Digital BW, The Print

Hi,

I recently scanned a 35 mm Tri X negative at 3200 DPI in 16 bit 
mode. The image was of my daughter sitting on swing in the park in 
daylight.

I was curious to see if the histogram of this file would be 
different if the file was converted to 8 bit. After converting the 
file to 8 bit, I reviewed the histogram and so no changes.

Shouldn't there be a different histogram once the file is converted 
to 8 bit? Going from 16 bit to 8 bit means that the tonal range has 
been reduced, correct?

More importantly, I could not detect any visible differences on the 
screen when the image was in 8 bit vs. 16 bit. I decided to work 
with the 8 bit file as there are more PS features available this way.

Should there be any discernable tonality differences on a print 
created with a 16 bit file vs. 8 bit when printing with an Epson 
2200 using the BO method?

I printed the 8 bit and the 16 bit file and saw no differences. Is 
this what you would expect?

Everything I've read, says to stick with 16 bit.

Does the rule "Stay in 16 bit" apply when printing BO on Epson 2200?

Thanks,
David

No. 8-bits per channel does not mean the tonal range is reduced in 
any way.

8-bits per channel means you have 256 different values for 
expressing the full tonal range of each channel. 12-bits per channel 
gives you 4,096 values (most DSLRs, for example). 16-bits per 
channel means you have 65,536 possible values.

More bits per channel means the tones in your image are more 
continuous. Hence, the image is less likely to posterize when you 
edit the image. (Nonlinear changes -- such as Levels, Curves, 
Hue/Saturation adjustments, etc. -- compress parts of the histogram 
and expand others, making it more likely you will see visible 
posterization in your image where there are gaps in your histogram.

Until you start editing the image, the histogram for an 8-bit, 12-
bit, or 16-bit scan are likely to look similar. The mischief doesn't 
begin until you start editing the image. ;)

I stay in 16-bits throughout when possible, and as long as possible 
otherwise.

My RAW files are converted to a 16-bit PSD file. With Photoshop CS, 
you can do nearly everything in terms of photo editing in 16-bit. If 
you want to preserve the highest image quality, stick with 16-bits.

My master file flattens the working file, removes the alpha channels 
(necessary, if you use ImagePrint RIP), and saves it as a TIFF.

My output files for the Epson 2200 are sharpened for printed output 
and any final Curve adjustment is applied. This is when I convert to 
8-bits. Sending a 16-bit file to the Epson 2200 printer (or the 
ImagePrint RIP, in my case) gains you nothing in image quality. All 
you get is an output files that's twice the size.

So, I wind up with:

1 RAW File (12-bits, compressed)
1 or more Photoshop working files: PSD with layers and channels (12-
bits, compressed)
1 Master file: TIFF flattened and sans channels (12-bits, 
uncompressed)
1 or more Output files: TIFF (8-bits, uncompressed

If I use a RAW converter instead of Adobe Camera Raw II in PS CS, 
then I also have:

1 Converted file: TIFF (12-bits, uncompressed)

I hope this helps!

Cheers,

Mitch

> From: David Sinai [mailto:dsinai@...]
>
> Shouldn't there be a different histogram once the file is converted
> to 8 bit? Going from 16 bit to 8 bit means that the tonal range has
> been reduced, correct?

The tonal range is the same, so the histogram remains the same. The only
thing that's changed is the fineness of the steps. In 8-bit mode, each step
is only 1/256 of full scale. If you do a curve in 16-bit mode, you'll still
have a clean histogram, but if you do a curve in 8-bit mode, some of the
steps, which were 1/256 of full scale, will inevitably be turned into 1/128
of full scale, which will show up as missing lines (combing) in the
histogram. In some cases, this can correspond to faint but visible
posterization in the print.

If you have PS CS, you can stay in 16-bit mode to the end. If not, just make
sure you do your major curve adjustments in 16-bit mode, and do lesser
tweaks (e.g., sharpening) after converting to 8-bit.

--

Ciao,               Paul D. DeRocco
Paul                mailto:pderocco@...

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Glenn 
Mitchell" <gmitchel850@y...> wrote:

> 
> Until you start editing the image, the histogram for an 8-bit, 12-
> bit, or 16-bit scan are likely to look similar.

One reason for this is that Photoshop's histogram display 
provides inadequate resolution to see many of the contunuity 
gaps. There's a free plug-in available to use with Photoshop that 
allows you to view histgrams at increased resolution (4 or 5 
times more than the 256 levels of PS's built-in histogram 
function):
http://www.reindeergraphics.com/free.shtml#widehisto

This plug-in also shows the histogram for saturation, hue and 
luminance mode, as well as RGB .

Phil

David Sinai writes:

> I was curious to see if the histogram of this file would be
> different if the file was converted to 8 bit. After converting the 
> file to 8 bit, I reviewed the histogram and so no changes.

The histogram only shows data with 256 levels, so you can't see any
difference between 8-bit and 16-bit modes with the histogram.  To show
the 16-bit data, you'd need a histogram 65,536 pixels wide, which
obviously is not practical.  In 16-bit mode, every pixel you see
horizontally in the histogram is actually the sum of 256 different
levels (256 x 256 = 65,536 levels).

> Shouldn't there be a different histogram once the file is converted
> to 8 bit?

It would be nice, but as explained above, it wouldn't fit on the screen.

> Going from 16 bit to 8 bit means that the tonal range has
> been reduced, correct?

Yes.

> More importantly, I could not detect any visible differences on the
> screen when the image was in 8 bit vs. 16 bit.

In black and white, very few monitors can show any difference.  And you
must be in at least a 32-bit color mode to even be able to hold any
difference in display memory (and that's still smaller than the 48-bit
mode you'd need to show full 16-bit images).

> I decided to work with the 8 bit file as there are more PS
> features available this way.

But there is also less data.

The advantage to 16-bit mode is that you have far more headroom for
modifications.  For example, if you need to drastically change the
curves in the image, you can do so in 16-bit mode with almost no loss of
quality, whereas you'll get serious posterization in 8-bit mode,
particularly after multiple operations, since the effect is cumulative.

> Should there be any discernable tonality differences on a print
> created with a 16 bit file vs. 8 bit when printing with an Epson 
> 2200 using the BO method?

No.  The advantage to 16-bit is in the extra data you have for
manipulation.  You don't need to stay in 16-bit mode once you are
finished retouching and adjusting the image, and you can store it in
8-bit mode if you want (just remember, once you move it to 8-bit mode,
you won't have as much data if you ever decide to modify it again).

> I printed the 8 bit and the 16 bit file and saw no differences. Is
> this what you would expect?

Yes.

> Everything I've read, says to stick with 16 bit.

As long as you are adjusting or retouching the image and/or you expect
to do so in the future, 16-bit is better than 8-bit.  For actually
printing the image and archiving it, though, it doesn't matter, since no
display or print device can really resolve more than 256 levels, anyway.

> Does the rule "Stay in 16 bit" apply when printing BO on Epson 2200?

No.  It doesn't apply for any kind of printer or display.  None of them
can come close to resolving 65,536 different gray levels.

Thank you to all for the informative replies.
--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Anthony G. 
Atkielski" <anthony@a...> wrote:
> David Sinai writes:
> 
> > I was curious to see if the histogram of this file would be
> > different if the file was converted to 8 bit. After converting 
the 
> > file to 8 bit, I reviewed the histogram and so no changes.
> 
> The histogram only shows data with 256 levels, so you can't see any
> difference between 8-bit and 16-bit modes with the histogram.  To 
show
> the 16-bit data, you'd need a histogram 65,536 pixels wide, which
> obviously is not practical.  In 16-bit mode, every pixel you see
> horizontally in the histogram is actually the sum of 256 different
> levels (256 x 256 = 65,536 levels).
> 
> > Shouldn't there be a different histogram once the file is 
converted
> > to 8 bit?
> 
> It would be nice, but as explained above, it wouldn't fit on the 
screen.
> 
> > Going from 16 bit to 8 bit means that the tonal range has
> > been reduced, correct?
> 
> Yes.
> 
> > More importantly, I could not detect any visible differences on 
the
> > screen when the image was in 8 bit vs. 16 bit.
> 
> In black and white, very few monitors can show any difference.  
And you
> must be in at least a 32-bit color mode to even be able to hold any
> difference in display memory (and that's still smaller than the 48-
bit
> mode you'd need to show full 16-bit images).
> 
> > I decided to work with the 8 bit file as there are more PS
> > features available this way.
> 
> But there is also less data.
> 
> The advantage to 16-bit mode is that you have far more headroom for
> modifications.  For example, if you need to drastically change the
> curves in the image, you can do so in 16-bit mode with almost no 
loss of
> quality, whereas you'll get serious posterization in 8-bit mode,
> particularly after multiple operations, since the effect is 
cumulative.
> 
> > Should there be any discernable tonality differences on a print
> > created with a 16 bit file vs. 8 bit when printing with an Epson 
> > 2200 using the BO method?
> 
> No.  The advantage to 16-bit is in the extra data you have for
> manipulation.  You don't need to stay in 16-bit mode once you are
> finished retouching and adjusting the image, and you can store it 
in
> 8-bit mode if you want (just remember, once you move it to 8-bit 
mode,
> you won't have as much data if you ever decide to modify it again).
> 
> > I printed the 8 bit and the 16 bit file and saw no differences. 
Is
> > this what you would expect?
> 
> Yes.
> 
> > Everything I've read, says to stick with 16 bit.
> 
> As long as you are adjusting or retouching the image and/or you 
expect
> to do so in the future, 16-bit is better than 8-bit.  For actually
> printing the image and archiving it, though, it doesn't matter, 
since no
> display or print device can really resolve more than 256 levels, 
anyway.
> 
> > Does the rule "Stay in 16 bit" apply when printing BO on Epson 
2200?
> 
> No.  It doesn't apply for any kind of printer or display.  None of 
them
> can come close to resolving 65,536 different gray levels.

David: on a practical basis, I usually save scans as 8 bit files and
do all work from there. BO printing requires usually involves such
minor adjustments that I see no difference in the final print starting
and working with an 8 bit or 16 bit file. All the theoretical
responses are 100% correct, I just can't see an difference in the
final print. 

However I find that using UT inks requires large adjustments and make
saving as 16 bit files, and adjusting as 16 bit makes it worthwhile. 

But for me, 8 bit works fine with close to normal negs and BO
printing. At least if the final print is the criteria. 

Bob Michaels

 --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "David Sinai"
<dsinai@y...> wrote:

> Hi,
> 
> I recently scanned a 35 mm Tri X negative at 3200 DPI in 16 bit 
> mode. The image was of my daughter sitting on swing in the park in 
> daylight.
> 
> I was curious to see if the histogram of this file would be 
> different if the file was converted to 8 bit. After converting the 
> file to 8 bit, I reviewed the histogram and so no changes.
> 
> Shouldn't there be a different histogram once the file is converted 
> to 8 bit? Going from 16 bit to 8 bit means that the tonal range has 
> been reduced, correct?
> 
> More importantly, I could not detect any visible differences on the 
> screen when the image was in 8 bit vs. 16 bit. I decided to work 
> with the 8 bit file as there are more PS features available this way.
> 
> Should there be any discernable tonality differences on a print 
> created with a 16 bit file vs. 8 bit when printing with an Epson 
> 2200 using the BO method?
> 
> I printed the 8 bit and the 16 bit file and saw no differences. Is 
> this what you would expect?
> 
> Everything I've read, says to stick with 16 bit.
> 
> Does the rule "Stay in 16 bit" apply when printing BO on Epson 2200?
> 
> Thanks,
> David

There should be absolutely no perceptable differences with an inkjet print no matter 
whether it came from an 8-bit or 16-bit file.

This has been discussed ad infinitum on other boards. But don't take my word on it: 
Dan Margulis' seminal book "Professional Photoshop" pretty much debunks the need 
for 16-bit.

J Michael Sullivan


--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "David Sinai" <dsinai@y...> 
wrote:

> Should there be any discernable tonality differences on a print 
> created with a 16 bit file vs. 8 bit when printing with an Epson 
> 2200 using the BO method?
> 
> I printed the 8 bit and the 16 bit file and saw no differences. Is 
> this what you would expect?
> 
> Everything I've read, says to stick with 16 bit.
> 
> Does the rule "Stay in 16 bit" apply when printing BO on Epson 2200?
> 
> Thanks,
> David

J Michael Sullivan wrote:

>This has been discussed ad infinitum on other boards. But don't take my word on it: 
>
It's been discussed ad nauseum here too.. It's one of those repeating 
discussions that is interesting the first time or two, but after that..




Keith Krebs

"Just some guy," caretaker of the Multiverse's largest EPSON printer 
User Community (highly recommended by Vogon Poets and MegaDodo 
Publications), at:
http://groups.yahoo.com/group/EPSON_Printers/
and  the Multiverse's largest Canon printer User  Community at:
http://groups.yahoo.com/group/Canon-printers
"For the rest of you out there, the secret is to bang the rocks together 
guys"

It's discussed ad inifitum and ad nauseum because a small number of 
people claim there is no perceptable difference between 8-bit and 16-
bit images. This just misleads people.

Let's debunk the idea of trotting out a name and saying so-and-so 
has debunked the argument. That's a fallacious argument. It's a 
textbook example of the ipse dixit ("he says so") fallacy.

So, let's leave names out and deal with the logic of their 
argument . . . Give us the logic why you feel we should expect to 
see no difference, after applying nonlinear edits to the image.

Here's my real world experience . . .

If you take an image and do no editing to it, except to convert to 8-
bits and print or make a Web image, you'll likely notice no 
difference at all.

If you do edit your images, conversion to 8-bits can quickly lead to 
posterization that will be evident even on a low-end monitor and 
with an inexpensive ink jet printer. All you have is 256 values for 
each channel, and you only have to compress/expand some of the 
values on one of the channels to get non-continuous tones that 
generate visible banding. Think of skin tones or flower petals. They 
can quickly show signs of banding that look like splotches.

You can even get visible banding with 16-bit images. It's less 
likely, but not impossible with extreme nonlinear transformations of 
the information can. Information is not uniformly distributed in the 
typical image, and it can be skewed in many images. So you can 
compress/expand tones so there are discrete breaks in the data large 
enough to be visible in a print even with 16-bits per channel. With 
8-bits, it's quite easy to do it.

I do agree with the comment that BO printing is much less likely to 
result in banding. My comments addressed color images more than BO 
images. However, few people do BO printing with inkjet printers. 
Instead, neutrals are typically generated by combinations of CMY 
with B and sometimes b and sometimes even cmy. Typically, neutrals 
greater than midtone have no B or b whatsover, and neutrals less 
than midtone have progressively more B or b as you move towards 
shadows.

Grayscale images that are not BO rely on localized contrast 
differences in the color image to map into contrast differences in 
the grayscale image. Working to increase those contrast differences 
can result in visible banding even after grayscale conversion.

Cheers,

Mitch


--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "J Michael 
Sullivan" <michael@h...> wrote:
> There should be absolutely no perceptable differences with an 
inkjet print no matter 
> whether it came from an 8-bit or 16-bit file.
> 
> This has been discussed ad infinitum on other boards. But don't 
take my word on it: 
> Dan Margulis' seminal book "Professional Photoshop" pretty much 
debunks the need 

> for 16-bit.
> 
> J Michael Sullivan
>

> From: Glenn Mitchell [mailto:gmitchel850@...]
>
> If you take an image and do no editing to it, except to convert to 8-
> bits and print or make a Web image, you'll likely notice no
> difference at all.
>
> If you do edit your images, conversion to 8-bits can quickly lead to
> posterization that will be evident even on a low-end monitor and
> with an inexpensive ink jet printer. All you have is 256 values for
> each channel, and you only have to compress/expand some of the
> values on one of the channels to get non-continuous tones that
> generate visible banding. Think of skin tones or flower petals. They
> can quickly show signs of banding that look like splotches.

It really depends upon how noisy the image is to begin with. If you start
with an image from a cheap digicam, then the noise will drown out any
posterization. If you start with an image from a good scanner or a DSLR,
then posterization is possible. I've seen it.

--

Ciao,               Paul D. DeRocco
Paul                mailto:pderocco@...

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "J Michael
Sullivan" <michael@h...> wrote:
snip...

> Dan Margulis' seminal book "Professional Photoshop" pretty much
debunks the need 
> for 16-bit.

http://www.brucelindbloom.com/index.html?DanMargulis.html

I'll take Lindbloom over Margulis any day. Many have much to offer
including Dan, this would be one area others are far ahead,
particularly regarding concerns members of this list may have.
Tyler

Noise becomes more apparent because of the effects of posterization 
as you edit the image.

Chromatic noise which is barely noticeable when the image is loaded, 
for example, can be quite obvious as clumping or graininess after 
levels adjustments, curves adjustments, hue/saturation adjustments 
precisely because you are selectively expanding some information and 
selectively compressing other information. What starts out as 
continuous or near-continuous tones becomes less so. This effect 
will be more exaggerated with an 8-bit image than a 16-bit image.

Cheers,

Mitch 

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Paul D. 
DeRocco" <pderocco@i...> wrote:
> It really depends upon how noisy the image is to begin with. If 
you start
> with an image from a cheap digicam, then the noise will drown out 
any
> posterization. If you start with an image from a good scanner or a 
DSLR,

> then posterization is possible. I've seen it.
> 
> --
> 
> Ciao,               Paul D. DeRocco
> Paul                mailto:pderocco@i...

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tyler Boley" 
<tyler@t...> wrote:
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "J Michael
> Sullivan" <michael@h...> wrote:
> snip...
> 
> > Dan Margulis' seminal book "Professional Photoshop" pretty much
> debunks the need 
> > for 16-bit.
> 
> http://www.brucelindbloom.com/index.html?DanMargulis.html
> 
> I'll take Lindbloom over Margulis any day. Many have much to offer
> including Dan, this would be one area others are far ahead,
> particularly regarding concerns members of this list may have.
> Tyler

More of the same ipse dixit fallacy.

Actually, Bruce Lindbloom admits that Dan Margulis' claim is based 
on a bogus challenge. One that involves taking a 16-bit image, 
converting it to 8-bits, then back to 16-bits, then doing your edits 
and comparing with a file converted to 8-bits and then edited.

As he says, anyone who took the challenge was baited and should be 
ashamed for not seeing it from the outset.

Why was this point left out of the discussion?! It was right there 
in the linked discussion.

Why leave out other experts, who clearly disagree? Like Fraser and 
Blatner, Kieran, etc. That makes the evidence selective.

If one understands what Bruce Lindbloom is saying -- quite clearly, 
BTW -- Dan Margulis' test has the same information on both sides.

By converting from 16-bits to 8-bits and then back to 16-bits, you 
knock out off the intervening values. You will only have every 
information for every 256th possible value. 0, 256, 512 . . . 
because everything from 0 to 255 is a 0 in 8-bits. Now, going 0 and 
then 256, when mapped back to 8-bits is the same as going 0 and then 
1.

As Lindbloom acknowledges, if you do not knock off the extra 
information in this way, Dan Margulis' challenge fails. Another 
oversight that's crucial.

The fair test is to take a 16-bit image. Make a copy. Convert the 
copy to 8-bits. Now, make the same editing changes to both images. 
This retains all of the information in the 16-bit image instead of 
truncating intervening values. You will find that visible 
posterization develops faster and more severely in the 8-bit image.

My practical experience is 100% congruent with the argument that 8-
bit images are more likely to posterize as I edit them than the 12-
bit images that DSLRs capture and the 12- to 16-bit images scanners 
capture.

I and others who work in 16-bits are not talking theory. Our very 
real experience is that there is more editing headroom with 
increased bit depth.

Cheers,

Mitch

hi all,

i've been playing around with different prints using both the Clayton 
BO workflow, and Rorak workflow using the curves etc. for Ultra Tone,
i just got my 1280 - and after some initial figuring out - the printing 
is pretty smooth - and am getting nice prints -

So, I played around with both BO, and using the full ultra tone inkset 
- and I kinda like BO - and ultra tone gives very good prints as 
well...
it seems that each has its strengths, and depends on what you wanna get 
out of the prints... i like the feel i get from BO... i'm gonna show 
off some of my
prints to some friends of mine tomorrow.. and see what they think... i 
did show off some of them yesterday, and got good reviews...

one thing I've noticed with BO - is that in one of the prints there are 
some kind of horizontal lines ... i'm not sure if this is what is 
called "banding" ...
(am too new to know all the terms used here... :-)) however, when i use 
the full inkset - those lines go away - and the area is smooth... these 
lines
are showing up in areas of the shot where is there is a lot of sky... 
(evening sky)... and appear to be located in one particular area of the 
print...

probably it has to do with the image itself, rather than the printer... 
but i'm wondering, why this would show up in BO, and not when using the 
full inkset?

thanx

Altaf

> From: Glenn Mitchell [mailto:gmitchel850@...]
> 
> Noise becomes more apparent because of the effects of posterization 
> as you edit the image.
> 
> Chromatic noise which is barely noticeable when the image is loaded, 
> for example, can be quite obvious as clumping or graininess after 
> levels adjustments, curves adjustments, hue/saturation adjustments 
> precisely because you are selectively expanding some information and 
> selectively compressing other information. What starts out as 
> continuous or near-continuous tones becomes less so. This effect 
> will be more exaggerated with an 8-bit image than a 16-bit image.

Sure, but at least you won't see outright banding.

-- 

Ciao,               Paul D. DeRocco
Paul                mailto:pderocco@...

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Glenn Mitchell"
<gmitchel850@y...> wrote:
snip...
> > > Dan Margulis' seminal book "Professional Photoshop" pretty much
> > debunks the need 
> > > for 16-bit.
> > 
> > http://www.brucelindbloom.com/index.html?DanMargulis.html
snip...
> 
> More of the same ipse dixit fallacy.

Sorry, I don't speak latin.

> Actually, Bruce Lindbloom admits...
snip..
> Why was this point left out of the discussion?! It was right there 
> in the linked discussion.

Providing the link to the entire discussion hardly leaves any points
out, I think it could be considered all inclusive. I don't recall
advising ignoring anything on the site.

> Why leave out other experts, who clearly disagree? Like Fraser and 
> Blatner, Kieran, etc. That makes the evidence selective.

A poster brought up Margulis, I recalled Lindbloom's page and provided
it. Searching for and providing links to all relevant discussions on
the planet didn't occur to me.
You go on to restate much on Lindbloom's site, so I assume you agree
with it.

snip...
> I and others who work in 16-bits are not talking theory. Our very 
> real experience is that there is more editing headroom with 
> increased bit depth.

That's obvious and indisputable, providing the science (hardly theory)
simply supports it.
Many, including Fraser, have noted that in RGB some editing can be
done in 8 bit without deterioration at output. There are 3 channels
blending, and multiple inks dithering at output.
This is the digital black and white printing list. Most of the users
here are editing in gray at some point in the workflow, that's one
channel, no others will blend and possibly mask tonal gaps. A priority
for most here is linear continuous tone output in some form, this gets
better by the day. Therefore as output becomes more critically
accurate, the more critical the tonal integrity of the file. When was
the last time Margulis scanned, prepped, and output a single channel
file on a high quality monochromatic output system? Most of these
people are offering nothing relevant to our concerns. Editing an 8 bit
single channel file is throwing out info rapidly.
So as we continue to obsess here about perfect continuous tone output,
inks, printers, drivers, RIPs, curves, whatever, the only relevant bit
depth question is-
Why would you not be going to every conceivable extreme to maintain as
much tonal information as possible in your files?
Tyler

> Dan Margulis' seminal book "Professional Photoshop" pretty much
> debunks the need
> for 16-bit.

I hope there was a HUGE caveat on that, that he was SPECIFICALLY talking
about color, NOT B&W.  You really *need* do tonal corrections to high bit
data for B&W data.

Regards,

Austin

Wow. What a firestorm. Even a claim of gratuitous "name dropping"!! hmph.

I will stand by the following facts (verified by me and others -- yes, my amigo Bruce 
Frazer too -- many years ago):

1.) almost all high-end scanners today scan INTERNALLY at 10, 12 or 14-bits (only 
the cheapest consumer scannersnow  scan at 8 bits internally. Almost NO scanners 
actually scan at 16 bits))

2.) assuming that the SCAN curve is optimal for the negarive in question (i.e. the 
instructions to the scanner, including white point, black point, and shape of curve) 
there will be ABSOLUTELY NO DIFFERENCE between an optimal "8 bit" scan and a raw 
"16 bit" scan that has been manipulated to equal the "look" obtained by the above 
optimized curve -- regardless of the number of bits. 

It can best be diagrammed as such:

scanner (12bit) + optimal curve --> optimal file (8bit) --> printer (8bit)
scanner (12bit) --> expanded file (16bit) --> manipulate (16bit) --> optimal file 
(any bit) --> printer (8bit)

All things being equal, the above two workflows will produce the EXACT SAME IMAGE 
at the printer (i.e. there will no perceptible difference at all)...

...However, I *WILL* conceed the following scenario where manipulating in 16 bits has 
a theoretical advantage (it is certainly not true for most B&W images):

- specifically, the case where one wants to push the image WAY BEYOND what an 
optimal 8-bit scan would deliver, especially using algorithmic means (e.g. equalize, 
extreme curves, etc,).  An example of such extreme manipulation might be what 
Ansel Adam would have needed to do with "Moonrise Over Hernandez" -- where he 
burned the (light) sky practically black in his (anlaogue) enlarger. Were he to do the 
same in Photoshop today in 8 bit, then yes, one MIGHT see enhanced noise and 
posterization -- however, much of this "might" is determined by what you are 
attempting to do and HOW you go about doing it. It is NOT TRUE AT ALL that if you 
perform extreme manipulations in 8 bits, then you are destined for posterization and 
enhanced noise. Again, it depends more on the user, not the technology. For 
example, actual "manual" Photoshop burning and dodging in 8bits gives perfect 
gradations, devoid of any banding. No advantage at all doing burning/dodging in 
16bits

So the sum of this is this:

- if you are trying to obtain an optimal digitization of a negative using modern 
scanning equipment, and your goal is to print this optimized image file as is, then 
there is no advantage to capturing, storing, printing in 16 bits. In oher words, for the 
vast majority of images (and users), 8 bits together with an optimal scan is more than 
adequate, in spite of all the theorists arguments to the contrary

- however, if your goal is to manipulate -- to an extreme degree -- an image way 
beyond its optimal "zone", then capturing and manipluating in 16 bit may have some 
advantages. 

- As others have noted, printing in 16 bits is a total waste.

Michael Sullivan
author of "Make Your Scanner a Great Design & Production Tool" published waaay 
back in 1994, now out of print

P.S. assuming someone has unlimited memory & harddisk space then go ahead, scan 
away at 16 bits -- at least there is no harm! However, my own experience with most 
users is that they do in fact have limited memory and harddisk space. For these 
people, I would instead recommend to them to learn how to do GREAT scans and save 
the files in 8bit

P.P.S in my experience, a lot of the 16bit vs 8bit argument involves a perverse form of 
bragging -- those with the horsepower (and latest tools) tend to brag about it. We've 
all heard a version of this before: "my car can blow the doors off of your car" kind of 
thing. Tedious. Boring.   In the end, RESULTS is all that counts, not HOW you did it.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "J Michael
Sullivan" <michael@h...> wrote:

big snip...
I'm sorry Michael, but I take issue with almost every point you make.
But it seems everyone is sick of this discussion so I won't respond
with the mini-novel point for point response simmering in my brain.

> P.S. assuming someone has unlimited memory & harddisk space then go
ahead, scan 
> away at 16 bits -- at least there is no harm!

Well there you go, and that's the point.
By the way, 16 bit gray files are still a snap compared to even the 8
bit RGB or CMYK files we deal with every day.

> P.P.S in my experience, a lot of the 16bit vs 8bit argument involves
a perverse form of 
> bragging --

Thank you sir, I guess I'm overdue for some soul searching. In the
meantime I'll continue labor under the illusion that most of us here
are trying to optimize our workflows to approach the very high
standards set by a legacy of amazing B&W photography. I can see
justifying working in 8 bit gray, but I can't see advocating it.
Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Paul D. 
DeRocco" <pderocco@i...> wrote:
> > From: Glenn Mitchell [mailto:gmitchel850@y...]
> > 
> > Noise becomes more apparent because of the effects of 
posterization 
> > as you edit the image.
> > 
> > Chromatic noise which is barely noticeable when the image is 
loaded, 
> > for example, can be quite obvious as clumping or graininess 
after 
> > levels adjustments, curves adjustments, hue/saturation 
adjustments 
> > precisely because you are selectively expanding some information 
and 
> > selectively compressing other information. What starts out as 
> > continuous or near-continuous tones becomes less so. This effect 
> > will be more exaggerated with an 8-bit image than a 16-bit image.
> 
> Sure, but at least you won't see outright banding.
> 

That's not necessarily true at all.

Noise comes in varying degrees. It's effects on an image are not 
always uniform.

Luminosity noise is often speckled about. But chromatic noise often 
affects sky, shadows, water and is less/non-evident elsewhere in an 
image.

Noise is a separate issue from posterization owing to nonlinear 
transformations during image editing.

There is no reason to expect noisy images from a digicam to 
obliterate all evidence of posterization.

Cheers,

Mitch

Michael,

> 1.) almost all high-end scanners today scan INTERNALLY at 10, 12
> or 14-bits (only
> the cheapest consumer scannersnow  scan at 8 bits internally.
> Almost NO scanners
> actually scan at 16 bits))

The new Nikon in fact does have a 16 bit A/D, but the sensor/AFE (Analog
Front End) simply won't provide 16 *clean* bits...so yes, it does scan at 16
bits, but they aren't 16 *good* bits.

> 2.) assuming that the SCAN curve is optimal for the negarive in
> question (i.e. the
> instructions to the scanner, including white point, black point,
> and shape of curve)
> there will be ABSOLUTELY NO DIFFERENCE between an optimal "8 bit"
> scan and a raw
> "16 bit" scan that has been manipulated to equal the "look"
> obtained by the above
> optimized curve -- regardless of the number of bits.
>
> It can best be diagrammed as such:
>
> scanner (12bit) + optimal curve --> optimal file (8bit) --> printer (8bit)
> scanner (12bit) --> expanded file (16bit) --> manipulate (16bit)
> --> optimal file
> (any bit) --> printer (8bit)

Except that not all scanner software has a decent setpoint/curve tool, so
scanning in high bit mode, and setting the setpoints and making the tonal
adjustments in PS is a better workflow for that circumstance.  My scanner
software happens to have a great setpoint, curve tool, and histogram...so I
do the do all that in the scanner driver and get 8 bit data.

> All things being equal, the above two workflows will produce the
> EXACT SAME IMAGE
> at the printer (i.e. there will no perceptible difference at all)...

Of course, because the "optimal curve" step you outline above is being done
to the high bit data in the scanner.

> ...It is NOT TRUE AT
> ALL that if you
> perform extreme manipulations in 8 bits, then you are destined
> for posterization and
> enhanced noise.

If you are talking about B&W, it may be true that you aren't "destined for
posterization, but any substantial (and sometimes not so substantial) tonal
manipulations done to 8 bit grayscale data *can* certainly result in
posterization...and certainly sometimes quite readily.  It very much depends
on where along the curve these manipulations take place, and the image.

> - As others have noted, printing in 16 bits is a total waste.

Well, that is debatable...for B&W that is.  It's true that the human eye can
only discern less than 200 graytones given any singular lighting
situation...but...in order to not distinguish "adjacent" tones (which is
what you want to do to get smooth gradations, with no two tones immediately
perceptible next to each other), you can introduce intermediate tones...and
it improves the smoothness of the gradation.  As usual, this is very image
dependant.  So, having a few extra bits, say 12 bit grayscale data, or at
least the ability of the dithering process in the print driver to
accommodate the transitions, may give *some* benefit in *some*
circumstances.

Regards,

Austin

What you are advocating is a wasteful workflow for those who may have limited 
resources. It certainly is an EXPENSIVE workflow -- hence my assertion of bragging 
rights (which you quickly fell into) for those who, by right of better equipment and 
the latest software, are somehow "greater than thou". Yuch.

There are many on this list who seek to achieve "best practices" without the  
wastefulness of your 16-bit dogma (i.e. 16-bit is always better -- and by implication: 
8-bit users are less than real photographers).

My recommended workfow for most users remains: generate an *optimal* scan, save 
in 8-bit and you have achieved the equivalent of the zone system. You seem to not 
understand the concept of putting the effort into the scan before it gets into 
Photoshop.

Your workflow: scan anyway you want as long as it is in 16-bits. Doesn't matter how 
good or bad it is. Even raw if fine. Then when you have the inclination, fix it up. 
Great. Good for you. But your result (for the vast majority of all B&W images) will be 
as good as an optimal 8-bit workflow. And it still remains wasteful and suggestive of 
"anything less is not worthy of consideration". Talk about demogogery. Not everyone 
has Photoshop CS. Not everyone has 2gb ram. Not everyone has 1 Terabyte of hd. 

(however I do).

mjs



--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tyler Boley" <tyler@t...> 
wrote:

> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "J Michael
> Sullivan" <michael@h...> wrote:
> 
> big snip...
> I'm sorry Michael, but I take issue with almost every point you make.
> But it seems everyone is sick of this discussion so I won't respond
> with the mini-novel point for point response simmering in my brain.
> 
> > P.S. assuming someone has unlimited memory & harddisk space then go
> ahead, scan 
> > away at 16 bits -- at least there is no harm!
> 
> Well there you go, and that's the point.
> By the way, 16 bit gray files are still a snap compared to even the 8
> bit RGB or CMYK files we deal with every day.
> 
> > P.P.S in my experience, a lot of the 16bit vs 8bit argument involves
> a perverse form of 
> > bragging --
> 
> Thank you sir, I guess I'm overdue for some soul searching. In the
> meantime I'll continue labor under the illusion that most of us here
> are trying to optimize our workflows to approach the very high
> standards set by a legacy of amazing B&W photography. I can see
> justifying working in 8 bit gray, but I can't see advocating it.
> Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tyler Boley" 
<tyler@t...> wrote:

> snip...
> > I and others who work in 16-bits are not talking theory. Our 
very 
> > real experience is that there is more editing headroom with 
> > increased bit depth.
> 
> That's obvious and indisputable, providing the science (hardly 
theory)
> simply supports it.
> Many, including Fraser, have noted that in RGB some editing can be
> done in 8 bit without deterioration at output. There are 3 channels
> blending, and multiple inks dithering at output.

Who has said some editing cannot be done in 8-bits without 
deterioration?! That's not the point of contention at all.

The question is, is there a significant difference between 8-bit and 
16-bit images when it comes to editing. My experience says 
unequivocally "Yes."

You can make linear changes in an 8-bit image and a 16-bit image and 
the effect on both will be identical. A linear change is one that 
affects every pixel equally. Brightness is one example. It simply 
increments every pixel by the same value.

The trouble is nonlinear changes. Levels, curves, hue/saturation, 
etc. affect some pixels more than others. Hence, information in 
portions of the histogram are compressed and/or expanded and others 
are unaffected or affected more/less. It is nonlinear 
transformations that are likely to encounter visible posterization. 
The more changes and the more extreme those changes, the more likely 
an 8-bit image will show more evidence of posterization than a 16-
bit image.

Inks are not necessarily dithered at all. Not all inkjet printers 
use dithering and many use it only for specific modes, like BO 
printing with the Epson 2200.


> This is the digital black and white printing list. Most of the 
users
> here are editing in gray at some point in the workflow, that's one
> channel, no others will blend and possibly mask tonal gaps.

Unless you use BO ink, neutrals are printed using predominantly CMY 
ink (or CcMmY, etc. depending on how many colors are used in 
printing). B&W printing remains color printing unless you go to BO 
and suffer the disadvantages of dithering.

No black ink is used at all for neutrals brighter than midtone. Even 
0,0,0 -- black -- will be something like 75,68,67,90 in CMYK. Black 
ink is just 30% of the total ink. Go to 64,64,64 -- a three-quarter 
tone -- and you can expect something like 68,61,60,49. Blank ink 
drops to 21% of the total ink.

There is also the issue of where B&W images come from when the 
original digital image is a color image. Information from the color 
channels is not necessarily combined to make a single gray channel. 
There are B&W conversion techniques that retain three RGB channels.

Even if the information in the three channels (and perhaps other 
channels, such as the "l" channel from a Lab conversion of the K 
channel from a CMYK conversion) are combined in various proportions 
to make a single gray channel, there can be considerable editing 
going on prior to the conversion. Those can lead to posterization in 
the original color image *AND in the subsequent B&W image.

Cheers,

Mitch

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "J Michael 
Sullivan" <michael@h...> wrote:
> What you are advocating is a wasteful workflow for those who may 
have limited 
> resources. It certainly is an EXPENSIVE workflow -- hence my 
assertion of bragging 
> rights (which you quickly fell into) for those who, by right of 
better equipment and 
> the latest software, are somehow "greater than thou". Yuch.
> 

I don't understand any of this at all.

Why all of this ad hominem? I offered nothing of the kind in my 
replies.

99% of my images come from a DSLR, not a scanner. So, I do not have 
the luxury of rescanning, and I'll admit not every image I take is 
dead on for exposure. So, I do expect to edit my images in PS using 
nonlinear transformations like levels, curves, etc. So, I shoot 
everything in RAW, which is a 16-bit image, and I keep them that way.

What you seem to suggest would make my workflow more complicated. I 
should convert them to 8-bits, see if I have problems when editing, 
and if so, then go back and work in 16-bits.

I want the best image possible, and I prefer to keep my options 
open. Working in 16-bits gives me more flexibility.

Where is the added cost to working in 16-bits? More storage space? 
More RAM. I archive on DVDs. So, storage costs are not really an 
issue. I have sufficient RAM to work with very large files and run 
multiple apps simultaneously. I did not go out and purchase 
additional stoarage space or RAM to work with 16-bit images.


> There are many on this list who seek to achieve "best practices" 
without the  
> wastefulness of your 16-bit dogma (i.e. 16-bit is always better -- 
and by implication: 
> 8-bit users are less than real photographers).

I never said any of this. Your charactgerization is completely 
unfair.


> 
> My recommended workfow for most users remains: generate an 
*optimal* scan, save 
> in 8-bit and you have achieved the equivalent of the zone system. 
You seem to not 
> understand the concept of putting the effort into the scan before 
it gets into 
> Photoshop.

My guess is that most of us are using DSLRs and digicams for most of 
our images and scanners infrequently (if at all).

The discussion about Ansel Adams' "Zone System" is a non sequitur 
here. A good scan of a shot with exposure problems is little help. 
And if you think Ansel Adams himself had perfect exposure on every 
shot, you should visit his gallery sometime. For every keeper, he 
had lots of shots with problems. He was quite famous for doing 
considerable work in the darkroom to correct images with exposure 
and contrast problems.

I do not assume all of my shots will come out with perfect exposure. 
I expect most will be close. Some will have problems. Some will also 
need global or localized changes to enhance contrast, remove color 
casts, etc. So, I expect my images will undergo editing in PS. It is 
a rare image when I do not make some nonlinear adjustment.

If I expect my images will have problems, and 8-bit images are more 
likely to posterize than 16-bit images when editing, it seems a very 
reasonable strategy to stay in 16-bits until I generate my output 
files for something like the Epson 2200.


> 
> Your workflow: scan anyway you want as long as it is in 16-bits. 
Doesn't matter how 
> good or bad it is. Even raw if fine. Then when you have the 
inclination, fix it up. 
> Great. Good for you. But your result (for the vast majority of all 
B&W images) will be 
> as good as an optimal 8-bit workflow. And it still remains 
wasteful and suggestive of 
> "anything less is not worthy of consideration". Talk about 
demogogery. Not everyone 
> has Photoshop CS. Not everyone has 2gb ram. Not everyone has 1 
Terabyte of hd. 
> 
> (however I do).

Good for you! [Mitch applauds Michael for all that RAM and harddrive 
space.]

You do not need lots of harddive space and RAW to work in 16-bits. I 
get by with 60GB of harddrive storage and 512MB of RAW and can work 
in 16-bits consistently and not feel constrained. Those are specs 
that even a vanilla PC meets these days.

Cheers,

Mitch

J Michael Sullivan writes:

> Not everyone has Photoshop CS. Not everyone has 2gb ram.
> Not everyone has 1 Terabyte of hd.

And not everyone is manipulating color images.  A 16-bit image in black
and white requires 33% less memory and disk than a color image in 8-bit;
so anyone who can manipulate color images at all already has enough
memory and disk to manipulate 16-bit grayscale images.  So why not do
so?

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "J Michael 
Sullivan" <michael@h...> wrote:
> What you are advocating is a wasteful workflow for those who may have limited 
> resources. It certainly is an EXPENSIVE workflow -- hence my assertion of bragging 
> rights (which you quickly fell into) for those who, by right of better equipment and 
> the latest software, are somehow "greater than thou". Yuch.
> 
> There are many on this list who seek to achieve "best practices" without the  
> wastefulness of your 16-bit dogma (i.e. 16-bit is always better -- and by implication: 
> 8-bit users are less than real photographers).
> 
> My recommended workfow for most users remains: generate an *optimal* scan, save 
> in 8-bit and you have achieved the equivalent of the zone system. You seem to not 
> understand the concept of putting the effort into the scan before it gets into 
> Photoshop.
> 
> Your workflow: scan anyway you want as long as it is in 16-bits. Doesn't matter how 
> good or bad it is. Even raw if fine. Then when you have the inclination, fix it up. 
> Great. Good for you. But your result (for the vast majority of all B&W images) will be 
> as good as an optimal 8-bit workflow. And it still remains wasteful and suggestive of 
> "anything less is not worthy of consideration". Talk about demogogery. Not everyone 
> has Photoshop CS. Not everyone has 2gb ram. Not everyone has 1 Terabyte of hd. 
> 
> (however I do).

Whopdeedoo for you. I have none of those things.
"bragging  rights (which you quickly fell into) for those who, by right 
of better equipment and  the latest software, are somehow "greater than 
thou"."
Hmm.
Excuse me but you have no idea what my workflow is, or what my dogma 
is, I have stated neither. You also mis-characterized my "implication".
I'm out of this one kids, good luck.
Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Glenn Mitchell" <
gmitchel850@y...> wrote:

snip...

> The question is, is there a significant difference between 8-bit and 
> 16-bit images when it comes to editing. My experience says 
> unequivocally "Yes."

Mitch! Buddy! I'm agreeeing with you man!

> Unless you use BO ink, neutrals are printed using predominantly CMY 
> ink (or CcMmY, etc. depending on how many colors are used in 
> printing). B&W printing remains color printing unless you go to BO 
> and suffer the disadvantages of dithering.

Uh...
Hello?! Is this not the B&W list? I'd say most of the people here are 
printing with quad or some variation of monochromatic inks.
I guess I've been away too long.
I'll just mosey on back outa here.
Tyler

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Tyler Boley" 
<tyler@t...> wrote:
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Glenn 
Mitchell" <
 
> > Unless you use BO ink, neutrals are printed using predominantly 
CMY 
> > ink (or CcMmY, etc. depending on how many colors are used in 
> > printing). B&W printing remains color printing unless you go to 
BO 
> > and suffer the disadvantages of dithering.
> 
> Uh...
> Hello?! Is this not the B&W list? I'd say most of the people here 
are 
> printing with quad or some variation of monochromatic inks.
> I guess I've been away too long.
> I'll just mosey on back outa here.
> Tyler

I don't know who's printing what, and it's irrelevant to the point I 
was making.

You can resort to black ink only, black and light black ink only, 
quad tone inks, sepia tone inks, etc. for your B&W images. With some 
printers, that might even be the only solution for neutral B&W 
images.

However, most inkjet printers do not use black ink only or black and 
light black ink only for B&W images. Their grayscale tones are 
generated from a combination of colors, a relatively small 
proportion of which is black or black and light black.

I use a RIP and the standard Epson Ultrachrome CcMmYKk inkset for my 
B&W images. The ImagePrint Lite RIP makes absolutely neutral B&W 
images on the Epson 2200. No special inks needed.

Since I get B&W without using special inks, I don't assume others 
are using them. I'm sure some are, but I have no idea what 
proportion is using them.

Cheers,

Mitch

> From: Glenn Mitchell [mailto:gmitchel850@...]
>
> Luminosity noise is often speckled about. But chromatic noise often
> affects sky, shadows, water and is less/non-evident elsewhere in an
> image.
>
> Noise is a separate issue from posterization owing to nonlinear
> transformations during image editing.
>
> There is no reason to expect noisy images from a digicam to
> obliterate all evidence of posterization.

... other than the fact that my old Minolta DiMage 7 does in fact have so
much noise that I can't manage to coax any visible posterization out of its
images, no matter what I try. I've done some rigorous testing of it, taking
raw data and rounding it to eight bits, and then trying to create
posterization by applying really nasty curves to it, and then trying the
same thing on gamma 2.2 TIFFs. I can get horrible noise levels, but never
anything that looks like posterization.

My 10D is a completely different story...

--

Ciao,               Paul D. DeRocco
Paul                mailto:pderocco@...

Aside from the other points of view expressed here, IMO the BO part of
this thread is irrelevent.  It's the image that is either posterized
or not as a result of 8 vs 16 bit editing.  

Regards,
Clayton


Info on black and white digital printing at    
http://www.cjcom.net/digiprnarts.htm

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Paul D. 
DeRocco" <pderocco@i...> wrote:
> > From: Glenn Mitchell [mailto:gmitchel850@y...]
> >
> > Luminosity noise is often speckled about. But chromatic noise 
often
> > affects sky, shadows, water and is less/non-evident elsewhere in 
an
> > image.
> >
> > Noise is a separate issue from posterization owing to nonlinear
> > transformations during image editing.
> >
> > There is no reason to expect noisy images from a digicam to
> > obliterate all evidence of posterization.
> 
> ... other than the fact that my old Minolta DiMage 7 does in fact 
have so
> much noise that I can't manage to coax any visible posterization 
out of its
> images, no matter what I try. I've done some rigorous testing of 
it, taking
> raw data and rounding it to eight bits, and then trying to create
> posterization by applying really nasty curves to it, and then 
trying the
> same thing on gamma 2.2 TIFFs. I can get horrible noise levels, 
but never
> anything that looks like posterization.
> 
> My 10D is a completely different story...
> 

I hear ya!

I bought a DiMage 7i for a small camera to carry around for street 
photography. I tried it for a day and returned it. Luminosity and 
chromatic noise were dreadful! Shutter delay was excruciating. 

I got my money back from the dealer right away!

Cheers,

Mitch

--- "Glenn Mitchell" <gmitchel850@y...> wrote:
> You do not need lots of harddive space and RAW to work in 16-bits. 
> I get by with 60GB of harddrive storage and 512MB of RAW and can 
> work in 16-bits consistently and not feel constrained. Those are 
> specs that even a vanilla PC meets these days.


So I finally see what the real issue here is. You are talking about 
essentially lo-res DSL RGB captures, converted to B&W. No wonder you 
keep pounding on the need for 16-bits -- by God, you need them with 
today's "model T" digital cameras! (in ten years you will 
reluctantly agree with me that the current state-of-the-art 6.3m 
Canon and Nikons are indeed Model Ts)

I, on the other hand, have been consistently talking about scanning 
4x5 negatives with high-end scanning equipment to generate the 
highest possible quality B&W (and yes, color) images. For example, 
my 8-bit scans are 90mb for a 4x5 B&W neg scanned at 2400spi. (Given 
the resolution characteristics of my Scheider lenses, you could even 
go up to 3200spi for most large format images, thereby generating an 
even bigger file size. However, 4000spi for large format is mostly a 
waste). Were you to scan the same above image at 16-bit you would be 
at 180mb. Add four layers and your 16-bit 4x5 image is now a 
whopping 720mb (as compared to a 4-layer 8-bit image that 
would "only" be 360mb). With such numbers you would quickly fill up 
your 60GB hard drive in a matter of months (or weeks or even days if 
you were a professional). Hence my (and others) need for a terabyte. 
Funny thing, backing up a terabyte ain't no easy task. Think about 
it. And here I can't disagree with you more: Harddisk space is NOT 
free -- there is a REAL cost associated with it, in spite of your 
admonitions to the contrary. And anyone who is prolific with 
photography will encounter this problem sooner rather than later. 16-
bit can indeed be wasteful.

As such, I stand by my assertion that scanning most images at 16-
bits is indeed wasteful, even if you have an unlimited budget for 
harddisk space (and backup). Further, recall that opening and saving 
every 16-bit image takes twice as long as its poor 8-bit cousin, 
thereby making you "pay" for this decision over and over. Time is 
money.

For those of you who do not have good scanning software, I can see 
the reason you might be tempted to scan in 12+ bits (please don't 
call it 16-bits as few of you have such equipment). So for y'all 
there are two choices: get better scanning software that directly 
drives the scanner at its rated hardware level, or scan at "12+ 
bits" then create an optimal curve in Photoshop and "save as" in 8-
bits (thereby allowing you to archive your 16-bit raw scan for the 
future.) Once you have an optimal image, you have eliminated the 
need for any "drastic" changes that would cause any banding or noise.

I will also remind you that the need for 16-bit scanned archives is 
waaaay over-rated. Todays scanners will be out-classed by tomorrow's 
scanners (just as yesterdays scanners pale in comparison to todays). 
You can count on that tomorrow's hard disks will be bigger and 
cheaper; Note that your negative is already in an archival form; As 
such, you can always go back and scan it at a later date with 
better/faster/cheaper equipment in the future. Now THIS is good 
advice.

MJS

Clayton, while I fully agree with your second sentence (postrized or
not....), I do think BO has some relevance to the discussion. My
experience is that BO printing a "normal" neg typically requires minor
adjustments, thereby 8 bit editing has a bearing. 

Bob Michaels

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Clayton Jones"
<cj@c...> wrote:

> Aside from the other points of view expressed here, IMO the BO part of
> this thread is irrelevent.  It's the image that is either posterized
> or not as a result of 8 vs 16 bit editing.  
> 
> Regards,
> Clayton
> 
> 
> Info on black and white digital printing at    
> http://www.cjcom.net/digiprnarts.htm