Friday, March 29, 2002 2:07 AM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


(snip)
>
> I never said it was a "law of physics", but it IS a specific measurement,
> and as I said, it IS defined by a fixed, non-negotiable (except different
> systems have some different multipliers, but the basic equation, and
> meaning, is the same) equation with defined relationships.

It is only good if those relationship exist or exist to a meaningful degree
in a system.
>
(snip)
>
> What does that have to do with dynamic range?  I'm missing your point
here.
> For "largest", of course you need to measure the darkest black and whitest
> white and "largest" is the difference of the two...which is the same as
> density range...but to derive dynamic range from that, you divide that by
> the "smallest discernable signal".

In photography the dynamic range has been takes as the density range which
happens to be the reflectance ratio and fulfills the general definition of
dynamic range as the ratio of the min and max signals for a photographic
system. This is not a sophisticated enough definition to define dynamic
range for digital imaging where your equation is appropriate.
>
> > Your equation is fine and works very well in electronic and
> > digital imaging
> > but since you have not defined or quantified the noise of a
> > static print, I
> > don't see how it can be aplied.
>
> OK.  I'm confused.  At first you said density range was the same as
dynamic
> range, and that dynamic range didn't require noise...are you now agreeing
> that density range and dynamic range ARE different, and that dynamic range
> does require noise?

First the definition for dynamic range is not a constant and have different
meaning for different fields. I understand that you don't agree with this
and see it as a universal concept. What I am saying here is let's apply your
equation to a photographic print and see if the values help us understand
what we see. In order to do that we need to have a definition of what would
be the noise in a photograph and a value for it.
>
> > > > I agree with this as part of the
> > > > process of determining the min and max.
> > >
> > > I don't follow you there.
> >
> > Knowing the noise floor and/or ceiling is part of determining the
> > usable min
> > and max.
>
> Not at all.  Noise floor and min are two entirely different things.  I've
> given examples of that being true.  Noise "ceiling" has nothing to do with
> max either...  I'm not getting your point here.

Yes but values that are lost below the noise floor are of no interest and
including them as the lower limit of the range would not be helpful.
>
> > > Well, then what do YOU believe is "dynamic" in dynamic range?  That is
a
> > > very important question.
> >
> > The ability to change tone.
>
> OK, then how does only using dMax and dMin give you anything dynamic?

They tell you the size tonal space you can change tones in and they tell you
the size of the largest change you can make. And no I do not believe that
there are a finite number of steps between them so I don't need an
expression to calculate what isn't there.
> >
> > I am sorry. I misunderstood you to mean that you could take a single
> > measurement to get a value for dynamic range. So you are taking multiple
> > measurements and getting the DR by calculation. Your idea for
determining
> > the noise is an interestin one and similar to what John suggested
earlier
> > regarding using variations in the paper at Dmin.
>
> Oh, oh...is this a breakthrough?
>
> > What tonal value do you print for your test patch and how do you know
that
> > the variation at this single tone will be representative of all
> > the tones in
> > the range?
>
> You can do various samples.

Okay, what would be a reasonable number of samples and which tones should we
look at first?
>
> > From a practical point of view I think that this will be difficult to
> > measure, at least with the densitometers I have seen but it
> > certainly could
> > be done in some fashion.
>
> OK, now I feel like we are getting somewhere ;-)
>
> > There is also the issue of what is
> > contributing to
> > this variation and to what degree. Variation in the paper, the ink or
the
> > way the ink is laid down. This could give you multiple noise sources
that
> > contribute differently at each end of the range.
>
> Absolutely agree.
>
> > Finally what do you conclude if this turned out to be a very small
number?
> > Would your concept of dynamic range still be of value in comparing two
> > different print media?
>
> I believe so, but I haven't thought about that at all.  I'm just trying to
> lay down the foundation for that part of the discussion.

Well since you have arrived at a possible concept for noise in a print,
design an experiment to determine a value. I have a Spectrocam
photospectrometer on order and I could try some readings across monotone
patches. I think that probably the variation may be much less that it can
measure but it can't hurt to check. I also hope to have a Howtek D4000
scanner soon. Would that give higher resolution or be usable to attempt such
a measurement?
>
> > > As I've said countless times, min and max WHAT?  What does min and max
> > mean?
> > > They are incomplete terms.
> >
> > Whitest white and blackest black.
>
> Those definitions only work for density range, not dynamic range.  min and
> max mean something different in the dynamic range equation.

Now I am confused again. So what attribute of the print would you use to
determine min and max if not the whitest white and blackest black
>
(snip)

> > I agree, but we have not clarified what the minmum discernable
> > signal might
> > be or if it is of significant magnitude.
>
> That's OK for now.
>
(snip)
> >
> > That you may have to describe a reationship based upon the values you do
> > know and neglect those that cannot be determined. In other words if you
> > cannot determine your "minimum discernable signal" you would need to
> > describe the density relationship in some other fashion.
>
> But minimum discernable signal only applies to dynamic range, not to
density
> range/relationship...

I am just saying that is we can only work with the information we have or
are able to obtain. From that we have to make the best choice we can. If you
can't determine a value for noise to use in your dynamic range equation then
all we may have left are the density or reflectance values.

(snip)
> >
> > From my own experience with a MacBeth densitometer the ability to
discern
> > tonal differences declined steadily as you increased print density
> > especially above 2 and stopped around 2.4. I was doing a lot of contact
> > exposures with a Stouffer step wedge to measure the different brands of
> > silver paper and the effects of different developers and toning
processes.
> > Why do you say 3?
>
> X-Rite 810 manual ;-)

Can you quote what they say or is it too lengthy? I'm curious because it
really is at odds with my own experience.
>
(snip)
> >
> > Yes it is the first time I have said that. But remember I do not believe
> > that our ability to discern tones has anything to do with the number of
> > tones present.
>
> No, not OUR ability to discern tones, but some methodology for discerning
> tones.

Then add into my statement above ..or our ability to discern tones with an
instrument
>
> > Well I would appreciate where the density of three comes in
> > and I assure that silver prints do go up to 2.6 from my own
> > measurements and
> > others have reported as high as 2.8. I am not convinced that it
> > is constant
> > since I personally can easily detect the difference between a density of
> > 0.02 and 0.04 but not so easily between 2.36 and 2.40. I believe we can
> > differentiate the highlights easier than the shadows.
>
> I completely agree with you on this.

Doesn't that suggest that the ability for the eye to discern tones is a
function of the tonal density?
>
> > >
> > > > From your earlier diagram you indicated
> > > > that the discernable difference was constant across the entire
range.
> > That
> > > > is not the case here.
> > >
> > > Well, it is the case, but even if it weren't, it's not
> > relevant.  Did you
> > > ever take Differential Equations?
> >
> > Well if you are going to do the division in your equation you
> > need to choose
> > a value. So if it varies what do you use? Yes but long ago and far away.
>
> You can integrate it very easily.  This is not a problem.  Your
> concern/point is entirely valid, but it can easily be dealt with.

What would the equation look like then?
>
(snip)
>
> Sigh ;-(
>
> > Well give me a rough estimate of the number of tones you think
> > there are in
> > an inkjet and silver prints.
>
> It depends on the print, of course!  Inkjet, using the Cone system is over
> 100.  Silver I have not "counted", but I can easily tell when one print
has
> "more" than another.  The same negative printed in a wet darkroom, that
has
> a high dynamic range (many tones ;-) shows more tones on the inkjet print
> than the wet print.

How who is being vague? <G>
>
(snip)
> >
> > Here it seems you are talking about zone placement with expanded and
> > contracted development. This is an issue of how do you map the real
world
> > tones you want to capture onto your film. Whether you do this
> > well or badly
> > the film will still go from base fog + film density to totaly blocked
and
> > contain every tone in between. You do not add or subtract the number of
> > tones in the negative by adjusting your exposure or or development,
> > (assuming you are at least in the ballpark and took your lens cap off
and
> > remembered to put the dark slide back in) you are adjusting which
specific
> > tones in the scene wind up in the usable range of the negative.
>
> Not true!  I have negatives that are completely blocked up, and almost
look
> like litho film!  The fact that there IS litho film backs up my statement.
> It doesn't necessarily have to do with Zone etc, but can be only related
to
> development...trust me!  I've done some lousy development in the past.

Perhaps this is analogous to stepping outside the limits of the range.
Severe over exposure or over development would be similar to pushing an amp
signal to the point of high distortion.
>
> > I sure do
> > want to know the DR of my scanners though!
>
> Get a high range target and scan it ;-)

What target do you use?

Best,
Martin

> 1.  I now think you do need both density range and dynamic range to
> characterise the tones in a print.  Clearly the blackest black is
> important
> and the whitest white is important -- these are measured by their density
> and the difference would be density range. Also the ability to
> render smooth
> transitions from one tone to another across the density range is important
> and this would be measured by Dynamic Range.

Bingo!  And very eloquently worded.  I hope it didn't take too many licks to
get there ;-)

> 4. The dynamic range of my prints (using the piezo driver v5, and epson
> 1200, and MIS FS inks) has a maximum of:
>
> -10log10((measured paper black density)-(measured paper white
> density)/256))
>
> Since the originating image only contains 256 shades of grey this caps the
> upper limit (ie smallest value of) on the minimum perceptible
> change.

You are correct, assuming your print HAS 256 tones though.  They dynamic
range of the printing system would always be 256 tones, but the print does
not have to contain 256 tonal steps, or 1/256th step per tonal variation.
Printing litho film comes to mind.

> If my
> eyes can only perceive 100 shades of grey then the dynamic range is lower.

I don't believe that lowers the dynamic range of the print...it is what it
is.  The dynamic range YOU can perceive is lower, no doubt.

> If my eyes can see 1000 shades of grey the dynamic range is still
> capped by
> the 256 shades that the print originated in.

Very true.

> To assess whether a silver print has more or less dynamic range you would
> have to use a "system" that can accurately discriminate as many distinct
> tones as our eyes and see if either print reaches its discrimination limit
> before the limit of our eyes is reached.

Some prints are blatantly obvious though.  I don't know that you need to
factor our eyes into this, as you really need to use a measurement technique
that is more accurate than what it is you are trying to measure.  As you
have said, our eyes can only distinguish 100 levels of gray (in any one
lighting condition that is).

Regards,

Austin

Austin,

First off I'd like to acknowledge you and Martin for the time you've put
into this topic. You both have gone far beyond the call of duty. Thanks.

Comments and questions below.
 
>> Earlier in the thread you have said (I'm paraphrasing) that the dynamic
>> range of a print is an expression of how many tones that print contains.
> 
> Basically correct.
> 
>> I don't see how the simple test you suggest, which only differs from the
>> more common test for dynamic range (or what you call density
>> range) is just
>> the additional step of measuring the variance across your test patch.
> 
> Exactly.  Since you can only discern by that value, that gives you the
> minimum discernable value.  How it differs is quite simple.  Density range
> does NOT take this into consideration at all.  It is merely a range between
> two static values.  Dynamic range is how many discernable values you get
> across that range.  Case in point is you can have a print with a high
> density range, pure black on a pure white background, but it has no dynamic
> range...no tones in between.
> 
>> I
>> don't see how that information tells you how many tones are in the print.
> 
> You can only discern what you can discern, and anything that can't be
> discerned is the same.
> 
> 
>> This would seem to be a case where, as you've said before, only 2 bits of
>> info is needed to establish dynamic range, no?
> 
> No, only two bits are needed to represent the DENSITY range.

I thought one bit could give you density range: maximum measurable (max
black), minus min measurable (paper white). Two tones, one bit. And that it
was the addition of the min discernable value (which I'll shorthand to
noise), which you sampled via deviation across the test patch, which gave
you the third component (noise) of your dynamic range equation?
 
>> But this then could suggest
>> the material is capable of a huge dynamic range, when in fact only three
>> tones are present.
> 
> No, huge density range, but very low dynamic range.

I'm confused.

I read you to mean that you get your dynamic range (I'll abbreviate that to
DyR from here on in this post) by measuring the min and max (I'm short
handing here, you know what I mean), subtract the min from the max (which is
your density range (DnR), then dividing that sum by your noise, which you
measured from your third tone. That gives you your DyR.

So you've sampled three tones, referred to as: min, max, and test (or min
discernable, or noise). Let's assume this print was comprised of ONLY these
three tones. You could calculate a very high DyR from these three tones,
implying the print has many intermediary tones, which it does not, no?
 
>> Furthermore, what does it say about a scenario where ones printer/workflow
>> happens to produce a certain granularity/banding/mottling/artifacting
>> precisely in the tonal region of your test patch, but nowhere else; or
>> worse, the converse, everywhere but your patch?
> 
> It's merely an example.  You can quantify your system any way you want.
> Those details are not relevant to the concepts involved that are trying to
> be discussed here.

Okay.
 
>> And what of all this inkjet artifacting?
> 
> We aren't talking about inkjet prints in this example...only chemical
> prints.

Really? First off I wasn't aware of that, when was that distinction made?
Second, so are you saying the two require different testing methodologies to
determine DyR? That's the first you've said of THAT.
 
>> Furthermore, I agree that there needs to be some term to
>> distinguish between
>> a litho-like substance, and a continuous tone substance, where
>> the two could
>> conceivably have the same density range, while one is capable of far more
>> shades in between. But I wonder now if dynamic range IS that descriptor,
> 
> It is.
> 
>> when in fact you could measure a three toned medium and presume it to have
>> the same number of tones and the continuous tone item, based upon
>> that test
>> and calculation. So, at best dynamic range can provide a
>> predictor of what a
>> medium is capable of, but it assumes a complete and linear set of tones
>> between the ends, which may or may not be so; at worst it could
>> mislead one
>> wildly.
> 
> Not at all.  Dynamic range can be measured/calculated as precisely as you
> like.  It's like measuring the width of a 4 x 8 sheet of plywood.  You can
> just say it's 4' wide, or you can take a measurement every 1' or every 1" or
> over any distance you want...  I'm just showing a measurement technique
> here...there is nothing about dynamic range, per se, in this plywood
> example.

I follow, but we can't proceed until we clear up my questions above.
 
>> I need to explore this subject of linearity more fully as it pertains to
>> dynamic range. For instance, a print with an expansion of highlights and
>> contraction of shadows (= high contrast) will contain fewer tones than the
>> material is capable of, but your measure of dynamic range will
>> not indicate
>> that has occurred. So, is dynamic range a measure of the number
>> of tones an
>> item contains, or a measure of what the material is capable of if perfect
>> linearity is present?
> 
> It entirely depends on what you are measuring and how you measure it.

Understood.
 
>> This has me wonder why you assume a Piezo print to have greater dynamic
>> range than a silver print given these considerations:
> 
> Because I have printed the exact same negative on both, and the Piezo print
> had more tones visible to the eye.

Okay, but that is one small sample, and without knowing your workflows for
both processes there is no way for anyone else to interpret that. But I
fully understand that in your tests you did better with Piezo.
 
>> A) We know a well handled glossy fiber print is capable of a
>> greater density
>> range than Piezo.
> 
> Which is not really relevant to dynamic range...

Okay, lets explore this. If DyR is the density range/min discernable value,
the greater the density range, the more noise the system can handle and
still yield a high DyR. (if your min discernable value is 1, and your DnR is
5 then your DyR would be 5/1 =5, if your DnR is 4 and noise is the same 4/1
=4. Like the extent of my math? <G>) So if two systems have equal noise, the
one with the higher DnR will have the higher DnR, no? That is the relevance
I'm speaking to.
 
>> B) We know the source material of a silver print (negative) is capapble of
>> more tones (millions? billions?) than the source material for an inkjet
>> print (256? 65,000?).
> 
> I don't believe we know that.

Well it's not a given in every case, but we know MOST people out here are
Piezo printing from 8-bit, 256 shade files, and I trust you agree most
continuous tone negative films are EASILY *capable* of containing more tones
than that. Tell me you don't disagree with that!
 
>> C) Noise can go either way, depending on which papers are used in each
>> respective system, and other workflow considerations.
> 
> Absolutely.
> 
>> I'm trying to get my head around my sense that the approach you
>> advocate for
>> determining the dynamic range of a print is only marginally more
>> informative
>> than the approach to determining density range (which is fine, you didn't
>> invent the formula <G>);
> 
> But density range only tells you the relative distance between the
> endpoints, and as we have discussed in the past, can be represented with
> only two bits...and as the simple equation for density range shows, it
> merely is (dMax - dMin) and is a static measurement.

Understood.
 
>> that measuring the dynamic range of a print still
>> doesn't tell you how many tones it contains;
> 
> It certainly does, as defined by the equation for dynamic range ((dMax -
> dMin) / smallest discernable value).  Note, it has an additional variable in
> the equation beyond what density range does.  The dynamic range of a
> chemical print certainly can be done as a range.  You can add any complexity
> to it you want...but it doesn't change the basic concept.

Okay, how many tones are in a print with a DyR of 25db? 10, 100, 1000?

Is it possible to measure a DyR of a print that contains less than that many
tones?, Say a print with paper white with a 95% reflectance, a dmax of 2.2,
a min discernable value of .1, and 10 solid shades of gray within the DnR?

What would that print's DyR be compared to a print with all the same
features but with 100 shades within the same DnR and noise parameters?
 
>> that I see no logic that a
>> Piezo print is any more capable of greater dynamic range than
>> silver,
> 
> It doesn't have logic associated with it, it's simply observation.

Okay, I'll let it go at that. ;-)
 
>> I told you this would be noisy/rambling. ;-)
> 
> You had a respectable dynamic range here ;-)

Lovely, you are kind, thank you!
 
>> Your thoughts please.
> 
> Anything else?
> 
> Regards,
> 
> Austin
> 
> P.S.  Please keep it short if you could...

I tried...

And thanks again.

Todd

on 3/29/02 11:42 AM, Austin Franklin wrote:

> BTW, the source of the noise, though it may be interesting to know, isn't
> really relevant to what the dynamic range is.  It is THE print that we are
> trying to determine the dynamic range of, and it is what it is...no matter
> what caused it to be what it is.  So, if the source is the paper, lense,
> negative, developer...what ever...won't matter.

Right. I understand that the source of noise is irrelevant to the
calculation of dynamic range, but understanding the sources will help us
know where to look for noise and how to measure it.

Todd

> > I never said it was a "law of physics", but it IS a specific
> measurement,
> > and as I said, it IS defined by a fixed, non-negotiable (except
> different
> > systems have some different multipliers, but the basic equation, and
> > meaning, is the same) equation with defined relationships.
>
> It is only good if those relationship exist or exist to a
> meaningful degree
> in a system.

Yeah.  I don't know that you can quantify the dynamic range of a pizza.

> > What does that have to do with dynamic range?  I'm missing your point
> here.
> > For "largest", of course you need to measure the darkest black
> and whitest
> > white and "largest" is the difference of the two...which is the same as
> > density range...but to derive dynamic range from that, you
> divide that by
> > the "smallest discernable signal".
>
> In photography the dynamic range has been takes as the density range

Since when?  By people who don't understand what the real meaning of dynamic
range is, perhaps...

> which
> happens to be the reflectance ratio and fulfills the general definition of
> dynamic range as the ratio of the min and max signals for a photographic
> system.

It depends on what you mean by "min" and "max".

> This is not a sophisticated enough definition to define dynamic
> range for digital imaging where your equation is appropriate.

The equation is system independent.

> First the definition for dynamic range is not a constant and have
> different
> meaning for different fields.

It does not, except for using different multiplication factors for other
reasons, but it is ALWAYS based on (largest signal divided by smallest
discernable signal).  Always.

> I understand that you don't agree with this
> and see it as a universal concept. What I am saying here is let's
> apply your
> equation to a photographic print and see if the values help us understand
> what we see.

Well, I have and it does ;-)

> In order to do that we need to have a definition of
> what would
> be the noise in a photograph and a value for it.

I've given a method for doing so.

> > > > Well, then what do YOU believe is "dynamic" in dynamic
> range?  That is
> a
> > > > very important question.
> > >
> > > The ability to change tone.
> >
> > OK, then how does only using dMax and dMin give you anything dynamic?
>
> They tell you the size tonal space you can change tones in and
> they tell you
> the size of the largest change you can make.

That is density range, not dynamic range.  Why on earth does the equation
for dynamic range (the ONE AND ONLY one that anyone has presented, and given
a succinct reference for, namely the one I provided) clearly base dynamic
range on smallest discernable signal?

> And no I do not believe that
> there are a finite number of steps between them so I don't need an
> expression to calculate what isn't there.

Yes, I know.  This is like trying to get Newtonian physicists to understand
Einsteinen physics...it goes against everything they based their beliefs on.

> > > What tonal value do you print for your test patch and how do you know
> that
> > > the variation at this single tone will be representative of all
> > > the tones in
> > > the range?
> >
> > You can do various samples.
>
> Okay, what would be a reasonable number of samples and which
> tones should we
> look at first?

This is completely irrelevant to the overall discussion.  This is purely an
implementation issue.

> > OK, now I feel like we are getting somewhere ;-)

I take that back ;-(

> > I believe so, but I haven't thought about that at all.  I'm
> just trying to
> > lay down the foundation for that part of the discussion.
>
> Well since you have arrived at a possible concept for noise in a print,
> design an experiment to determine a value. I have a Spectrocam
> photospectrometer on order and I could try some readings across monotone
> patches. I think that probably the variation may be much less that it can
> measure but it can't hurt to check.

That sounds like a plan ;-)  What are the specs for the unit?

> I also hope to have a Howtek D4000
> scanner soon. Would that give higher resolution or be usable to
> attempt such
> a measurement?

No, I don't believe so.  I would just create a test pattern with darkest
black and a probably 4 or so different solid test patches in PS and print
that.  It's as "pure" as you're going to get!

> > Those definitions only work for density range, not dynamic
> range.  min and
> > max mean something different in the dynamic range equation.
>
> Now I am confused again. So what attribute of the print would you use to
> determine min and max if not the whitest white and blackest black

Good!  Confusion IS good...  Seriously.  OK, in the density range equation,
"max" IS simply the maximum density value you can measure by direct
measurement, and "min" IS simply the minimum density value you can measure
by direct measurement.  Simply put, measure the darkest patch, get your max
value, measure the paper, get your min value.

For dynamic range, take dMax (from your density range measurement) and
subtract your dMin (from your density range measurement), and that gives you
"max" for the dynamic range equation.  "min", basically, will be the
variance across the patch that I've suggested above.

> (snip)
> > >
> > > From my own experience with a MacBeth densitometer the ability to
> discern
> > > tonal differences declined steadily as you increased print density
> > > especially above 2 and stopped around 2.4. I was doing a lot
> of contact
> > > exposures with a Stouffer step wedge to measure the different
> brands of
> > > silver paper and the effects of different developers and toning
> processes.
> > > Why do you say 3?
> >
> > X-Rite 810 manual ;-)
>
> Can you quote what they say or is it too lengthy? I'm curious because it
> really is at odds with my own experience.

Be happy to.

Quoted from X-Rite 810 Operator's Manual, p. 51, section 6.0
SPECIFICATIONS/6.1 Instrument Specifications

REFLECTION SPECIFICATIONS

Measuring Range: 0-2.5D
Accuracy: +- .02D
Repeatability: +- .01D

TRANSMISSION SPECIFICATION

Measuring Range: 0-4.0D
Accuracy: +- .02D (0-3.50D)
          +-   1% (3.1D-3.5D)
          +-   3% (3.6D-4.0D)
Repeatability: +- .01D

Note, I was quoting the reflection specs, but note the transmission specs
show it consistent across the entire range, up to the max of 2.5D.

> Doesn't that suggest that the ability for the eye to discern tones is a
> function of the tonal density?

Of course it is...but I don't see how that applies.

> > > Well if you are going to do the division in your equation you
> > > need to choose
> > > a value. So if it varies what do you use? Yes but long ago
> and far away.
> >
> > You can integrate it very easily.  This is not a problem.  Your
> > concern/point is entirely valid, but it can easily be dealt with.
>
> What would the equation look like then?

Don't know, I'd have to work it out...but as I've said, this is a red
herring and completely irrelevant to the overall discussion.  It is merely
an implementation detail.

> > > Well give me a rough estimate of the number of tones you think
> > > there are in
> > > an inkjet and silver prints.
> >
> > It depends on the print, of course!  Inkjet, using the Cone
> system is over
> > 100.  Silver I have not "counted", but I can easily tell when one print
> has
> > "more" than another.  The same negative printed in a wet darkroom, that
> has
> > a high dynamic range (many tones ;-) shows more tones on the
> inkjet print
> > than the wet print.
>
> How who is being vague? <G>

Not me.  I am simply relating my direct experience.  It's not scientific,
but it IS my experience.

> Severe over exposure or over development would be similar to
> pushing an amp
> signal to the point of high distortion.
> >
> > > I sure do
> > > want to know the DR of my scanners though!
> >
> > Get a high range target and scan it ;-)
>
> What target do you use?

Q-60...but also check with Edmund's or Stoffers.  Your sensitometer might
come with one.  Some scanners do too...

Regards,

Austin

> First off I'd like to acknowledge you and Martin for the time you've put
> into this topic. You both have gone far beyond the call of duty. Thanks.

Ha!  He's as tenacious and as verbose as you are ;-)

> > No, only two bits are needed to represent the DENSITY range.
>
> I thought one bit could give you density range: maximum measurable (max
> black), minus min measurable (paper white). Two tones, one bit.

Here's the thought, call it as you may.  You need to represent two
thresholds, and ANY value in between.  That's three numbers you need to
represent.  You need two bits to do that.

> And that it
> was the addition of the min discernable value (which I'll shorthand to
> noise), which you sampled via deviation across the test patch, which gave
> you the third component (noise) of your dynamic range equation?

No.  Bits and dynamic range are different than representing density range
with N bits.  The number of bits IS the dynamic range (of that number of
bits), as you need the intermediate tones to ALL be represented.  You do not
for density range.

> I read you to mean that you get your dynamic range (I'll
> abbreviate that to
> DyR from here on in this post) by measuring the min and max (I'm short
> handing here, you know what I mean), subtract the min from the
> max (which is
> your density range (DnR), then dividing that sum by your noise, which you
> measured from your third tone. That gives you your DyR.

Yes, that is correct.

> So you've sampled three tones, referred to as: min, max, and test (or min
> discernable, or noise). Let's assume this print was comprised of
> ONLY these
> three tones. You could calculate a very high DyR from these three tones,
> implying the print has many intermediary tones, which it does not, no?

You need to separate the dynamic range of a particular image from the
dynamic range achievable from the "system".  They are different.

> Really? First off I wasn't aware of that, when was that distinction made?

Martin said that chemical prints don't have dynamic range since you can't
measure any discernment in tones, he claims they are infinitely continuous.

> Second, so are you saying the two require different testing
> methodologies to
> determine DyR? That's the first you've said of THAT.

A single test methodology can be arrived at.

> >> A) We know a well handled glossy fiber print is capable of a
> >> greater density
> >> range than Piezo.
> >
> > Which is not really relevant to dynamic range...
>
> Okay, lets explore this.

It IS relevant in that it bounds the "largest" signal, but you can have a
lower "largest" signal, and have a higher dynamic range.  That's why I said
it isn't relevant.  Relevant really isn't the right word perhaps.

> >> B) We know the source material of a silver print (negative) is
> capapble of
> >> more tones (millions? billions?) than the source material for an inkjet
> >> print (256? 65,000?).
> >
> > I don't believe we know that.
>
> Well it's not a given in every case, but we know MOST people out here are
> Piezo printing from 8-bit, 256 shade files, and I trust you agree most
> continuous tone negative films are EASILY *capable* of containing
> more tones
> than that. Tell me you don't disagree with that!

That's negative, not print.

> >> that measuring the dynamic range of a print still
> >> doesn't tell you how many tones it contains;
> >
> > It certainly does, as defined by the equation for dynamic range ((dMax -
> > dMin) / smallest discernable value).  Note, it has an
> additional variable in
> > the equation beyond what density range does.  The dynamic range of a
> > chemical print certainly can be done as a range.  You can add
> any complexity
> > to it you want...but it doesn't change the basic concept.
>
> Okay, how many tones are in a print with a DyR of 25db? 10, 100, 1000?

25 = 10log10((dMax - dMin) / sds)

Now, I claim that ((dMax - dMin) / sds) ARE the number of tones, so it
doesn't matter what those three values are ;-)

So, 25 = 10log10X, makes X 310.

> Is it possible to measure a DyR of a print that contains less
> than that many
> tones?, Say a print with paper white with a 95% reflectance, a
> dmax of 2.2,
> a min discernable value of .1, and 10 solid shades of gray within the DnR?

You can always measure the dynamic range of one print...or the system.
Different tests though.

> > P.S.  Please keep it short if you could...
>
> I tried...

Hum ;-)

Regards,

Austin

On 3/29/02 10:39 AM, "Austin Franklin" <darkroom@...> wrote:

>> 1.  I now think you do need both density range and dynamic range to
>> characterise the tones in a print.  Clearly the blackest black is
>> important
>> and the whitest white is important -- these are measured by their density
>> and the difference would be density range. Also the ability to
>> render smooth
>> transitions from one tone to another across the density range is important
>> and this would be measured by Dynamic Range.
> 
> Bingo!  And very eloquently worded.  I hope it didn't take too many licks to
> get there ;-)
> 
>> 4. The dynamic range of my prints (using the piezo driver v5, and epson
>> 1200, and MIS FS inks) has a maximum of:
>> 
>> -10log10((measured paper black density)-(measured paper white
>> density)/256))
>> 
>> Since the originating image only contains 256 shades of grey this caps the
>> upper limit (ie smallest value of) on the minimum perceptible
>> change.
> 
> You are correct, assuming your print HAS 256 tones though.  They dynamic
> range of the printing system would always be 256 tones, but the print does
> not have to contain 256 tonal steps, or 1/256th step per tonal variation.
> Printing litho film comes to mind.
> 
>> If my
>> eyes can only perceive 100 shades of grey then the dynamic range is lower.
> 
> I don't believe that lowers the dynamic range of the print...it is what it
> is.  The dynamic range YOU can perceive is lower, no doubt.
> 
When you say "it is what it is" do you mean that given a set of increasingly
accurate measuring devices there will be a point where the next most
accurate device will not yield a result different from its predecessor?

Also, how do you deal with the increasingly precise measurement when the
precision of the measurements starts to exceeds 2x (or is it 1/2) the
physical characteristics of that which you are measuring (ink droplets,
silver grains etc.

>> If my eyes can see 1000 shades of grey the dynamic range is still
>> capped by
>> the 256 shades that the print originated in.
> 
> Very true.
> 
>> To assess whether a silver print has more or less dynamic range you would
>> have to use a "system" that can accurately discriminate as many distinct
>> tones as our eyes and see if either print reaches its discrimination limit
>> before the limit of our eyes is reached.
> 
> Some prints are blatantly obvious though.  I don't know that you need to
> factor our eyes into this, as you really need to use a measurement technique
> that is more accurate than what it is you are trying to measure.  As you
> have said, our eyes can only distinguish 100 levels of gray (in any one
> lighting condition that is).
> 
> Regards,
> 
> Austin
> 
> 
> 
> Please visit the Group Homepage to check the Files, Bookmarks, Polls and other
> resources as they are often being updated. The page is at:
> 
> http://groups.yahoo.com/group/DigitalBlackandWhiteThePrint
> 
> Please follow these basic guidelines:
> - Include your full name with your message.
> - Include the address of your website, if you have one.
> - As threads develop, trim off excess portions of earlier messages to keep
> them short.
> - As the topic of a thread changes remember to change the subject header.
> - Good manners are required at all time. No personal attacks or "flames."
> - Complete your Yahoo profile.
> - Before posting a question, search the message archives and the various
> resources on the homepage.
> 
> 
> 
> 
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
> 
> 
>

Hi Kevin,

> > I don't believe that lowers the dynamic range of the print...it 
> is what it
> > is.  The dynamic range YOU can perceive is lower, no doubt.
> > 
> When you say "it is what it is" do you mean that given a set of 
> increasingly
> accurate measuring devices there will be a point where the next most
> accurate device will not yield a result different from its predecessor?

Yes, that's an interesting way to put it!

> Also, how do you deal with the increasingly precise measurement when the
> precision of the measurements starts to exceeds 2x (or is it 1/2) the
> physical characteristics of that which you are measuring (ink droplets,
> silver grains etc.

I don't quite get what you mean here...

Regards,

Austin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Friday, March 29, 2002 11:06 AM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


(snip)
> Yeah.  I don't know that you can quantify the dynamic range of a pizza.

It may have something to do with how fast I can grab a piece and how much
beer you got to go with it.

(snip)
> >
> > In photography the dynamic range has been takes as the density range
>
> Since when?  By people who don't understand what the real meaning of
dynamic
> range is, perhaps...

Since Ansel Adams and Minor White worked out the zone system and by
definition of the term dynamic range since the density difference gives you
the ratio of the minimum reflectance to the maximum reflectance.

From NASA website:

\ufffdThe dynamic range of an image is the ratio between the maximum and minimum
brightness levels in the image.\ufffd

http://nasatech.com/Briefs/Aug98/NPO20254.html



From the Cornell University Library:

\ufffdDYNAMIC RANGE is the range of tonal difference between the lightest light
and darkest dark of an image.\ufffd

http://www.library.cornell.edu/preservation/tutorial/intro/intro-05.html



From Digital Photography Review:

\ufffdDynamic range is the ratio between the brightest and darkest recordable
parts of an image or scene.\ufffd

http://www.dpreview.com/learn/Glossary/Digital_Imaging/Dynamic_Range_01.htm


(snip)>
> It depends on what you mean by "min" and "max".

See above.
>
(snip)
>
> It does not, except for using different multiplication factors for other
> reasons, but it is ALWAYS based on (largest signal divided by smallest
> discernable signal).  Always.

But the determination of what is the meaningful largest and smallest signals
is dependent on the system.
>
> > I understand that you don't agree with this
> > and see it as a universal concept. What I am saying here is let's
> > apply your
> > equation to a photographic print and see if the values help us
understand
> > what we see.
>
> Well, I have and it does ;-)

You have not been able to supply a value for the noise in your equation yet
so how did you manage that.
>
(snip)

> > > OK, then how does only using dMax and dMin give you anything dynamic?
> >
> > They tell you the size tonal space you can change tones in and
> > they tell you
> > the size of the largest change you can make.
>
> That is density range, not dynamic range.  Why on earth does the equation
> for dynamic range (the ONE AND ONLY one that anyone has presented, and
given
> a succinct reference for, namely the one I provided) clearly base dynamic
> range on smallest discernable signal?

Austin, that is simply not true. I have provided several of descriptions and
definitions of dynamic range from several sources. I have taken those an
written equations based upon the definitions. You simply say they are wrong
because they do not match your equation.
>
(snip)
>
> Yes, I know.  This is like trying to get Newtonian physicists to
understand
> Einsteinen physics...it goes against everything they based their beliefs
on.

Or Einsteinen physicists to under stand Newtonian physics. This is actually
and ongoing debate but that is another story. These things cut both ways.
>
(snip)
> > Okay, what would be a reasonable number of samples and which
> > tones should we
> > look at first?
>
> This is completely irrelevant to the overall discussion.  This is purely
an
> implementation issue.

Austin, I offer you a chance to look at what might be the noise value of a
print to use in your equation and you tell me that is irrelevant to the
overall discussion? If you cannot give this term a value than how can you
determine if your dynamic range tells you something significant about a
print?
>
(snip)
>
> >
> > Well since you have arrived at a possible concept for noise in a print,
> > design an experiment to determine a value. I have a Spectrocam
> > photospectrometer on order and I could try some readings across monotone
> > patches. I think that probably the variation may be much less that it
can
> > measure but it can't hurt to check.
>
> That sounds like a plan ;-)  What are the specs for the unit?

http://www.spectrostar.com/indexmsie.htm
>
(snip)
>
> > Now I am confused again. So what attribute of the print would you use to
> > determine min and max if not the whitest white and blackest black
>
> Good!  Confusion IS good...  Seriously.  OK, in the density range
equation,
> "max" IS simply the maximum density value you can measure by direct
> measurement, and "min" IS simply the minimum density value you can measure
> by direct measurement.  Simply put, measure the darkest patch, get your
max
> value, measure the paper, get your min value.

Austin, I have been saying that for post after post. What else could the
Dmin and Dmax of a print would be?
>
> For dynamic range, take dMax (from your density range measurement) and
> subtract your dMin (from your density range measurement), and that gives
you
> "max" for the dynamic range equation.  "min", basically, will be the
> variance across the patch that I've suggested above.

Okay, looks like the density range (or the log expression of ratio of the
reflectance) divided by the noise.
>
> > (snip)
> > > > Why do you say 3?
> > >
> > > X-Rite 810 manual ;-)
> >
> > Can you quote what they say or is it too lengthy? I'm curious because it
> > really is at odds with my own experience.
>
> Be happy to.
>
> Quoted from X-Rite 810 Operator's Manual, p. 51, section 6.0
> SPECIFICATIONS/6.1 Instrument Specifications
>
> REFLECTION SPECIFICATIONS
>
> Measuring Range: 0-2.5D
> Accuracy: +- .02D
> Repeatability: +- .01D
>
> TRANSMISSION SPECIFICATION
>
> Measuring Range: 0-4.0D
> Accuracy: +- .02D (0-3.50D)
>           +-   1% (3.1D-3.5D)
>           +-   3% (3.6D-4.0D)
> Repeatability: +- .01D
>
> Note, I was quoting the reflection specs, but note the transmission specs
> show it consistent across the entire range, up to the max of 2.5D.

Okay that is the X-Rite 810's uncertainly when it measures density but it is
much better at this than the human eye. Like many instruments it is only
linear over a specific range. This may be the case with our eyes as well but
believe me the human eye has a tough time discerning tonal differences above
2.4.
>
> > Doesn't that suggest that the ability for the eye to discern tones is a
> > function of the tonal density?
>
> Of course it is...but I don't see how that applies.

Just that it varies from light to dark so that you could not take a simple
seperation between discernable tones but would need to mathematically
account for this variation.
>
(snip)

> > What would the equation look like then?
>
> Don't know, I'd have to work it out...but as I've said, this is a red
> herring and completely irrelevant to the overall discussion.  It is merely
> an implementation detail.

If it can't be implemented what good is it? If you don't have the values you
need to apply your dynamic range equation, what use is it?
>
(snip)
> > > It depends on the print, of course!  Inkjet, using the Cone
> > system is over
> > > 100.  Silver I have not "counted", but I can easily tell when one
print
> > has
> > > "more" than another.  The same negative printed in a wet darkroom,
that
> > has
> > > a high dynamic range (many tones ;-) shows more tones on the
> > inkjet print
> > > than the wet print.
> >
> > How who is being vague? <G>
>
> Not me.  I am simply relating my direct experience.  It's not scientific,
> but it IS my experience.

And this had nothing to do with the operator but only the materials. <G>
>
(snip)
> > > Get a high range target and scan it ;-)
> >
> > What target do you use?
>
> Q-60...but also check with Edmund's or Stoffers.  Your sensitometer might
> come with one.  Some scanners do too...

Thanks,
Martin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Friday, March 29, 2002 11:22 AM
Subject: RE: [Digital BW] Dynamic Range: For Austin


(snip)
>
> Martin said that chemical prints don't have dynamic range since you can't
> measure any discernment in tones, he claims they are infinitely
continuous.

No I said the dynamic range of the silver print is Dmax minus Dmin which is
the ratio of the the minimum and maximum reflectance. I said that a silver
print may not have a DyR as Austin defines it because the noise in his
equation is either zero, approaches zero or is so low as to be reasonably
neglected.

Martin
(snip)

> > Yeah.  I don't know that you can quantify the dynamic range of a pizza.

You put the white eyedropper on the lightest part of the cheese, and the
black 
eyedropper in the shadow of a piece of pepperoni. Or hamburger. Or
whatever is the darkest part of the pieza.

:)

>

--- In DigitalBlackandWhiteThePrint@y..., "Martin Wesley" 
<mwesley250@e...> wrote:
> 
> ----- Original Message -----
> From: "Austin Franklin" <darkroom@i...>
> To: <DigitalBlackandWhiteThePrint@y...>
> Sent: Friday, March 29, 2002 11:22 AM
> Subject: RE: [Digital BW] Dynamic Range: For Austin
> 
> 
> (snip)
> >
> > Martin said that chemical prints don't have dynamic range since 
you can't
> > measure any discernment in tones, he claims they are infinitely
> continuous.
> 
> No I said the dynamic range of the silver print is Dmax minus Dmin 
which is
> the ratio of the the minimum and maximum reflectance. I said that a 
silver
> print may not have a DyR as Austin defines it because the noise in 
his
> equation is either zero, approaches zero or is so low as to be 
reasonably
> neglected.
> 
> Martin
> (snip)
I think a silver print does have a DyR.  I'm not sure what noise 
would be in a silver print but I think the minimum perceptible change 
could be determined -- with some difficulty.  Perhaps it could be 
done by precisely exposing  and developing greyscale step wedges with 
veryfine increments in the grey.  If they were in say 1cm squares 
then the point at which you could no longer tell the difference 
between the measured densities of the squares would be the smallest 
discernable change.  No? 

Kevin Gulstene

--- In DigitalBlackandWhiteThePrint@y..., "Austin Franklin" 
<darkroom@i...> wrote:
> Hi Kevin,
> 
> > > I don't believe that lowers the dynamic range of the print...it 
> > is what it
> > > is.  The dynamic range YOU can perceive is lower, no doubt.
> > > 
> > When you say "it is what it is" do you mean that given a set of 
> > increasingly
> > accurate measuring devices there will be a point where the next 
most
> > accurate device will not yield a result different from its 
predecessor?
> 
> Yes, that's an interesting way to put it!
> 
> > Also, how do you deal with the increasingly precise measurement 
when the
> > precision of the measurements starts to exceeds 2x (or is it 1/2) 
the
> > physical characteristics of that which you are measuring (ink 
droplets,
> > silver grains etc.
> 
> I don't quite get what you mean here...
> 
I'm not surprised.  For some reason I started getting confused 
between precision of density measurement and the size of the area 
being sampled.  Erroneously thinking that the measurment had to be 
smaller to be better.  Hence the reference to physical size.

> Regards,
> 
> Austin

The thread has been interesting and after thinking a lot and
reading about it I've got a few comments.

1) Dynamic Range Issue

At first I was very skeptical about Austin's views about
Dynamic Range.  I've got a couple of degrees in EE from way back
but its all pretty rusty since I haven't used it in a long time.
So I figured it deserved some brushing up on the subject.
Naturally, I decide to look at the book by Higgins, that Austin 
referenced for his formula.  As expected Austin's formula is
straight out of the book.

Dynamic Range (dB) = log10 (largest signal/smallest discernable signal)

Now the big question is "What's the signal?".  All my own
recollections and everything I saw in the book has an audio/video
type flavor.  I.e the diagrams show nice sinusoidal signals in
the time domain.  Signals are measured by amplitude of the sine
waves.  All this is well and good, but of course for our purposes
we want to map all this into photographic prints in the spacial
domain.

Austin, you've made a clear choice in this mapping from the time
domain to the space domain.  For you "signal" is difference
between to tonal densities, i.e:
    largest signal = dMax - dMin
    smallest discernable signal = smallest density difference
or maybe I can paraphrase "signal is basically contrast".
(I hope I'm not putting words in your mouth!)

As far as I can tell your derivations from this and all the
discussions are consistant and mathematically sound.  But I've
been a little unsure of picking this mapping rather than 
something else.

Obviously, what we are trying to do is make a mathematical
model of a human perception.  In audio, the main thing is
loudness: with dynamic range we're characterizing how much
louder the loud parts are from the soft parts.  So the ratio
inside the parentheses is (loudest/softest). Makes sense.
For pictures, what's the basic perception?  It seems you've picked
contrast --  so the ratio is (highest contrast/lowest contrast).
Why not the more basic perception: darkness versus lightness?
The ratio would be (darkest/lightest). I guess you can see
where this is going->> it reduces the whole thing to
Density Range.

DynRange = log10 (darkest/lightest)
         = log10 (darkest) - log10 (lightest)
         = dMax - dMin
         = Density Range

I doubt that you're ready to switch definitions but intuitively
using "dynamic range" to means "number of tones" rather
than some "range" of values seems weird.  And by the prevalence
of the "dynamic range" threads that go on regularly many
others don't feel comfortable with your definition.
I'm not up to trying to buck a whole industry's definition so
are there references that show this as wide spread?  I'd
love to read some.  


2) Number of Print Tones

There was some discussion about number of grays on a monitor
screen.  For current monitors they all have a maximum of 256
grays.  Every RGB monitor I know of takes 8 bits for each
color (Red,Green,Blue).  Grayscales make R=G=B i.e. the same
8 bits are sent to each color gun.  There's only 256 possibilities.

To get an interesting feel for how many grays make up a 
picture, try this.  Take a 8-bit grayscale picture small
enough to fit on the screen at 100%. Then posterize it with
128,64,32,16 levels.  Try a couple different pictures.

For both darkroom and inkjet prints, gray tones are really
just a human perception.  There is just white paper with
lots of small black specks of silver or carbon.  The
perception of grays is based on optically (eyeball) and
the brain integrating the blacks and whites over a finite
area. So its based on things other than the print itself.

I would guess that less the brain feels a need to integrate
over a larger area the sharper the image seems. But the
larger the area the more opportunity for perceiving more
different gray tones.

Here's an interesting thought experiment, Say we a printing
out gray swatches from an inkjet printer.  Each swatch
is a small square of each grayscale value.  Assuming we're
talking 8 bit files and printer drivers, we can have only
256 possible gray swatches.  If its all perfectly 
calibrated ideally we ought to be able to distinguish
each swatch.  Now imagine a swatch which is a checkboard
with two different adjacent gray values, for instance
137 and 138 mixed together.  What does it look like?  Is
it appear as a checkerboard of two different grays or
does it merge into a new and different 137.5 gray swatch?
I think the answer is "it depends" ... i.e. how close you
get to it, got your bifocals on? etc.

The point is "number of tones" is kind of nebulous.  This
also adds to my dissatisfication with "dynamic range"
meaning "number of tones".


Hopefully someone will find this useful.

Roy

-- 
Roy Harrington
roy@...
Black & White Photography Gallery
http://www.harrington.com

Hiiiii again Martin ;-)

> Since Ansel Adams and Minor White worked out the zone system and by
> definition of the term dynamic range

Do you have a reference for EXACTLY what you claim?  I AM a Zone boy (as in
I do all Zone work, with development compensation etc.), and I see that the
Zone system has some basis in dynamic range, simply by defining 10 steps,
that's giving it SOME dynamic range...

> From NASA website:
> From the Cornell University Library:
> From Digital Photography Review:

And you conveniently left out the ones that were "better" ;-)  Why do you
keep going back to taking some one's interpretation of a REAL EQUATION and
not just use the real equation?  Talk about adding noise to the system!

> > It does not, except for using different multiplication factors for other
> > reasons, but it is ALWAYS based on (largest signal divided by smallest
> > discernable signal).  Always.
>
> But the determination of what is the meaningful largest and
> smallest signals
> is dependent on the system.

It is defined by the equation...we are going around in circles here...

> > > I understand that you don't agree with this
> > > and see it as a universal concept. What I am saying here is let's
> > > apply your
> > > equation to a photographic print and see if the values help us
> understand
> > > what we see.
> >
> > Well, I have and it does ;-)
>
> You have not been able to supply a value for the noise in your
> equation yet
> so how did you manage that.

Er, yes I did.  I gave you an example using 0.01D as the "noise".

> > > > OK, then how does only using dMax and dMin give you
> anything dynamic?
> > >
> > > They tell you the size tonal space you can change tones in and
> > > they tell you
> > > the size of the largest change you can make.
> >
> > That is density range, not dynamic range.  Why on earth does
> the equation
> > for dynamic range (the ONE AND ONLY one that anyone has presented, and
> given
> > a succinct reference for, namely the one I provided) clearly
> base dynamic
> > range on smallest discernable signal?
>
> Austin, that is simply not true. I have provided several of
> descriptions and
> definitions of dynamic range from several sources. I have taken those an
> written equations based upon the definitions. You simply say they
> are wrong
> because they do not match your equation.

You can NOT take someone's interpretation of an equation and just write the
equation from that!  My God.  As I've said a dozen times, and you, for some
reason won't get it, the dynamic range equation IS the equation for dynamic
range, period.  No matter how many times some pedestrian has tried to put
verbosity to it to try to describe it TO LAY PEOPLE using ambiguous terms.

> (snip)
> > > Okay, what would be a reasonable number of samples and which
> > > tones should we
> > > look at first?
> >
> > This is completely irrelevant to the overall discussion.  This is purely
> an
> > implementation issue.
>
> Austin, I offer you a chance to look at what might be the noise value of a
> print to use in your equation and you tell me that is irrelevant to the
> overall discussion? If you cannot give this term a value than how can you
> determine if your dynamic range tells you something significant about a
> print?

The point is there IS noise, it exists everywhere.  We do not have to
quantify it to know it exists.  It is also not a far stretch to believe it
is significant, knowing the difference in reflectance that paper has alone!

> > > Now I am confused again. So what attribute of the print would
> you use to
> > > determine min and max if not the whitest white and blackest black
> >
> > Good!  Confusion IS good...  Seriously.  OK, in the density range
> equation,
> > "max" IS simply the maximum density value you can measure by direct
> > measurement, and "min" IS simply the minimum density value you
> can measure
> > by direct measurement.  Simply put, measure the darkest patch, get your
> max
> > value, measure the paper, get your min value.
>
> Austin, I have been saying that for post after post. What else could the
> Dmin and Dmax of a print would be?

We are not in disagreement on how to derive the density range...and that
wasn't the important part of what I was saying, the next paragraph is.

> >
> > For dynamic range, take dMax (from your density range measurement) and
> > subtract your dMin (from your density range measurement), and that gives
> you
> > "max" for the dynamic range equation.  "min", basically, will be the
> > variance across the patch that I've suggested above.
>
> Okay, looks like the density range (or the log expression of ratio of the
> reflectance) divided by the noise.

YES.  Grrr.  As you said, and I quote: "I have been saying that for post
after post." ;-)

> Okay that is the X-Rite 810's uncertainly when it measures
> density but it is
> much better at this than the human eye. Like many instruments it is only
> linear over a specific range. This may be the case with our eyes
> as well but
> believe me the human eye has a tough time discerning tonal
> differences above
> 2.4.

I agree completely.

> > > Doesn't that suggest that the ability for the eye to discern
> tones is a
> > > function of the tonal density?
> >
> > Of course it is...but I don't see how that applies.
>
> Just that it varies from light to dark so that you could not take a simple
> seperation between discernable tones but would need to mathematically
> account for this variation.

Sure, but again, that's just an implementation detail ;-)

> > > What would the equation look like then?
> >
> > Don't know, I'd have to work it out...but as I've said, this is a red
> > herring and completely irrelevant to the overall discussion.
> It is merely
> > an implementation detail.
>
> If it can't be implemented what good is it?

Who said it can't be implemented?  Not I!

> If you don't have the
> values you
> need to apply your dynamic range equation, what use is it?

Having a "correct" value or not doesn't mean it doesn't exist!

> And this had nothing to do with the operator but only the materials. <G>

I was a VERY good B&W printer in the darkroom.  If it did have anything to
do with operator, it was negligible.  I was a real purist, with very good
equipment (still have it BTW, need to sell it ;-) and knew how to use it, as
I do my scanner.

Do you ever sleep?

Austin

Hi Roy,

> Now the big question is "What's the signal?".

Very simply, for a photographic image, it is the density value.

> All my own
> recollections and everything I saw in the book has an audio/video
> type flavor.  I.e the diagrams show nice sinusoidal signals in
> the time domain.

But time has never been a part of the dynamic range equation at all, even in
the case of audio...it is purely voltage.

> Austin, you've made a clear choice in this mapping from the time
> domain to the space domain.

Not at all...see above.  Neither time or space have anything to do with
this.  It is purely a (any in fact) measurable quality that is required,
whether voltage or density or what ever.  It is what ever property YOU want
to figure out the dynamic range of.

> For you "signal" is difference
> between to tonal densities, i.e:
>     largest signal = dMax - dMin
>     smallest discernable signal = smallest density difference
> or maybe I can paraphrase "signal is basically contrast".
> (I hope I'm not putting words in your mouth!)

Sure, I'll fly with that.

> As far as I can tell your derivations from this and all the
> discussions are consistant and mathematically sound.  But I've
> been a little unsure of picking this mapping rather than
> something else.

I disagree with your use of the term "mapping".  Again, you pick the
property you want to figure out the dynamic range of...note the dynamic
range equation does NOT use any units for largest/smallest...they simply
have to be on the same scale, of the same unit.

> For pictures, what's the basic perception?  It seems you've picked
> contrast --  so the ratio is (highest contrast/lowest contrast).

Yes, since contrast (as in delta change) is all that distinguishes tones
from one another.

> Why not the more basic perception: darkness versus lightness?

What, exactly, do you mean by "darkness" and "lightness"?

> The ratio would be (darkest/lightest). I guess you can see
> where this is going->> it reduces the whole thing to
> Density Range.
>
> DynRange = log10 (darkest/lightest)
>          = log10 (darkest) - log10 (lightest)
>          = dMax - dMin
>          = Density Range

But you have now removed, arbitrarily from what I can tell, the smallest
discernable signal from the equation, that above, you agreed was part of the
equation.  Why?  This seems like one of those Gary Larson "and then magic
happened" on the chalk board...

> I doubt that you're ready to switch definitions but intuitively
> using "dynamic range" to means "number of tones" rather
> than some "range" of values seems weird.

Not at all, my guess is it really means dynamic OVER THE particular
range...but I do not know the history of the term.  It isn't really A range,
but it describes a property OF the given range.  Density range isn't a range
either, but it does describe the width OF the range of density.

> And by the prevalence
> of the "dynamic range" threads that go on regularly many
> others don't feel comfortable with your definition.

Yeah, but I didn't make up the definition, as you well know!  Remember, this
is a very technical issue, being described to and by and used by a lot of
lay people.  There are a LOT of misperceptions/understandings when this type
of thing happens...especially these days with the proliferation of the
Internet.  You can find "compelling" (as in decent presentations, that look
well thought out) sources for both sides of near any argument!

> Here's an interesting thought experiment, Say we a printing
> out gray swatches from an inkjet printer.  Each swatch
> is a small square of each grayscale value.  Assuming we're
> talking 8 bit files and printer drivers, we can have only
> 256 possible gray swatches.  If its all perfectly
> calibrated ideally we ought to be able to distinguish
> each swatch.  Now imagine a swatch which is a checkboard
> with two different adjacent gray values, for instance
> 137 and 138 mixed together.  What does it look like?  Is
> it appear as a checkerboard of two different grays or
> does it merge into a new and different 137.5 gray swatch?
> I think the answer is "it depends" ... i.e. how close you
> get to it, got your bifocals on? etc.

It does depend.  More so on lighting than anything else.  We actually can
distinguish a lot more than 100 gray tones, but the 100 number is in "a"
given light...but vary that light, and you increase the number of tones you
can see...and your standard photographic print may not have tones that
extend into that range, since most people don't view a photo in the dark or
extreme light, so they weren't designed TO be seen in those conditions.

> The point is "number of tones" is kind of nebulous.  This
> also adds to my dissatisfication with "dynamic range"
> meaning "number of tones".

Very interesting point, but I have an answer ;-)  You are right, in your
example, you will be able to shift the tones all by .5 of a tone, and get a
whole new set of shifted "discernable" tones...but you can STILL only
discern the same number, as well as have the same separation!  This does
fail at some point, when your number of tones is high enough (where adjacent
tones aren't discernable).  Does that make sense?

> Hopefully someone will find this useful.

Yes.

Regards,

Austin

No offense Austin, but you really snipped out all my relevant points and
didn't answer much concretely.

I guess you're getting tired? ;-)

My main point was that for a formula that you suggests indicates the number
of tones present in a print, based on sampling as few as three tones, may
not yield a different number for each of two prints, even though one print
contains three tones and one contains 1000.

And I never spoke to a system in my post, precisely because I'm probing to
see if DyR holds up as useful (in the way you suggest = describing the
number of tones present) specifically for an existing print. I have no doubt
that DyR is useful for describing some aspects of a system's potential.

You also did not speak to my premise that I think that a calculation of
tones from DyR is only possible if linearity is assumed, which is not
something I would take for granted in a silver print.

Todd

>> So you've sampled three tones, referred to as: min, max, and test (or min
>> discernable, or noise). Let's assume this print was comprised of
>> ONLY these
>> three tones. You could calculate a very high DyR from these three tones,
>> implying the print has many intermediary tones, which it does not, no?
> 
> You need to separate the dynamic range of a particular image from the
> dynamic range achievable from the "system".  They are different.
> 
>> Really? First off I wasn't aware of that, when was that distinction made?
> 
> Martin said that chemical prints don't have dynamic range since you can't
> measure any discernment in tones, he claims they are infinitely continuous.
> 
>> Second, so are you saying the two require different testing
>> methodologies to
>> determine DyR? That's the first you've said of THAT.
> 
> A single test methodology can be arrived at.
> 
>>>> A) We know a well handled glossy fiber print is capable of a
>>>> greater density
>>>> range than Piezo.
>>> 
>>> Which is not really relevant to dynamic range...
>> 
>> Okay, lets explore this.
> 
> It IS relevant in that it bounds the "largest" signal, but you can have a
> lower "largest" signal, and have a higher dynamic range.  That's why I said
> it isn't relevant.  Relevant really isn't the right word perhaps.
> 
>>>> B) We know the source material of a silver print (negative) is
>> capapble of
>>>> more tones (millions? billions?) than the source material for an inkjet
>>>> print (256? 65,000?).
>>> 
>>> I don't believe we know that.
>> 
>> Well it's not a given in every case, but we know MOST people out here are
>> Piezo printing from 8-bit, 256 shade files, and I trust you agree most
>> continuous tone negative films are EASILY *capable* of containing
>> more tones
>> than that. Tell me you don't disagree with that!
> 
> That's negative, not print.
> 
>>>> that measuring the dynamic range of a print still
>>>> doesn't tell you how many tones it contains;
>>> 
>>> It certainly does, as defined by the equation for dynamic range ((dMax -
>>> dMin) / smallest discernable value).  Note, it has an
>> additional variable in
>>> the equation beyond what density range does.  The dynamic range of a
>>> chemical print certainly can be done as a range.  You can add
>> any complexity
>>> to it you want...but it doesn't change the basic concept.
>> 
>> Okay, how many tones are in a print with a DyR of 25db? 10, 100, 1000?
> 
> 25 = 10log10((dMax - dMin) / sds)
> 
> Now, I claim that ((dMax - dMin) / sds) ARE the number of tones, so it
> doesn't matter what those three values are ;-)
> 
> So, 25 = 10log10X, makes X 310.
> 
>> Is it possible to measure a DyR of a print that contains less
>> than that many
>> tones?, Say a print with paper white with a 95% reflectance, a
>> dmax of 2.2,
>> a min discernable value of .1, and 10 solid shades of gray within the DnR?
> 
> You can always measure the dynamic range of one print...or the system.
> Different tests though.
> 
>>> P.S.  Please keep it short if you could...
>> 
>> I tried...
> 
> Hum ;-)
> 
> Regards,
> 
> Austin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Friday, March 29, 2002 7:16 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


> Hiiiii again Martin ;-)
>
> > Since Ansel Adams and Minor White worked out the zone system and by
> > definition of the term dynamic range
>
> Do you have a reference for EXACTLY what you claim?  I AM a Zone boy (as
in
> I do all Zone work, with development compensation etc.), and I see that
the
> Zone system has some basis in dynamic range, simply by defining 10 steps,
> that's giving it SOME dynamic range...

I think Mike already posted to that.
>
> > From NASA website:
> > From the Cornell University Library:
> > From Digital Photography Review:
>
> And you conveniently left out the ones that were "better" ;-)  Why do you
> keep going back to taking some one's interpretation of a REAL EQUATION and
> not just use the real equation?  Talk about adding noise to the system!

Okay I will add:

From X-rite\ufffds \ufffdThe Color Guide and Glossary\ufffd

\ufffdDynamic Range: An instrument\ufffds range of measurable values, from the lowest
amount it can detect to the highest amount it can handle.\ufffd


and:

From CCD Direct:

\ufffdThe dynamic range is often represented as a log ratio of well depth to the
readout noise in decibels. For example, a system with a well depth of 45,000
electrons and a readout noise of 15 electrons would have a dynamic range =
20 log (45,000/15), or 69dB.\ufffd

http://www.ccddirect.com/online-store/scstore/dynamic.html



You never did say what you thought of this last one which seems to be in
your field.

>
(snip)

>
> It is defined by the equation...we are going around in circles here...



Yep. I am rather dizzy.
>
> >
> > You have not been able to supply a value for the noise in your
> > equation yet
> > so how did you manage that.
>
> Er, yes I did.  I gave you an example using 0.01D as the "noise".



Hmmm. Somehow an example just doesn't satisfy like an experimental value.
What if I say for example that the noise in a silver print is
0.00000000000000001 dB can I use that? <G>
>
(snip)

>
> You can NOT take someone's interpretation of an equation and just write
the
> equation from that!  My God.  As I've said a dozen times, and you, for
some
> reason won't get it, the dynamic range equation IS the equation for
dynamic
> range, period.  No matter how many times some pedestrian has tried to put
> verbosity to it to try to describe it TO LAY PEOPLE using ambiguous terms.



You say they are lay people or it is a simplified explaination and I don't
agree. I think we are not going anywhere in this direction.
>
> > (snip)
> > Austin, I offer you a chance to look at what might be the noise value of
a
> > print to use in your equation and you tell me that is irrelevant to the
> > overall discussion? If you cannot give this term a value than how can
you
> > determine if your dynamic range tells you something significant about a
> > print?
>
> The point is there IS noise, it exists everywhere.  We do not have to
> quantify it to know it exists.  It is also not a far stretch to believe it
> is significant, knowing the difference in reflectance that paper has
alone!



I have no reason to believe that noise exists everywhere and I can't take
that as a given. The noise you described earlier would be variations within
a given paper base rather than the variation from brand-to-brand and may be
so small as to push the results of your equation towards infinity.
>
(snip)

>
> We are not in disagreement on how to derive the density range...and that
> wasn't the important part of what I was saying, the next paragraph is.
>
> > >
> > > For dynamic range, take dMax (from your density range measurement) and
> > > subtract your dMin (from your density range measurement), and that
gives
> > you
> > > "max" for the dynamic range equation.  "min", basically, will be the
> > > variance across the patch that I've suggested above.
> >
> > Okay, looks like the density range (or the log expression of ratio of
the
> > reflectance) divided by the noise.
>
> YES.  Grrr.  As you said, and I quote: "I have been saying that for post
> after post." ;-)
>
(snip)

> >
> > Just that it varies from light to dark so that you could not take a
simple
> > seperation between discernable tones but would need to mathematically
> > account for this variation.
>
> Sure, but again, that's just an implementation detail ;-)



Well I am a very practical engineer. Chemical process and project
engineering remember, so I want to know all those details.
>
(snip)

>
> Who said it can't be implemented?  Not I!



True but you need a meaningful value for the noise to do that.
>
> > If you don't have the
> > values you
> > need to apply your dynamic range equation, what use is it?
>
> Having a "correct" value or not doesn't mean it doesn't exist!
>
No but lack of a correct value makes it of little practical use.



> > And this had nothing to do with the operator but only the materials. <G>
>
> I was a VERY good B&W printer in the darkroom.  If it did have anything to
> do with operator, it was negligible.  I was a real purist, with very good
> equipment (still have it BTW, need to sell it ;-) and knew how to use it,
as
> I do my scanner.



I did not mean to impugn your printing skills. I am proud of my own
abilities but still manage to have some real disasters.
>
> Do you ever sleep?



No. That would be terrible ineffiecient and waste of valuable time.  <G>



I'm just on the west coast and a bit of an insomniac, and can keep posting
past your bed time but you leave me a whole stack to look at with my morning
coffee.



Martin

----- Original Message -----

From: "Austin Franklin" <darkroom@...>
To: <DigitalBlackandWhiteThePrint@yahoogroups.com>
Sent: Friday, March 29, 2002 7:44 PM
Subject: RE: [Digital BW] Dynamic Range Definitions and Print Tones


(snip)
>
> Very interesting point, but I have an answer ;-)  You are right, in your
> example, you will be able to shift the tones all by .5 of a tone, and get
a
> whole new set of shifted "discernable" tones...but you can STILL only
> discern the same number, as well as have the same separation!  This does
> fail at some point, when your number of tones is high enough (where
adjacent
> tones aren't discernable).

As in continuous tone? <G>

Martin

> Here's an interesting thought experiment, Say we a printing
> out gray swatches from an inkjet printer.  Each swatch
> is a small square of each grayscale value.  Assuming we're
> talking 8 bit files and printer drivers, we can have only
> 256 possible gray swatches.  If its all perfectly
> calibrated ideally we ought to be able to distinguish
> each swatch.  Now imagine a swatch which is a checkboard
> with two different adjacent gray values, for instance
> 137 and 138 mixed together.  What does it look like?  Is
> it appear as a checkerboard of two different grays or
> does it merge into a new and different 137.5 gray swatch?
> I think the answer is "it depends" ... i.e. how close you
> get to it, got your bifocals on? etc.

It's a good demonstration of perception. For those to lazy to make your own
test targets, go into the Files section of this group and look at Tylers
Zees. It's a series of boxes with Zees in each. It represents 100 tones. I
originally thought that the Zees has a gradient through them, but in fact
they are a solid tone, but the box which houses them has a gradient through
it (5%). So the box, which is a gradient looks solid, unless you see it's
tone relative to the solid tone of the Z.

It speaks to the methodology of a teest which determines our ability to
distinguish between tones. If each tone is compared to the tone directly
next to it and they quite close in density it may be very difficult to
distinguish between the two, but if each is compared to a tone of a
significantly large enough difference you will. That is what gives the
illusion of the Z having a gradient.

Thus, in Tyler's target I dare say it's relatively easy to discern 100 tones
(due to the nonlinearity of our monitors gamma it's easier to see in print
than on screen), but if you took a linear gradient and posterized it into
100 steps I believe it would be very difficult to distinguish between many
of the tones.

Todd

Hi Austin,


--- In DigitalBlackandWhiteThePrint@y..., "Austin Franklin" <darkroom@i...> wrote:
> Hi Roy,
> 
> > Now the big question is "What's the signal?".
> 
> Very simply, for a photographic image, it is the density value.

This is precisely what I'm trying to say.   Let's take the original definition
equation and substitute "density value" for "signal".  We have:

Dynamic Range = log10 (largest signal/smallest discernable signal)
            =  log10 (largest density value/smallest discernable density value)

Isn't that precisely the same as: dMax - dMin  as I derived last post?

> 
> > All my own
> > recollections and everything I saw in the book has an audio/video
> > type flavor.  I.e the diagrams show nice sinusoidal signals in
> > the time domain.
> 
> But time has never been a part of the dynamic range equation at all, even in
> the case of audio...it is purely voltage.
> 
> > Austin, you've made a clear choice in this mapping from the time
> > domain to the space domain.
> 
> Not at all...see above.  Neither time or space have anything to do with
> this.  It is purely a (any in fact) measurable quality that is required,
> whether voltage or density or what ever.  It is what ever property YOU want
> to figure out the dynamic range of.

It not that time or space figure into the calculation, its just that they
built into the properties that we are interested in.

You say I can figure the dynamic range of whatever property I want.
Well I want the dynamic range of the density of the print!

It seems that you are insisting that I can only calculate (i.e. have) the
dynamic range of the contrasts of the print.

> 
> > For you "signal" is difference
> > between to tonal densities, i.e:
> >     largest signal = dMax - dMin
> >     smallest discernable signal = smallest density difference
> > or maybe I can paraphrase "signal is basically contrast".
> > (I hope I'm not putting words in your mouth!)
> 
> Sure, I'll fly with that.
> 
> > As far as I can tell your derivations from this and all the
> > discussions are consistant and mathematically sound.  But I've
> > been a little unsure of picking this mapping rather than
> > something else.
> 
> I disagree with your use of the term "mapping".  Again, you pick the
> property you want to figure out the dynamic range of...note the dynamic
> range equation does NOT use any units for largest/smallest...they simply
> have to be on the same scale, of the same unit.

Right.  And I want the property "density".

> 
> > For pictures, what's the basic perception?  It seems you've picked
> > contrast --  so the ratio is (highest contrast/lowest contrast).
> 
> Yes, since contrast (as in delta change) is all that distinguishes tones
> from one another.
> 
> > Why not the more basic perception: darkness versus lightness?
> 
> What, exactly, do you mean by "darkness" and "lightness"?

Can't you go with: darkness = high density, lightness = low density??

> 
> > The ratio would be (darkest/lightest). I guess you can see
> > where this is going->> it reduces the whole thing to
> > Density Range.
> >
> > DynRange = log10 (darkest/lightest)
> >          = log10 (darkest) - log10 (lightest)
> >          = dMax - dMin
> >          = Density Range
> 
> But you have now removed, arbitrarily from what I can tell, the smallest
> discernable signal from the equation, that above, you agreed was part of the
> equation.  Why?  This seems like one of those Gary Larson "and then magic
> happened" on the chalk board...

I figured dMin = smallest discernable density  = smallest discernable signal
I haven't any problem keeping the word "discernable" throughout.

The fact that density, as commonly used, already has a log10 built into it
makes the math manipulation slightly different order.  The idea is to distribute
the log10 function or the ratio.  The log10(largest density) becomes dMax,
log10(smallest discernable density) becomes dMin and finally the divide
becomes a subtract.

> 
> > I doubt that you're ready to switch definitions but intuitively
> > using "dynamic range" to means "number of tones" rather
> > than some "range" of values seems weird.
> 
> Not at all, my guess is it really means dynamic OVER THE particular
> range...but I do not know the history of the term.  It isn't really A range,
> but it describes a property OF the given range.  Density range isn't a range
> either, but it does describe the width OF the range of density.

"Range" as commonly used in math is used to describe the set of possible
values.  Density range given by dMax and dMin entirely describes the
possible density values i.e. they all have to lie between dMax and dMin.

> 
> > And by the prevalence
> > of the "dynamic range" threads that go on regularly many
> > others don't feel comfortable with your definition.
> 
> Yeah, but I didn't make up the definition, as you well know!  Remember, this
> is a very technical issue, being described to and by and used by a lot of

True, but, being technical does NOT exempt it from scrutany and passing the test
of usefulness. 

I'll acknowledge that you didn't make up the definition taken from the Higgins
book, but I've yet to see corroborating info for your interpretation of
"signal" in the equation.   Higgins most definitely does NOT address how to
apply the equation to the photographic prints, so I'm assuming you 
decided how to do that.   As you say below there's lots of info about.
I'd like to see more supporting your  interpretation.

> lay people.  There are a LOT of misperceptions/understandings when this type
> of thing happens...especially these days with the proliferation of the
> Internet.  You can find "compelling" (as in decent presentations, that look
> well thought out) sources for both sides of near any argument!
> 
> > Here's an interesting thought experiment, Say we a printing
> > out gray swatches from an inkjet printer.  Each swatch
> > is a small square of each grayscale value.  Assuming we're
> > talking 8 bit files and printer drivers, we can have only
> > 256 possible gray swatches.  If its all perfectly
> > calibrated ideally we ought to be able to distinguish
> > each swatch.  Now imagine a swatch which is a checkboard
> > with two different adjacent gray values, for instance
> > 137 and 138 mixed together.  What does it look like?  Is
> > it appear as a checkerboard of two different grays or
> > does it merge into a new and different 137.5 gray swatch?
> > I think the answer is "it depends" ... i.e. how close you
> > get to it, got your bifocals on? etc.
> 
> It does depend.  More so on lighting than anything else.  We actually can
> distinguish a lot more than 100 gray tones, but the 100 number is in "a"
> given light...but vary that light, and you increase the number of tones you
> can see...and your standard photographic print may not have tones that
> extend into that range, since most people don't view a photo in the dark or
> extreme light, so they weren't designed TO be seen in those conditions.
> 
> > The point is "number of tones" is kind of nebulous.  This
> > also adds to my dissatisfication with "dynamic range"
> > meaning "number of tones".
> 
> Very interesting point, but I have an answer ;-)  You are right, in your
> example, you will be able to shift the tones all by .5 of a tone, and get a
> whole new set of shifted "discernable" tones...but you can STILL only
> discern the same number, as well as have the same separation!  This does

Well I was figuring that you're adding all these new tones -- i.e. double the
number of tones.  But I do agree we are probably beyond the capability
of human discerning of all these tones.  My real point here was just that its
hard to get a handle on "how many grays?".   Here's a system which is
based on 8 bit values -- i.e. 256 values and there's at least a potential for
actually having more the 256 grays.  On the surface that seems very weird.

> fail at some point, when your number of tones is high enough (where adjacent
> tones aren't discernable).  Does that make sense?
> 
> > Hopefully someone will find this useful.
> 
> Yes.
> 
> Regards,
> 
> Austin

-Roy

> > > Now the big question is "What's the signal?".
> >
> > Very simply, for a photographic image, it is the density value.
>
> This is precisely what I'm trying to say.   Let's take the
> original definition
> equation and substitute "density value" for "signal".  We have:
>
> Dynamic Range = log10 (largest signal/smallest discernable signal)
>             =  log10 (largest density value/smallest discernable
> density value)
>
> Isn't that precisely the same as: dMax - dMin  as I derived last post?

Well, no, you're taking a "leap" here to arrive where you did.  As I've said
a dozen times, "largest" is an ambiguous term, as well as "smallest".  It is
open to interpretation if left undefined.  The equation for dynamic range
clearly uses largest to mean the "absolute value of the maximum detectible
signal", in other words, the amplitude...dMax - dMin, as opposed to just the
single measurement for dMax...

> It not that time or space figure into the calculation, its just that they
> built into the properties that we are interested in.

That's my point, they don't.  You can take a measurement in and of it self,
and you don't have to relate it at all to time and/or space.  They aren't
part of the dynamic range equation at all.

> You say I can figure the dynamic range of whatever property I want.
> Well I want the dynamic range of the density of the print!

But that's what this entire discussion has been about, hasn't it?

> It seems that you are insisting that I can only calculate (i.e. have) the
> dynamic range of the contrasts of the print.

Contrast (in out case) simply means delta density, right?  In order to
distinguish density, you MUST have contrast.

> > I disagree with your use of the term "mapping".  Again, you pick the
> > property you want to figure out the dynamic range of...note the dynamic
> > range equation does NOT use any units for largest/smallest...they simply
> > have to be on the same scale, of the same unit.
>
> Right.  And I want the property "density".

That's no different than we have been talking about.  Contrast merely means
there is a difference, and as far as the discussion goes, density is the
same as contrast.  High contrast means large distinguishable difference, low
contrast means small distinguishable difference.

> > > For pictures, what's the basic perception?  It seems you've picked
> > > contrast --  so the ratio is (highest contrast/lowest contrast).
> >
> > Yes, since contrast (as in delta change) is all that distinguishes tones
> > from one another.
> >
> > > Why not the more basic perception: darkness versus lightness?
> >
> > What, exactly, do you mean by "darkness" and "lightness"?
>
> Can't you go with: darkness = high density, lightness = low density??
>
> >
> > > The ratio would be (darkest/lightest). I guess you can see
> > > where this is going->> it reduces the whole thing to
> > > Density Range.
> > >
> > > DynRange = log10 (darkest/lightest)
> > >          = log10 (darkest) - log10 (lightest)
> > >          = dMax - dMin
> > >          = Density Range
> >
> > But you have now removed, arbitrarily from what I can tell, the smallest
> > discernable signal from the equation, that above, you agreed
> was part of the
> > equation.  Why?  This seems like one of those Gary Larson "and
> then magic
> > happened" on the chalk board...
>
> I figured dMin = smallest discernable density  = smallest
> discernable signal
> I haven't any problem keeping the word "discernable" throughout.

Yeah, but that's wrong.  dMin is merely the smallest signal level that has
been measured, and has nothing to do with the ability to discern anything.
You could have a dMin of 2...but your ability to distinguish a signal could
be much smaller, like .01.  dMin and smallest discernable signal are
entirely different.  This goes back to my statement that smallest is
ambiguous in the definitions given, except mine ;-) which included a diagram
that showed what was meant by smallest...

> > > I doubt that you're ready to switch definitions but intuitively
> > > using "dynamic range" to means "number of tones" rather
> > > than some "range" of values seems weird.
> >
> > Not at all, my guess is it really means dynamic OVER THE particular
> > range...but I do not know the history of the term.  It isn't
> really A range,
> > but it describes a property OF the given range.  Density range
> isn't a range
> > either, but it does describe the width OF the range of density.
>
> "Range" as commonly used in math is used to describe the set of possible
> values.  Density range given by dMax and dMin entirely describes the
> possible density values i.e. they all have to lie between dMax and dMin.

No, not at all.  It describes the MAXIMUM and MINIMUM value, but not the
resolution of values in between.  Counting from 0-100, you could count by
10's or 5's or 1's, you have the same range, but different number of steps.


> I'll acknowledge that you didn't make up the definition taken
> from the Higgins
> book, but I've yet to see corroborating info for your interpretation of
> "signal" in the equation.

Signal means nothing but some distinguishable quality.

> Higgins most definitely does NOT
> address how to
> apply the equation to the photographic prints, so I'm assuming you
> decided how to do that.

I didn't have to decide on anything.  Density values are "signals", as they
signal a quality of the density.  Though signal is commonly used in
electronics, that doesn't preclude other things from being considered a
signal.  In fact, dynamic range is commonly used with simple numbers, like
number of bits has a particular dynamic range...and there is no electronics
involved in that!

> As you say below there's lots of info about.
> I'd like to see more supporting your  interpretation.

I don't believe this is something that people write down, it's just an
inherent understanding.  There is no book on how to count from 2 to 3 that I
know of, or someone substantiating that you can count trees with integer
values....  Hum.  I'll have to think about that.

I guess if you understand the concept of dynamic range, it should be obvious
how it can be applied to photographic prints with no problem...I mean, even
Martin said dynamic range has been spoken about with respect to photography
for a long time, so did Ansel Adams...so others apparently believe dynamic
range applies to chemical photographic prints...  As far as interpretation
goes, well, as I've said, I didn't make up the dynamic range equation, so
it's not up to me to interpret it.  The use and definitions I've given to it
are common amongst any engineers I know of who understand dynamic range...
So I am on very firm ground I've gotten the equations and definitions of
terms used in the equation correct, and others have agreed dynamic range
applies to chemical prints...so I don't know what more I can substantiate.

Another thing is dynamic range does in fact describe a unique
property...that is NOT defined by density range.  If dynamic range were
simply the same thing as density range, they there probably would not be two
distinct terms, but there are.  Density range does not describe the same
thing that dynamic range does, and that is why dynamic range exists...

Austin

> > Very interesting point, but I have an answer ;-)  You are right, in your
> > example, you will be able to shift the tones all by .5 of a
> tone, and get
> a
> > whole new set of shifted "discernable" tones...but you can STILL only
> > discern the same number, as well as have the same separation!  This does
> > fail at some point, when your number of tones is high enough (where
> adjacent
> > tones aren't discernable).
>
> As in continuous tone? <G>

That doesn't make it continuous.  Continuous is merely a perception...so
whether it "appears" continuous is solely based on the sensor used to
perceive it.

Neither chemical prints, or inkjet prints ARE in reality continuous tone.
As I've pointed out, you would need a sheet of paper of infinite size to
represent your definition of continuous tone (being entirely stepless), and
that does not exist.

Austin

> Okay I will add:
>
> From X-rite\ufffds \ufffdThe Color Guide and Glossary\ufffd
>
> \ufffdDynamic Range: An instrument\ufffds range of measurable values, from
> the lowest
> amount it can detect to the highest amount it can handle.\ufffd

Yeah, what's wrong with that?  It's exactly as I have said.  The lowest
AMOUNT (not least amount mind you) it can detect IS the "smallest
discernable signal".  Highest AMOUNT it can handle is still a bit dubious.

As I've said, provenance is the equation, period ;-)

> From CCD Direct:
>
> \ufffdThe dynamic range is often represented as a log ratio of well
> depth to the
> readout noise in decibels. For example, a system with a well
> depth of 45,000
> electrons and a readout noise of 15 electrons would have a dynamic range =
> 20 log (45,000/15), or 69dB.\ufffd
>
> http://www.ccddirect.com/online-store/scstore/dynamic.html
>
>
>
> You never did say what you thought of this last one which seems to be in
> your field.

It's fine, it includes noise, as it should, and the equation is correct...
The RANGE is 45,000, and the noise is 15, and they divide range by
noise...exactly as I have been saying every time...  A depth of 45,000 means
a Max of 45,000 and a min of 0...

I didn't comment on it because it's entirely correct, and fits everything
I've been saying to a "T".

> > > You have not been able to supply a value for the noise in your
> > > equation yet
> > > so how did you manage that.
> >
> > Er, yes I did.  I gave you an example using 0.01D as the "noise".
>
>
>
> Hmmm. Somehow an example just doesn't satisfy like an experimental value.
> What if I say for example that the noise in a silver print is
> 0.00000000000000001 dB can I use that? <G>

If you want, I have no problem with you using ANY value, all I'm trying to
get you to understand is chemical prints HAVE dynamic range.

> I have no reason to believe that noise exists everywhere and I can't take
> that as a given.

OK, but I can state unequivocally it does...at least in the physical world.
Noise doesn't exist in the number 1, but trying to measure 1", it does.  Can
you name any physical property that doesn't have any noise?

> The noise you described earlier would be
> variations within
> a given paper base rather than the variation from brand-to-brand
> and may be
> so small as to push the results of your equation towards infinity.

Source of noise isn't relevant to the dynamic range OF the print, the
print's dynamic range is the print's dynamic range, irrespective of where
the noise came from.  It is relevant to the dynamic range of the system
though.

> Well I am a very practical engineer. Chemical process and project
> engineering remember, so I want to know all those details.

I am a very practical engineer too...I like to understand what it is I am
trying to do here before getting into the details.  You and I haven't worked
out the basic concepts here, so the details aren't relevant.  Perhaps you
are having trouble seeing this through the details?

> True but you need a meaningful value for the noise to do that.

Yes you do.

> > > If you don't have the
> > > values you
> > > need to apply your dynamic range equation, what use is it?
> >
> > Having a "correct" value or not doesn't mean it doesn't exist!
> >
> No but lack of a correct value makes it of little practical use.

Yeah, but I'm trying to convince you it even exists, not anything about the
practicality of it...that's a waste of time until you understand the concept
and it's existence!

> I did not mean to impugn your printing skills.

I didn't take it that way at all...it's an entirely legitimate
question/concern!

> No. That would be terrible ineffiecient and waste of valuable time.  <G>

Doesn't that create more noise in your system though, or are you just so
damped that it's just all din ;-)

> I'm just on the west coast and a bit of an insomniac, and can keep posting
> past your bed time but you leave me a whole stack to look at with
> my morning
> coffee.

Oh, sorry, I don't drink coffee ;-)

BTW, do you have children?

Regards,

Austin

> No offense Austin, but you really snipped out all my relevant points and
> didn't answer much concretely.

Sorry, I answered what I thought was important, and, what I believe, was
quite clearly!

> My main point was that for a formula that you suggests indicates
> the number
> of tones present in a print, based on sampling as few as three tones,

No, my formula does not suggest that at all.  It does suggest measuring TWO,
dMax and dMin, but noise/smallest discernable signal may take many tones to
measure.

> You also did not speak to my premise that I think that a calculation of
> tones from DyR is only possible if linearity is assumed, which is not
> something I would take for granted in a silver print.

You don't need linearity at all.  Noise can be calculated using calculus,
and integrated over the entire range of the print...  Knowing math is not
your strong point?  I am guessing you don't know what integration is, in a
mathematical sense?  I can explain if you like.

Regards,

Austin

--- In DigitalBlackandWhiteThePrint@y..., "Austin Franklin" <darkroom@i...> wrote:
> 
> > > > Now the big question is "What's the signal?".
> > >
> > > Very simply, for a photographic image, it is the density value.
> >
> > This is precisely what I'm trying to say.   Let's take the
> > original definition
> > equation and substitute "density value" for "signal".  We have:
> >
> > Dynamic Range = log10 (largest signal/smallest discernable signal)
> >             =  log10 (largest density value/smallest discernable
> > density value)
> >
> > Isn't that precisely the same as: dMax - dMin  as I derived last post?
> 
> Well, no, you're taking a "leap" here to arrive where you did.  As I've said
> a dozen times, "largest" is an ambiguous term, as well as "smallest".  It is
> open to interpretation if left undefined.  The equation for dynamic range
> clearly uses largest to mean the "absolute value of the maximum detectible
> signal", in other words, the amplitude...dMax - dMin, as opposed to just the
> single measurement for dMax...

I don't understand why you're having trouble here.   "largest", "smallest",
"maximum", "minimum"  are all pretty basic math concepts, I can't believe
you really need me to define them.

The real issue here is the "leap" of interpretation.  Let's go back to the Higgins
book -- to the pages about dynamic range.  There's the formula on one page
(using the word signal) and there's the sine wave diagram on the next page
(the same one you posted way back).   The sine wave diagram talks about
"signal" and shows a large amplitude and a small amplitude.  In the audio world
its real easy to look at it and think "voltage" oscillating back and forth over time. 
But --- we want to apply the idea to photographic prints.  Hmm!! How shall
we do this??   Density sure seems like an important concept but the diagram
certainly doesn't have any labels that say "density", so what to do?

Here's where the "leap" of interpretation HAS to occur. You've come up with
a way to put density onto the signal diagram and so have I. We both had to
"interpret" the signal diagram as it applies to density.  The trouble is we've
come up with different interpretations!!  There's nothing in the book that
addresses or supports one interpretation over another.  I think this is the
real issue here.  You have CHOSEN what you think is the most obvious
interpretation, in fact it seems so obvious to you that its hard to think of
it as a choice.   I, likewise, have chosen an interpretation but it is different from
yours.  I believe both interpretations are mathematically sound and can both
lead to lots of consistent derivations (within their own framework).

First of all, I guess I have to convince you that we have both made choices.
If that works, we can agree to have different interpretations leading to two
definitions of Dynamic Range or somehow decide which one is more appropriate
for modeling prints as human view them.  I contend that there isn't mathematical
argument because both are sound, its just a difference (better/worse) argument.

My idea is to show how you put the labels on the graph and how I put them
on the graph.  The signal diagram in the book looks like a voltage sine wave
with time along the x-axis, right?  It seems like we both are going to replace
the x-axis with spacial dimension across the paper (I think we OK so far).
Now the tricky part, the large amplitude portion is the "loud" part in audio.
You are looking at the large amplitude and putting dMin at the bottom trough
and dMax at the top of the sine wave so that "signal" means (dMax-dMin).
Likewise, the  small amplitude is "soft" in audio language and for you its two
densities that are very close and just discernable.  How we doing so far?
Now, my turn.  What I want is to use the absolute value of the amplitude
to be a measure of the density.  To be precise I'll borrow some terminology
from the audio world  -- I want density = RMS (RootMeanSquare) of the
signal amplitude.  Well maybe that wasn't so clear, let me try with a 
what the signal diagram looks like in both audio world and print world.  In
audio its: loud section followed by soft section.  On the print I want it
to be a dark section followed by light section (i.e.  paper half black and then
half white).  In audio we talk about RMS Power output and it is max in the
"loud" section and min in the "soft" section.  And so in the print we have
density is max in the black section and min in the white section --- and I
want signal strength to be exactly analogous to density.
---  Question: what does your paper look like?  half high contrast and
      half low contrast?  It sure doesn't "feel" analogous to my audio
       example: half loud, half soft.

> Contrast (in out case) simply means delta density, right?  In order to
> distinguish density, you MUST have contrast.

I don't have much problem envisioning an all black paper or an all white paper.
No contrast but I sure know which is which.  The really weird thing about
contrast is that its JUST the transition.  There's no extent, it can't keep going.


> > "Range" as commonly used in math is used to describe the set of possible
> > values.  Density range given by dMax and dMin entirely describes the
> > possible density values i.e. they all have to lie between dMax and dMin.
> 
> No, not at all.  It describes the MAXIMUM and MINIMUM value, but not the
> resolution of values in between.  Counting from 0-100, you could count by
> 10's or 5's or 1's, you have the same range, but different number of steps.

Range is a really, really basic mathematical word.  I just looked it up in my
7th grade daughter's math book, its there.   Giving a Max and a Min is by
definition a precise description of a range.  Resolution is totally irrelevent.
dMax and dMin define an interval (i.e. range) where any density between
dMax and dMin is "in the range".  There's an infinity number of possible
densities (distinguishability is also irrelevant to what a range is).


> 
> > As you say below there's lots of info about.
> > I'd like to see more supporting your  interpretation.
> 
> I don't believe this is something that people write down, it's just an
> inherent understanding.  There is no book on how to count from 2 to 3 that I
> know of, or someone substantiating that you can count trees with integer
> values....  Hum.  I'll have to think about that.

Come on now.  There's lots of references to dynamic range as it applies 
to the audio world.  Where are the books about dynamic range as it
applies to prints?  If there aren't any maybe there aren't "accepted"
ways to look at it yet.  Is that why we are having this discussion?  To
add our two cents to the issue.  

> 
> I guess if you understand the concept of dynamic range, it should be obvious
> how it can be applied to photographic prints with no problem...I mean, even
> Martin said dynamic range has been spoken about with respect to photography
> for a long time, so did Ansel Adams...so others apparently believe dynamic

Have you read Adams short paragraph about it lately?   He pretty definitely
talks about it as max and min zones i.e a range of "useful values".
Put that on a negative or paper and we have exactly a "density range".

> range applies to chemical photographic prints...  As far as interpretation
> goes, well, as I've said, I didn't make up the dynamic range equation, so
> it's not up to me to interpret it.

I think you have most definitely "interpreted" it for usage with prints.  If it's
"common amongst engineers", there's most definitely got to be a book.

  The use and definitions I've given to it
> are common amongst any engineers I know of who understand dynamic range...
> So I am on very firm ground I've gotten the equations and definitions of
> terms used in the equation correct, and others have agreed dynamic range
> applies to chemical prints...so I don't know what more I can substantiate.
> 
> Another thing is dynamic range does in fact describe a unique
> property...that is NOT defined by density range.  If dynamic range were
> simply the same thing as density range, they there probably would not be two
> distinct terms, but there are.  Density range does not describe the same
> thing that dynamic range does, and that is why dynamic range exists...

You're really stretching here.  Dynamic range is just the generic term that
can be applied to many areas of study.  It just happens that it's already
got a well-known and commonly used name is the specific area of photography.

> 
> Austin

Well, Austin, I hope we're not wasting our time, here.  I think I really understand
how you've used the technical definitions and interpreted them.  I don't
think they are "wrong" per se, but I definitely do think there are other
ways to interpret them and potentially have a more useful concept.
All using precisely the same tried-and-true technical definitions that you
and engineers at least in some industies are very familiar.

Roy

Hi Roy,

> I don't understand why you're having trouble here.   "largest",
> "smallest",
> "maximum", "minimum"  are all pretty basic math concepts, I can't believe
> you really need me to define them.

That's the problem!  They ARE basic concepts, but they can mean more than
one thing, there are ambiguous terms being used.  I've cited this countless
times.  Largest can mean the maximum/largest "level" of the signal related
to 0, OR it can be the that I just mentioned, minus the lowest/minimum
"level" of the signal, again, related to 0, and it is therefore a range.
Why is that so hard to understand?

> The real issue here is the "leap" of interpretation.  Let's go
> back to the Higgins
> book -- to the pages about dynamic range.  There's the formula on one page
> (using the word signal) and there's the sine wave diagram on the next page
> (the same one you posted way back).

The diagram that I posted a picture of, was actually NOT a sine wave (there
is a sine wave to the left of it on the page, but that's not what I posted),
but a picture of a grove in an analog record ;-)

> The sine wave diagram talks about
> "signal" and shows a large amplitude and a small amplitude.  In
> the audio world
> its real easy to look at it and think "voltage" oscillating back
> and forth over time.

Hum.  I didn't know a record had any voltage ;-)

> But --- we want to apply the idea to photographic prints.  Hmm!! How shall
> we do this??   Density sure seems like an important concept but
> the diagram
> certainly doesn't have any labels that say "density", so what to do?

It also doesn't have a label for voltage either.  You are missing the
importance of the diagram.  It is merely an example showing what largest and
smallest are.

This is just SO simple.  Dynamic range is a concept.  It does not apply to
any particular thing, and because I, or anyone, uses an example of audio,
does not mean it only applies to audio.

All that is required for dynamic range are the variables listed in the
equation, largest signal and smallest discernable signal.  Largest meaning
the absolute value of the overall signal, and smallest discernable signal
being just what that means.

> You've
> come up with
> a way to put density onto the signal diagram and so have I. We both had to
> "interpret" the signal diagram as it applies to density.

I didn't have to interpret anything, the meaning of largest and smallest
discernable signal are clearly defined.

> The
> trouble is we've
> come up with different interpretations!!  There's nothing in the book that
> addresses or supports one interpretation over another.

But that's just not true.

> You have CHOSEN what you think is the most obvious
> interpretation, in fact it seems so obvious to you that its hard
> to think of
> it as a choice.

It is only obvious to me what it means, simply because I have worked with
dynamic range for over 25 years, and it has been exactly the same every
time, from the exercises in class, and professors and classmates that I've
had discussions with, for the hundreds of engineers I've work with, and for
the thousands of customers that have read my specifications, equations and
analysis.  It's a pretty sound base!

> First of all, I guess I have to convince you that we have both
> made choices.

You can't do that, because my terms are not by choice.  They are, and have
always been, clearly defined.

> If that works, we can agree to have different interpretations
> leading to two
> definitions of Dynamic Range

But any interpretation that does not match what I know dynamic range to be
is wrong.  It won't be dynamic range.  I did not make up dynamic range, nor
did I make up the definitions of the terms used in it.

> I contend that there
> isn't mathematical
> argument because both are sound, its just a difference
> (better/worse) argument.

I don't understand how claiming largest means highest level of signal and
smallest discernable signal means lowest level of signal is "sound".  If you
use those definitions, you are not describing DYNAMIC range.  Your
calculation will be in error.

Let's just talk about "smallest".  You CAN have a very high dMin, say 2, but
have the ability to distinguish .1 density, can you not?  Also some of the
definitions of dynamic range (the correct ones that is ;-) say noise is the
smallest discernable signal, and for the most part, it is.  Given that, are
you saying that dMin is simply noise?

> My idea is to show how you put the labels on the graph and how I put them
> on the graph.  The signal diagram in the book looks like a
> voltage sine wave
> with time along the x-axis, right?

Yeah, but it isn't.  It's a record grove....and you can just take a single
slice out of the diagram, it is not necessary to be a wave at all.

> It seems like we both are
> going to replace
> the x-axis with spacial dimension across the paper (I think we OK so far).

No, I don't use an XY graph at all, I don't graph anything.  That
"graph"/image was solely to show an EXAMPLE of what smallest discernable
signal and largest mean.

> Now the tricky part, the large amplitude portion is the "loud"
> part in audio.

Fine...

> You are looking at the large amplitude and putting dMin at the
> bottom trough
> and dMax at the top of the sine wave so that "signal" means (dMax-dMin).

Well, it doesn't matter if there is a picture of a sine wave or not.  The
point of the example is to show the extents/bounds of the signal, and it
doesn't matter if it's voltage or density.  I do agree that the two bounding
limits can be equated to dMax and dMin, and the "largest" in the dynamic
range equation, when applied to an image is dMax - dMin.

> Likewise, the  small amplitude is "soft" in audio language

Do you mean the vertical distance between the top and bottom of the "signal"
(not the bounds)?  That is neither loud or soft.  Loud is if the signal is
at the top of the bounds, and soft is if the signal is at the bottom of the
bounds.  The, what I believe you are calling "small amplitude" has nothing
at all to do with loud or soft.

**** (tagged for below)
> and
> for you its two
> densities that are very close and just discernable.  How we doing so far?

See above.  I think your term "soft" may be ambiguous...see below too ;-)

> Now, my turn.  What I want is to use the absolute value of the amplitude
> to be a measure of the density.  To be precise I'll borrow some
> terminology
> from the audio world  -- I want density = RMS (RootMeanSquare) of the
> signal amplitude.

You can't RMS does not apply to dynamic range.  It does apply to SNR, but we
are not talking about SNR...

> Well maybe that wasn't so clear,

Well, you're right there ;-)

> let me try with a
> what the signal diagram looks like in both audio world and print
> world.  In
> audio its: loud section followed by soft section.

Sorry, you lost me.  I don't believe your terms "loud section" and "soft
section" are really right...

> On the print I want it
> to be a dark section followed by light section (i.e.  paper half
> black and then
> half white).

He he, that I understand ;-)

> In audio we talk about RMS Power output and it is max in the
> "loud" section and min in the "soft" section.

As I've said, RMS etc. doesn't apply to dynamic range, but I'll see where
you're going with this anyway...  It seems you are using '"soft" section'
ambiguously.  Here it almost appears as you mean it as being a signal that
is close to the bottom bounds...but above, you say...see **** above.

> And so in the print we have
> density is max in the black section and min in the white section

Yes.

> --- and I
> want signal strength to be exactly analogous to density.

Fine.

> ---  Question: what does your paper look like?

You said above the paper was half black, then half white...so that's what
the paper looks like, right?

>  half high contrast and
>  half low contrast?  It sure doesn't "feel" analogous to my audio
>  example: half loud, half soft.

Where do you get "half" from?  You lost me here...

> The really weird thing about
> contrast is that its JUST the transition.  There's no extent, it
> can't keep going.

What do you mean by that?

>
> > > "Range" as commonly used in math is used to describe the set
> of possible
> > > values.  Density range given by dMax and dMin entirely describes the
> > > possible density values i.e. they all have to lie between
> dMax and dMin.
> >
> > No, not at all.  It describes the MAXIMUM and MINIMUM value, but not the
> > resolution of values in between.  Counting from 0-100, you
> could count by
> > 10's or 5's or 1's, you have the same range, but different
> number of steps.
>
> Range is a really, really basic mathematical word.  I just looked
> it up in my
> 7th grade daughter's math book, its there.   Giving a Max and a Min is by
> definition a precise description of a range.

Agreed.

> Resolution is
> totally irrelevent.

To the range, of course.  I never said any differently.

> dMax and dMin define an interval (i.e. range) where any density between
> dMax and dMin is "in the range".

Yes, of course.

> There's an infinity number of possible
> densities

...you didn't finish that sentence.  If you mean within the range, yes, you
are right, there is an infinite number of POSSIBLE densities, but in
reality, the number is bounded.

> (distinguishability is also irrelevant to what a range is).

Absolutely, and I've never said differently.


> Come on now.  There's lots of references to dynamic range as it applies
> to the audio world.

Yes, because engineers design audio equipment, and because companies have to
test this gear and produce specs for it, it is a specified piece of
information.

> Where are the books about dynamic range as it
> applies to prints?

Ansel Adams did ;-)  I am sure that the companies that make paper/chemicals
etc. have information on this, and it may or may not be published.  This is
a very different field than audio.  To expect there to be the same level or
requirement of specsmanship in photographic paper/printing as there is in
audio is really asking too much.  Most people who are interested in this
type of information for prints are few and far between, but for audio it's
every college kid...it's far more "pedestrian".

I am sure that somewhere dynamic range of prints is discussed.  I have about
a dozen books on subjects that may contain some information on it...  The
one book I'd start with is a well known "bible", "The Handbook of Modern
Halftone Photography" by Ewald Fred Noemer.  It, unfortunately, has no
bloody index...but it would be my first guess for a book that might discuss
dynamic range with respect to photography.

> If there aren't any maybe there aren't "accepted"
> ways to look at it yet.  Is that why we are having this discussion?  To
> add our two cents to the issue.

I'm sure it's been discussed...and I'm sure we aren't the first.

> Have you read Adams short paragraph about it lately?

I read what was posted, but I haven't read the source material to see
exactly what he was talking about.

>
> > range applies to chemical photographic prints...  As far as
> interpretation
> > goes, well, as I've said, I didn't make up the dynamic range
> equation, so
> > it's not up to me to interpret it.
>
> I think you have most definitely "interpreted" it for usage with
> prints.  If it's
> "common amongst engineers", there's most definitely got to be a book.

But Higgins shows the definition of dynamic range is...so it IS in a book!

> Dynamic range is just the generic
> term that
> can be applied to many areas of study.

I believe it is a very specific property, as defined by the dynamic range
equation.

> I don't
> think they are "wrong" per se, but I definitely do think there are other
> ways to interpret them and potentially have a more useful concept.

That's what I don't understand.  The definition you want to ascribe to it
isn't useful, as we already have a term for that...density range.  To me,
dynamic range is a very valuable property for imaging (as well as audio).

Phew!

Regards,

Austin

>> No offense Austin, but you really snipped out all my relevant points and
>> didn't answer much concretely.
> 
> Sorry, I answered what I thought was important, and, what I believe, was
> quite clearly!

Fair enough. Perhaps I'm not stating my points clearly enough, or it is I
who missed your point.

BTW, I still think your doing a great job trying to explain yourself to
everyone. Few have your patience and commitment.
 
>> My main point was that for a formula that you suggests indicates
>> the number
>> of tones present in a print, based on sampling as few as three tones,
> 
> No, my formula does not suggest that at all.  It does suggest measuring TWO,
> dMax and dMin, but noise/smallest discernable signal may take many tones to
> measure.

You're right, your formula did not state it, but back when I started this
line of inquiry I quoted the methodology you suggested was sufficient for
calculating the DyR of a print, which was in short: measure your prints
dmax, dmin, and various points from a solid tone between. This will give you
the values needed for your DyR calculation. Then from your DyR you have
another calculation that will tell you how many tones are present in a
print. Remember, you said that DyR is, de facto, the number of tones
present.

From this I took it that three tones where required/sufficient to make an
assumption of how many tones are present in the print.

Do you see where I'm going with this yet?

If you do have linearity of all the tones between your dmin and dmax your
assessment of the number of tones present will be correct. HOWEVER, that
calculation, at least if you take it so far as to assume a tone count from
it, will be way wrong if the print ONLY contains those three tones.

I don't have a big point here other than something doesn't compute and I'm
curious where the break in logic is. I'm going to restate it to be sure you
follow. What you've presented is this: a sampling from three tones is
sufficient to assess a print's DyR , and from that calculated value you can
calculate the number of tones present. It would seem to me that those
calculations work well for a system that produces a full and linear
placement of tones between your dmin and dmax, but when those conditions are
not met it would yield an inaccurate tone count. If I'm wrong please explain
why.

>> You also did not speak to my premise that I think that a calculation of
>> tones from DyR is only possible if linearity is assumed, which is not
>> something I would take for granted in a silver print.
> 
> You don't need linearity at all.  Noise can be calculated using calculus,
> and integrated over the entire range of the print...  Knowing math is not
> your strong point?  I am guessing you don't know what integration is, in a
> mathematical sense?  I can explain if you like.

Thanks for the offer. You are perceptive indeed about my math skills <G>,
but I consider my sense of logic to be pretty good, so if you can answer my
point above through logical explanation (or math as a last resort) I'd be
most appreciative. :-)

Thanks Austin!

Todd

> BTW, I still think your doing a great job trying to explain yourself to
> everyone. Few have your patience and commitment.

Thanks, but you three (Martin, you and Kevin) can really pack a LONG
response, though insightful and interesting...which is what keeps me at
this.  I also appreciate you guys ability to stick with this, and so far, I
believe I've made some headway in the discussion.  If anything, you guys
have helped to clarify/understand some gray (no pun intended) areas that
I've had in this my self.

> From this I took it that three tones where required/sufficient to make an
> assumption of how many tones are present in the print.

It gives you "some" information, yes, but upon further discussion, it became
apparent to me that it is prudent to actually characterize the noise to see
if it is variable over the density range, and if so, what is the equation
that defines the noise.  My belief is it's simply a linear gain adjustment,
but it could also be a ramp of some kind.  I don't really know, which is why
I think it would be really interesting to characterize it.  Martin seems to
know more about this than I do, and I'd like to get his input on this after
he gets his new spectrometer.  (are you listening, Martin ;-)

> I don't have a big point here other than something doesn't compute and I'm
> curious where the break in logic is. I'm going to restate it to
> be sure you
> follow. What you've presented is this: a sampling from three tones is
> sufficient to assess a print's DyR , and from that calculated
> value you can
> calculate the number of tones present. It would seem to me that those
> calculations work well for a system that produces a full and linear
> placement of tones between your dmin and dmax,

You are correct.

> but when those
> conditions are
> not met it would yield an inaccurate tone count. If I'm wrong
> please explain
> why.

You are absolutely right, and I'm happy that you have an understanding of
this to make such an observation!  Bravo!

This is where calculus comes into play.  It is a way of taking into account
the variability of the noise throughout the system.  Here's the deal.  The
noise "change" could be deterministic (possibly linear), which means it's
predictable.  If that is true (and I believe it is), then taking some number
of points (what ever are warranted by the type of equation that defines the
noise over the range, and if the noise is a simple gain, as in it's linear,
two points would do) in the "middle" will give you enough information to
derive the dynamic range.  Let's not talk about what that "result" or
equations would look like yet...

Regards,

Austin

>> but when those
>> conditions are
>> not met it would yield an inaccurate tone count. If I'm wrong
>> please explain
>> why.
> 
> You are absolutely right, and I'm happy that you have an understanding of
> this to make such an observation!  Bravo!

Thanks, and just think, it it only took you about 200 replies and a year of
your life to bring me to this point...
 
> This is where calculus comes into play.  It is a way of taking into account
> the variability of the noise throughout the system.  Here's the deal.  The
> noise "change" could be deterministic (possibly linear), which means it's
> predictable.  If that is true (and I believe it is), then taking some number
> of points (what ever are warranted by the type of equation that defines the
> noise over the range, and if the noise is a simple gain, as in it's linear,
> two points would do) in the "middle" will give you enough information to
> derive the dynamic range.  Let's not talk about what that "result" or
> equations would look like yet...

Okay, I'd be interested to see what you might come up with, because DyR as a
tone count has always been the hang-up in the whole conversation for me.

I fully appreciate that noise is an issue worthy of considering when
considering the ability to discern between tones in a system, and that IF a
system performs as expected you can make certain assumptions about the
capability of a print produced by that system.

BUT when assessing a print of unknown provenance, where we have no knowledge
of the capability of the system, wow, before you could compute a tone count
you'd need to assess the noise within virtually every tone within the DnR.

For instance, silver prints have characteristic curves which are non-linear,
and each brand and grade has it's own shape, which is further altered by
development...it'd be very difficult to make anything more than the most
general assumptions. Therefor to accurately determine the DyR of a "generic"
print of unknown origin the required measurements could be overwhelming.
Likewise for and inkjet print that contains mottling through some tones,
microbanding in others, etc. Plus all that is without the consideration of
the dynamic range of the
subject/material/file/negative/artistic-interpretation, the print is trying
to represent. Meaning not all prints are intended to be full toned and
linear!

Therefor, I find the concept of DyR much more useful for assessing the
CAPABILITY, of a well controlled system, but much less useful when assessing
singular prints. I look at it like this, you can pretty well assess the
dynamic range of a well behaved scanner, especially if it's linear across
the range of all the films on the market. But assessing the DyR of the same
scanner when there is an intermittent short in one or more of the
components, making for infinite variability between scans, would be much
harder. The infinite variability between scans represents the unknown
characteristics of a print of unknown origin.

I hope that made sense. ;-)

Todd

Austin

I think I'm winding down here on my participation in this thread, so I'm
taking stock of where we stand at this point. I'm curious after all that's
been discussed on this topic, what is your position on the usefulness of a
calculation of dynamic range on an "existing print of unknown origin", at
least with regard to giving an indication of how many tones THAT print
contains.

Lets face it DyR might be a useful measure of a SYSTEM, where you know the
values of a range you put through it, and you compare it to the values that
come out. That's how you test the DyR of a scanner right? You scan a target
of known values and you measure how the noise and non-linearity of the
system distorts those values upon output? I don't know, I'm asking. For a
print system you'd print a target of known values, measure those values at
the print, and compare?

If so, it's a comparison of input to output values that is relevant; or
calculations based upon assumed linearity, and if the system is linear, and
you can measure the noise, you can make an assumption or calculation of tone
count.

But in the case of a print of unknown origin, the only way to count the
tones present would be to do a full "pixel count" along the lines of a
Photoshop histogram calculation. I can see no way that a few well chosen
samples will get you there. Not without have a great understanding of the
materials underlying the process, but a print of unknown origin rules that
out.

Furthermore, tones will either be present or they will not, but how many are
present almost certainly be based upon far more powerful determinants than
"noise". For instance, the number of tones present in the original file or
negative - or the printmakers intent!

So really, what do you think?

Now this part I don't expect an answer on (though feel free), but I'm tying
my thoughts in with Martin's, just to summarize what I've taken from this
extremely thorough and enlightening discussion.

It seems to me that the formula you site for dynamic range may well be
useful in some small measure WRT a print, or large measure when speaking to
a SYSTEM'S CAPABILITY. But I think it's been shorthanded by the printmakers
to dmax - dmin, not because they are too ignorant to know better ,though
surely that plays it's part, but because they have found that given the
strong tonal manipulations they have control over, and the ability to place
tones as desired within a density range, they consider the "noise" in their
"system" something they have control over, while having a measure of the
minor issue of "noise" at the print surface, (like a bit of tooth to the
paper), is insignificant in scale, and moot to their concerns, and thus
assigned a value near zero and dropped.

Furthermore, as we see in inkjet prints, you can have microbanding,
mottling, etc, (i.e. noise) and still perceive a tone. It's really a
function of the amount of noise relative to the geometric dimensions of a
tone; or the amount of noise of the system relative to the detail of the
print. (I know that's vague, but it's very tangential, so I'll leave it at
that.)

IOW, it's about perception, and the photographer feels that within the
density range of his print s/he can place the tones where they need be, and
if there is noise enough that it becomes visibly distracting, they know it
without a spectro or densitometer. And if they can't perceive it they don't
mind?

Obviously materials designers and engineers will feel otherwise, and take a
different approach. In that scenario a controlled test would be set up,
input/output values would be compared, and your formula would be
appropriately applied.

Now to be fair to you, you never said photographers should make DyR
measurements, you were just explaining how the dynamic range formula you
know so well is applied, and making the distinction between it and density
range. You did a great job with that. I know I will be more careful how I
use those terms in the future. I will consider strongly the dynamic range of
a "printing system", but honestly, I don't expect to refer to the dynamic
range of a "print" much - density range speaks more solidly to my concern
there. Furthermore, I think other descriptive terms, like tonal separation,
linearity, microbanding, tooth, refraction, etc, - and their negative
counterparts - are much more useful and less nebulous than a measurement of
"noise".

Thanks,
Todd

Hi Austin,

Wow!  There's definitely two different brain wave lengths in progress.
But, seriously, I've spent a fair amount of time understanding your position
on the whole matter.  I've seen your long discussions with others in this thread
and awhile ago on the Piezo list.  My gut feel was against some of your claims
but gut feel certainly isn't much of an argument (is it:) ).  So I didn't jump in
right away.  I figured out what you were saying and thought a lot about what
"made sense" to me.  I understand and acknowledge the concepts and
formulas you've used and derived.  You've contributed a lot to this group and
probably have a fair amount of influence.  Heck, you are persistant:)  

Anyway, I'm hoping I can convince you to try and understand my position
as opposed to re-explaining what you are doing.  I'm sorry that I have an unfair
advantage since I'm been reading your stuff much longer than you've seen mine.
You've got "over 25 years" of experience in dynamic range related stuff.  This is
great but I'm assuming most if not all is in the audio/video world.  I'm not usually
a name dropper but I do have two EE degrees from MIT and Stanford so I'm
not a techno-neophyte.  

I sure you have a very good intuitive and technical feel for audio and similar
systems.  I'll only claim pretty good for me.  While concept of dynamic range of
prints has been around, I think the technical definition and calculation of dynamic
range of prints is relatively new.

The way humans perceive sound versus a print is very different.  How you map
concepts in one to the other is not all trivial and predefined.


--- In DigitalBlackandWhiteThePrint@y..., "Austin Franklin" <darkroom@i...> wrote:
> Hi Roy,
> 
> > The real issue here is the "leap" of interpretation.  Let's go
> > back to the Higgins
> > book -- to the pages about dynamic range.  There's the formula on one page
> > (using the word signal) and there's the sine wave diagram on the next page
> > (the same one you posted way back).
> 
> The diagram that I posted a picture of, was actually NOT a sine wave (there
> is a sine wave to the left of it on the page, but that's not what I posted),
> but a picture of a grove in an analog record ;-)

Sorry, I was going from memory.  But the gist is exactly the same: when you
put the record on a player the moving needle creates a voltage that mimics
the same shape.

> 
> > The sine wave diagram talks about
> > "signal" and shows a large amplitude and a small amplitude.  In
> > the audio world
> > its real easy to look at it and think "voltage" oscillating back
> > and forth over time.
> 
> Hum.  I didn't know a record had any voltage ;-)
> 
> > But --- we want to apply the idea to photographic prints.  Hmm!! How shall
> > we do this??   Density sure seems like an important concept but
> > the diagram
> > certainly doesn't have any labels that say "density", so what to do?
> 
> It also doesn't have a label for voltage either.  You are missing the
> importance of the diagram.  It is merely an example showing what largest and
> smallest are.

Sure its an example but without labels (explicit or at least implied) its totally
useless.  With a background in audio, its probably impossible to look at the
diagram without intuitively seeing a record needle track (for us older guys),
a speaker cone motion, a voltage oscillation or something similar.

If we try to use this diagram as an example for the print domain, don't you
need to put some labels on the axes??  How else can we look at the "largest"
label and tell what it measures?

> 
> This is just SO simple.  Dynamic range is a concept.  It does not apply to
> any particular thing, and because I, or anyone, uses an example of audio,
> does not mean it only applies to audio.
> 
> All that is required for dynamic range are the variables listed in the
> equation, largest signal and smallest discernable signal.  Largest meaning
> the absolute value of the overall signal, and smallest discernable signal
> being just what that means.

Yep, but you still have to DECIDE what constitutes "signal".  

> 
> > You've
> > come up with
> > a way to put density onto the signal diagram and so have I. We both had to
> > "interpret" the signal diagram as it applies to density.
> 
> I didn't have to interpret anything, the meaning of largest and smallest
> discernable signal are clearly defined.

Like I said: You still have to DECIDE what constitutes "signal". I.e what property
do we measure to plug into the equation.  You can't just say "density".
You're "largest" is Dmax - Dmin
Mine is just Dmax.
You're "smallest" variation within a single gray
Mine is just Dmin.

> 
> > The
> > trouble is we've
> > come up with different interpretations!!  There's nothing in the book that
> > addresses or supports one interpretation over another.
> 
> But that's just not true.
> 
> > You have CHOSEN what you think is the most obvious
> > interpretation, in fact it seems so obvious to you that its hard
> > to think of
> > it as a choice.
> 
> It is only obvious to me what it means, simply because I have worked with
> dynamic range for over 25 years, and it has been exactly the same every
> time, from the exercises in class, and professors and classmates that I've
> had discussions with, for the hundreds of engineers I've work with, and for
> the thousands of customers that have read my specifications, equations and
> analysis.  It's a pretty sound base!

Is that 25 years of experience in dynamic range of photographic prints??
Remember my issue is NOT whether you understand dynamic range in general,
I'm concerned with how it's applied to prints.

> 
> > First of all, I guess I have to convince you that we have both
> > made choices.
> 
> You can't do that, because my terms are not by choice.  They are, and have
> always been, clearly defined.
> 
> > If that works, we can agree to have different interpretations
> > leading to two
> > definitions of Dynamic Range
> 
> But any interpretation that does not match what I know dynamic range to be
> is wrong.  It won't be dynamic range.  I did not make up dynamic range, nor
> did I make up the definitions of the terms used in it.

You gotta realize that you know lots about dynamic range in one area, but
that applying it to a completely different area takes some decisions about mapping
concepts from one to the other.

> 
> > I contend that there
> > isn't mathematical
> > argument because both are sound, its just a difference
> > (better/worse) argument.
> 
> I don't understand how claiming largest means highest level of signal and
> smallest discernable signal means lowest level of signal is "sound".  If you

----> I guess poor choice of word.    Definition:  sound: valid

> use those definitions, you are not describing DYNAMIC range.  Your
> calculation will be in error.
> 
> Let's just talk about "smallest".  You CAN have a very high dMin, say 2, but
> have the ability to distinguish .1 density, can you not?  Also some of the
> definitions of dynamic range (the correct ones that is ;-) say noise is the
> smallest discernable signal, and for the most part, it is.  Given that, are
> you saying that dMin is simply noise?
> 
> > My idea is to show how you put the labels on the graph and how I put them
> > on the graph.  The signal diagram in the book looks like a
> > voltage sine wave
> > with time along the x-axis, right?
> 
> Yeah, but it isn't.  It's a record grove....and you can just take a single
> slice out of the diagram, it is not necessary to be a wave at all.

The groove gets converted to that at playing, right?  Try to play that groove
without moving it in time.  Can't take a single slice out of sound, can you?
You can't have sound without time.


> > You are looking at the large amplitude and putting dMin at the
> > bottom trough
> > and dMax at the top of the sine wave so that "signal" means (dMax-dMin).
> 
> Well, it doesn't matter if there is a picture of a sine wave or not.  The
> point of the example is to show the extents/bounds of the signal, and it
> doesn't matter if it's voltage or density.  I do agree that the two bounding
> limits can be equated to dMax and dMin, and the "largest" in the dynamic
> range equation, when applied to an image is dMax - dMin.
> 
> > Likewise, the  small amplitude is "soft" in audio language
> 
> Do you mean the vertical distance between the top and bottom of the "signal"
> (not the bounds)?  That is neither loud or soft.  Loud is if the signal is
> at the top of the bounds, and soft is if the signal is at the bottom of the
> bounds.  The, what I believe you are calling "small amplitude" has nothing
> at all to do with loud or soft.

Well if the diagram is a record groove, the loud music is when the groove swings
back and worth from the top bound to the bottom bound i.e. the amplitude
of the audio signal is max.  The soft music is where groove moves thru a
"small amplitude".  (Did you really say that small amplitude signal has nothing
to do soft (i.e. quiet) audio?? Did you?)  I'm sorry but I keep reading that
last sentence you wrote.  You do know how records work, don't you?


> > let me try with a
> > what the signal diagram looks like in both audio world and print
> > world.  In
> > audio its: loud section followed by soft section.
> 
> Sorry, you lost me.  I don't believe your terms "loud section" and "soft
> section" are really right...
> 
> > On the print I want it
> > to be a dark section followed by light section (i.e.  paper half
> > black and then
> > half white).
> 
> He he, that I understand ;-)
> 
> > In audio we talk about RMS Power output and it is max in the
> > "loud" section and min in the "soft" section.
> 
> As I've said, RMS etc. doesn't apply to dynamic range, but I'll see where
> you're going with this anyway...  It seems you are using '"soft" section'
> ambiguously.  Here it almost appears as you mean it as being a signal that
> is close to the bottom bounds...but above, you say...see **** above.
> 
> > And so in the print we have
> > density is max in the black section and min in the white section
> 
> Yes.
> 
> > --- and I
> > want signal strength to be exactly analogous to density.
> 
> Fine.
> 
> > ---  Question: what does your paper look like?
> 
> You said above the paper was half black, then half white...so that's what
> the paper looks like, right?
> 
> >  half high contrast and
> >  half low contrast?  It sure doesn't "feel" analogous to my audio
> >  example: half loud, half soft.
> 
> Where do you get "half" from?  You lost me here...
> 
> > The really weird thing about
> > contrast is that its JUST the transition.  There's no extent, it
> > can't keep going.
> 
> What do you mean by that?

The more I keep reading your responses, the more you seem to be glossing
over what you are doing.  You've referenced the signal diagram lots of times
saying it completely defines how to measure "signal".  You've got to relate
the squiggles in the diagram to something in the real world.  For audio
interpretation it has to MEAN something i.e. relate to sound.  I thought
the audio interpretation was obvious but the little loud/soft discussion above 
makes me wonder about even that.  The print interpretation is even more
abstract.  The diagram (or make a new one) somehow  has to relate to
a print i.e. show me what a largest signal versus a smallest signal looks
like on paper.  If you can't say what a "largest signal" looks like or
what a "smallest signal" looks like, how are you going to recognize it in
a print?   Briefly going back to the audio world, would you agree that
"largest signal" is the loudest sound you get from the speakers and
"smallest signal" is the quietest (noise included, of course)??
For me in the print world, my "largest signal" looks like pure black and
my "smallest signal" looks like pure white.  Can you describe yours
as simply?

> 
> 
> > Come on now.  There's lots of references to dynamic range as it applies
> > to the audio world.
> 
> Yes, because engineers design audio equipment, and because companies have to
> test this gear and produce specs for it, it is a specified piece of
> information.
> 
> > Where are the books about dynamic range as it
> > applies to prints?
> 
> Ansel Adams did ;-)  I am sure that the companies that make paper/chemicals
> etc. have information on this, and it may or may not be published.  This is
> a very different field than audio.  To expect there to be the same level or
> requirement of specsmanship in photographic paper/printing as there is in
> audio is really asking too much.  Most people who are interested in this
> type of information for prints are few and far between, but for audio it's
> every college kid...it's far more "pedestrian".
> 
> I am sure that somewhere dynamic range of prints is discussed.  I have about
> a dozen books on subjects that may contain some information on it...  The
> one book I'd start with is a well known "bible", "The Handbook of Modern
> Halftone Photography" by Ewald Fred Noemer.  It, unfortunately, has no
> bloody index...but it would be my first guess for a book that might discuss
> dynamic range with respect to photography.

Actually, I've looked at a few books that use the term "dynamic range" as
related to prints.  None have formal definitions, but all use it as
a  range (i.e. min and max) of useful zones, exposures, densities,
brightness/darkness, etc.


> 
> But Higgins shows the definition of dynamic range is...so it IS in a book!

  Certainly not for prints!

> 
> > Dynamic range is just the generic
> > term that
> > can be applied to many areas of study.
> 
> I believe it is a very specific property, as defined by the dynamic range
> equation.
> 
> > I don't
> > think they are "wrong" per se, but I definitely do think there are other
> > ways to interpret them and potentially have a more useful concept.
> 
> That's what I don't understand.  The definition you want to ascribe to it
> isn't useful, as we already have a term for that...density range.  To me,
> dynamic range is a very valuable property for imaging (as well as audio).

What you calculate may be valuable in some way, but I don't think its 
what others would use the term "dynamic range of a print" for. 

> 
> Phew!
> 
> Regards,
> 
> Austin

Phew is right!!
Regards,

Roy

> I think I'm winding down here on my participation in this thread, so I'm
> taking stock of where we stand at this point. I'm curious after all that's
> been discussed on this topic, what is your position on the usefulness of a
> calculation of dynamic range on an "existing print of unknown origin", at
> least with regard to giving an indication of how many tones THAT print
> contains.

Obviously, it will differ from print to print.  I certainly believe you can
measure it, and the first method that comes to mind is to scan it on a drum
and see what the numbers come out like!

> For a
> print system you'd print a target of known values, measure those values at
> the print, and compare?

Sounds good.  I believe Polaroid are the only ones who have actually
characterized these things via a testing procedure.  I'll ask David if he
happens to have a copy of the testing procedure.

> But in the case of a print of unknown origin, the only way to count the
> tones present would be to do a full "pixel count" along the lines of a
> Photoshop histogram calculation. I can see no way that a few well chosen
> samples will get you there.

I agree...that's a matter of measurement though...but I have certainly, and
I believe most anyone has, been able to say that this print has more/less of
a dynamic range than that print.  Whether that was intentional in the image
or not, isn't relevant, the point is, you can visually see with your own
eyes that two prints have different dynamic ranges.

> Furthermore, tones will either be present or they will not, but
> how many are
> present almost certainly be based upon far more powerful determinants than
> "noise". For instance, the number of tones present in the original file or
> negative - or the printmakers intent!

Right, but again, you are measuring THE print, and only THE print, you don't
care about the origin of the "missing tones".

> But I think it's been shorthanded by the
> printmakers
> to dmax - dmin,

Well, that I don't agree with.  That is clearly density range, and dynamic
range will give you a different "dynamic" than density range will.  I don't
believe it's "shorthand" at all.  I believe density range IS density range,
and dynamic range IS dynamic range.

> And if they can't perceive it
> they don't
> mind?

Right, but as I've said, because your eyes don't perceive it, doesn't mean
it isn't there.

> Now to be fair to you, you never said photographers should make DyR
> measurements, you were just explaining how the dynamic range formula you
> know so well is applied, and making the distinction between it and density
> range.

Correct ;-)

Regards,

Austin

> Therefor, I find the concept of DyR much more useful for assessing the
> CAPABILITY, of a well controlled system, but much less useful
> when assessing
> singular prints.

Using dynamic range in a "gross" (meaning not scientifically determined, but
purely visual) sense, I find dynamic range very important in comparing
prints.  In a very scientific sense, it is obviously useful for system
analysis.

> The infinite variability between scans represents the unknown
> characteristics of a print of unknown origin.
>
> I hope that made sense. ;-)

Yes, but I really don't believe it's infinitely variable.  I believe the
variability (if it even exists) can easily be characterized.

Regards,

Austin

Hi Roy,

> If we try to use this diagram as an example for the print domain,
> don't you
> need to put some labels on the axes??

I actually believe axes aren't very useful for our discussion.  I believe
you can just use one axis and describe what you need only along that one
axis, as follows:


----|..........|=|.......|-----

    ^ dMin      ^ noise  ^ dMax

and the axis is marked in density values.

> Yep, but you still have to DECIDE what constitutes "signal".

All that does is effect accuracy of your results...  I think one can
reasonably "find" the signal without much "decision" ;-)

> > > You've
> > > come up with
> > > a way to put density onto the signal diagram and so have I.
> We both had to
> > > "interpret" the signal diagram as it applies to density.
> >
> > I didn't have to interpret anything, the meaning of largest and smallest
> > discernable signal are clearly defined.
>
> Like I said: You still have to DECIDE what constitutes "signal".
> I.e what property
> do we measure to plug into the equation.

Ah, you mean in a descriptive sense, not measurement...well, scratch the
above ;-)

>  You can't just say "density".

Why not?  If that's what I am measuring the dynamic range OF, then why can't
I just "say" it?

> You're "largest" is Dmax - Dmin
> Mine is just Dmax.
> You're "smallest" variation within a single gray
> Mine is just Dmin.

And...yours don't work in the dynamic range equation...

> Is that 25 years of experience in dynamic range of photographic prints??
> Remember my issue is NOT whether you understand dynamic range in general,
> I'm concerned with how it's applied to prints.

25 years of using the concept of dynamic range.  I really can't say when
with respect to specifically photographic prints.

> You gotta realize that you know lots about dynamic range in one area, but
> that applying it to a completely different area takes some
> decisions about mapping
> concepts from one to the other.

That's what I disagree with.  I believe dynamic range is a very specific
property, and has a very specific purpose, and in your use, it does not
follow that purpose, it is merely density range, which is an entirely
different property than dynamic range.

> Can't take a single slice out of
> sound, can you?

But that's a property of the medium.  You can perceive sound without light,
but you can't perceive images without light.  I believe that "property" is
not important to the dynamic range.  You CAN get a one dimensional "image"
of signal min, signal max and noise completely without time.  "signal min"
and "signal max" as used in the dynamic range equation.

> You can't have sound without time.

But it isn't sound you're measuring, you don't have to hear it to measure
it, in fact, when measuring dynamic range of audio gear, I never hear it ;-)

> Well if the diagram is a record groove, the loud music is when
> the groove swings
> back and worth from the top bound to the bottom bound i.e. the amplitude
> of the audio signal is max.

Correct.  Gets that magnet all charged up ;-)

> The soft music is where groove moves thru a
> "small amplitude".

Yes.

> (Did you really say that small amplitude
> signal has nothing
> to do soft (i.e. quiet) audio?? Did you?)

If I did, it was because I didn't understand what you meant by "soft"...I
believe I thought you were using "soft" ambiguously, and asked you to
qualify your meaning.  I certainly know what you mean by it here.

> You do know how records work, don't you?

Yes, very well, thank you.

> You've referenced the signal diagram
> lots of times
> saying it completely defines how to measure "signal".

Well, if that's what I said, it really should say it defines the terms, by
example, used in the dynamic range equation.  It doesn't tell you how to
measure anything, though it does tell you what TO measure.

> The diagram (or make a new one) somehow  has to relate to
> a print i.e. show me what a largest signal versus a smallest signal looks
> like on paper.  If you can't say what a "largest signal" looks like or
> what a "smallest signal" looks like, how are you going to recognize it in
> a print?

But I have qualified what the terms mean, with respect to the dynamic range
equation.  Largest is dMax - dMin, very simply.

>  Briefly going back to the audio world, would you agree that
> "largest signal" is the loudest sound you get from the speakers

It depends on what you mean by "largest signal".  As it pertains to dynamic
range, it's the maximum amplitude.  This requires both the loudest AND
softest.

> and
> "smallest signal" is the quietest (noise included, of course)??
> For me in the print world, my "largest signal" looks like pure black and
> my "smallest signal" looks like pure white.  Can you describe yours
> as simply?

Well, yes.  I have dozens of times.  The terms "largest" and "smallest" are
different for density range then they are for dynamic range.

For the dynamic range equation, Dynamic range (dB) = 10log10(largest
signal/smallest discernable signal), the terms are defined as follows:

"largest signal" is the (highest value measured minus the lowest value
measured).

"smallest discernable signal" can be measured/derived in different ways, and
that depends on the application.  In an audio system, you can typically see,
and measure the amplitude of the noise on a scope.  In an image, it would
have to be derived, possibly as I've described in other posts.

> Actually, I've looked at a few books that use the term "dynamic range" as
> related to prints.  None have formal definitions, but all use it as
> a  range (i.e. min and max) of useful zones, exposures, densities,
> brightness/darkness, etc.

Would you mind providing a list of these references?

> > But Higgins shows the definition of dynamic range is...so it IS
> in a book!
>
>   Certainly not for prints!

In fact, he says "Dynamic Range EXAMPLE".

> What you calculate may be valuable in some way, but I don't think its
> what others would use the term "dynamic range of a print" for.

I understand, and that is simply because of misunderstanding/misuse.  Just
like "resolution".  It is also very misused, especially when it is related
to monitors.  It drives the monitor engineers absolutely batty.  I know a
few ;-)

> Phew is right!!

phew**2!

;-)

Regards,

Austin

> I actually believe axes aren't very useful for our discussion.  I believe
> you can just use one axis and describe what you need only along that one
> axis, as follows:
>
>
> ----|..........|=|.......|-----
>
>     ^ dMin      ^ noise  ^ dMax
>
> and the axis is marked in density values.

In thinking about that, and the discussion with Todd, I believe that, though
a single measurement is useful, in order to better characterize the system,
you could extend that to an XY plot, with that diagram being the X axis, and
density value that the noise was measured at, being the Y axis.  That would
give you a nice diagram of the noise over the entire range of the system.

Austin

Okay, 

I think we are in much agreement here. I think we both agree that if a
measuring scheme of sufficient complexity is used one can plug those numbers
into your dynamic range formula and get a result. My only quibble is that
the complexity, or points of sampling, would be so extensive that the data
collected might be better used in something like a histogram, which shows a
scale of range, with standard deviation etc, rather than the relatively
simplistic dynamic range formula.

So I agree you could perform a DyR calculation on a print, but it wouldn't
be worth while for the complexity of measuring, and for the limited
conveyance of what those many measurements represent. But we are entitled to
our own feelings about this. You obviously have a close relationship with
DyR and it holds a lot of meaning for you.


>> But in the case of a print of unknown origin, the only way to count the
>> tones present would be to do a full "pixel count" along the lines of a
>> Photoshop histogram calculation. I can see no way that a few well chosen
>> samples will get you there.
> 
> I agree...that's a matter of measurement though...but I have certainly, and
> I believe most anyone has, been able to say that this print has more/less of
> a dynamic range than that print.  Whether that was intentional in the image
> or not, isn't relevant, the point is, you can visually see with your own
> eyes that two prints have different dynamic ranges.

Wow, that is so variable I wouldn't be willing to go out on a limb like
that! Since your concept of DyR is based upon the density range of the
print, and the number of tones between, if one has a greater range, but the
other has more intermediary tones, which has a higher DyR? Not an easy
choice.

I could conceive of dozens of scenarios where one couldn't eyeball the
difference in DyR as you've defined it, however, density range is a much
simpler call to eyeball.
 
>> Furthermore, tones will either be present or they will not, but
>> how many are
>> present almost certainly be based upon far more powerful determinants than
>> "noise". For instance, the number of tones present in the original file or
>> negative - or the printmakers intent!
> 
> Right, but again, you are measuring THE print, and only THE print, you don't
> care about the origin of the "missing tones".

Exactly my point. That you have no idea how many tones a print SHOULD have
is what diminishes the importance of "noise" as the defining difference
between DyR and DnR in a print. How do you characterize Artistic intent"?
 
>> But I think it's been shorthanded by the
>> printmakers
>> to dmax - dmin,
> 
> Well, that I don't agree with.  That is clearly density range, and dynamic
> range will give you a different "dynamic" than density range will.  I don't
> believe it's "shorthand" at all.  I believe density range IS density range,
> and dynamic range IS dynamic range.

And I agree with you in a technical sense. We can have two different
definition, and use them accordingly. But as you vacillate between what
needs to be measured vs what can be eyeballed, so do others, and I am
suggesting that most print makers have decided the noise component of a
print is less than relevant.

I'm not saying they are using the term dynamic range "properly", I'm just
trying to discover why your more thorough concept of DyR hasn't carried over
to printing. I believe it's because the technology has not existed until
recently to do the full "pixel count" that is required in order to overcome
the non-linearity issue. Thus your formula for DyR when applied to a print
(not a system) would as often as not yield inaccurate or misleading results!

For some reason you have not been willing to consider that different fields
may have different jargons. I don't think we can limit the scope of
discussion to your 25 years in the field, when the fields of both
photography and electrical engineering are closer to 150 years old. In
earlier years, though your concept of DyR may have held some validity, I'm
guessing a practical process to implement it did not exist, thus the
terminology was allowed to drift into a less rigid interpretation.

Or, perhaps, more than likely, it was the other way around. Perhaps the
concept of DyR had it's origins in density range, and only when measuring
instruments evolved to the point where values as small as noise could be
assessed, they invented an equation to incorporate those values. Perhaps
even then the concept of noise at the print level was not deemed a large
enough factor to bother with by anyone less than an engineer.

At this point, if you care, I really think it is incumbent upon you to
demonstrate that anyone else applies your definition of DyR to a print,
because any web search using the terms "print (dynamic range)" will inundate
you with results that suggest otherwise. At a certain point you just have to
accept that one field uses the term differently than the other, pedestrian
or not.

Regardless, in the case of a print of unknown origin, the ability to
accurately measure the entirety of a print, which is required in the face of
non-linearity, which I believe must always be considered (because it
includes artistic intent) is very young indeed. Thus, until recently it
would have been virtually impossible to apply the DyR equation and expect
accurate results.

> 
>> And if they can't perceive it
>> they don't
>> mind?
> 
> Right, but as I've said, because your eyes don't perceive it, doesn't mean
> it isn't there.

Understood, I'm just speaking from a human/artistic perspective.

Thanks,
 
Todd