Digital BW, The Print

Hi all.
I must confess I'm a little lost here. I ordered an Epson 1400 and  
UT-14 from MIS. The possibilitie of printing in matte,luster and  
baryta without swapping carts seem great to me. I just read Jacob's   
post on http://photo3dart.com/wordpress/?p=334 and got a little  
confused.
So, visually, is Eboni 6 much smoother then UT-14.? My parameter is  
what I used to do, MIS FSN with an Epson 1160. How would you compare  
them all?
best,
Andre

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

Oh, and for the sake of clarification, one more thing: Looking at the Profiles folder of QTR, I see three sets for the 1400: CL(ok,it's for OEM Epson inks) K6(Ebony 6 I suppose) and MIS. IS the last one for the UT-14?

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Andre Vallejo da Silva <vallejodasilva@...> wrote:

>
> Hi all.
> I must confess I'm a little lost here. I ordered an Epson 1400 and  
> UT-14 from MIS. The possibilitie of printing in matte,luster and  
> baryta without swapping carts seem great to me. I just read Jacob's   
> post on http://photo3dart.com/wordpress/?p=334 and got a little  
> confused.
> So, visually, is Eboni 6 much smoother then UT-14.? My parameter is  
> what I used to do, MIS FSN with an Epson 1160. How would you compare  
> them all?
> best,
> Andre
> 
> [Non-text portions of this message have been removed]
>

"andrevallejo" <vallejodasilva@...> wrote:


> I ordered an Epson 1400 and UT-14 from MIS. 
> The possibilitie of printing in matte,luster and  
> baryta without swapping carts seem great to me. 

> I just read Jacob's   
> post on http://photo3dart.com/wordpress/?p=334 and got a little  
> confused.

> So, visually, is Eboni 6 much smoother then UT-14? 
> My parameter is what I used to do, MIS FSN with an Epson 1160. 
> How would you compare them all?

Eboni-6 has very light inks and can print more smoothly than UT14, which stresses flexibility.  Whether you'd be aware of any difference without very close inspection -- or even then -- is questionable.  Ut14 is probably smoother than your old 1160 with FSN.

> Looking at the Profiles folder of QTR, I see three sets 
> for the 1400: CL(ok,it's for OEM Epson inks) 
> K6(Ebony 6 I suppose)

Maybe not.  I did not supply QTR profiles.  Those K6 profiles may be for Jon's set.

> and MIS. IS the last one for the UT-14

I'm not sure how many people on the QTR forum have contributed profiles.

I mostly made some basic Photoshop curves that people could use with Create ICC RGB.  My main concern was to get a good variable tone inkset out there, but since I don't really use it, I didn't leave it in the 1400 for long enough to make a lot of different profiles.  

One of the design criteria was to make UT14 run well enough with the Epson driver than profiling would be very easy.  For example, Premier Art Smooth BW, with no profile, but the Epson driver set to Color Controls, gamma 2.2, prints with essentially a straight line Lab L curve -- very close to QTR's.  Photo Rag prints with somewhat compressed shadows and more of a split tone.

The PS curves are probably close enough to linearize with most papers, and that is about the extent of what I did with the inkset.

Paul
www.PaulRoark.com

Actually it's a good question. I tried to measure the difference between BO mode that prints with one black color and EB6 which prints with 6 colors.
 If you look at high resolution scans of 21-step wedge print  you'll see it as combination of white and gray spots. I tried to calculate the size of the area of all white spots and their brightness, the same I did for the gray spots. To do this I've used Photoshop's analyze tool on a cropped 1" each side rectangular. Then I calculated:
roughness = (white area size * white area RGB)/(gray area size * gray area RGB)

So I've got numbers showing how much of whites we have relative to gray.
Here are some numbers for 50% step on the 21-step wedge printed with different combinations of BO and EB6:

0%BO:100EB6  - 0.12
20%BO:80%EB6 - 0.18
40%BO:60%EB6 - 0.27
60%BO:40%EB6 - 0.3
80%BO:20%EB6 - 0.49
100%BO:0%EB6 - 0.54

The bigger the number - the rougher the ink. As you see 100%BO is 4 times less smooth than EB6. I didn't measure for UT14 but my guess it would be in the middle between BO and EB6.

Now the next question is: how rough ink should be to so the roughness would become visible for the viewer? 
I don't have answer to that, to my eyes the print made with BO looks the same quality as EB6 (sure EB6 is warmer, BO whites are whiter).
I'm using 1400, Paul mentioned that on 1800 BO prints had some banding. 

It is possible that large prints will have noise more visible with rougher ink.

It looks to me that the combination of BO and EB6 I've recommended in my article http://photo3dart.com/wordpress/?p=370 gives the best compromise between coolness of BO and smoothness of EB6.

Still I think that difference in quality between BO and EB6 is very small, between UT14 and EB6 it's even smaller.

Actually UT14 is a different ball game - it introduces colorants, so here we have to consider compromise between image tone and longevity. 
If you are worried about UT14 smoothness you can replace GLOP ink on Y position with EB6 Y ink (as Paul once recommended) and get smoothness close to EB6, if you want to print on glossy paper you can always put GLOP back.

If somebody is interested in details of measuring smoothness of the ink please let me know and I will publish it.

Jacob Mann
http://www.photo3dart.com

Thanks Paul and Jacob, as my stuff arrives I'll post my impressions...
best,
Andre

"Jacob" <flashinet@...> wrote:

[Re: smoothness measures]

> ... I tried to measure the [smoothness] difference between 
> BO mode that prints with one black color and EB6 which 
> prints with 6 colors.

> ...To do this I've used Photoshop's analyze tool ...

I used the standard deviation measure in the Photoshop Histogram tool.  The results do not always seem to be what I'd estimate from simple visual comparison, but it often seems to correlate.

I'm wondering how the correlation would be between your method and the standard deviation measure.


Paul
www.PaulRoark.com

Paul, here are the numbers:

                          Std. Deviation

0%BO:100EB6 - 0.12       10.40
20%BO:80%EB6 - 0.18      12.45 
40%BO:60%EB6 - 0.27      14.20 
60%BO:40%EB6 - 0.3       14.89
80%BO:20%EB6 - 0.49      17.29
100%BO:0%EB6 - 0.54      17.85

The numbers are correlated but my numbers show 4 times difference between BO and EB6, std dev show less then 2 times difference, to my eye 4 times looks like more correct number.

To make it crystal clear what is correct I've created an empty square image in Photoshop with gray background RGB 209, then selected 2 squares, 1/8 of full square each. Filled them with Black color RGB 38.

Now with analysis menu measurement I've got following:

  black area size  1.42  RGB 38
  gray area size    4.25  RGB 209

Smoothness: (1.42 * 38)/(4.25 * 209) = 0.0607

Std dev shows: 74.18

Then I've converted 2 squares 1/8 of the big square each into black ones. So now I have a half of square black, a half is gray.

  black area size  2.85  RGB 38
  gray area size    2.83  RGB 209

Smoothness: (2.85 * 38)/(2.83 * 209) = 0.183

std. dev shows: 85

So calculated smoothness increased 3 times which is about right, standard dev increased 10% which is not even close.

So you can't trust standard. dev, it makes inks looks too smooth relatively. 

I'm not sure what the right term should be for "smoothness", the best I can come up with is "relative specularity".

Jacob Mann
http://www.photo3dart.com

Humm...Jacob, could you please translate all this for the non-mathematical capable like me (I'm a doctor...) into some "so it means that..." form ?!
(8-)


--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Jacob" <flashinet@...> wrote:

>
> Paul, here are the numbers:
> 
>                           Std. Deviation
> 
> 0%BO:100EB6 - 0.12       10.40
> 20%BO:80%EB6 - 0.18      12.45 
> 40%BO:60%EB6 - 0.27      14.20 
> 60%BO:40%EB6 - 0.3       14.89
> 80%BO:20%EB6 - 0.49      17.29
> 100%BO:0%EB6 - 0.54      17.85
> 
> The numbers are correlated but my numbers show 4 times difference between BO and EB6, std dev show less then 2 times difference, to my eye 4 times looks like more correct number.
> 
> To make it crystal clear what is correct I've created an empty square image in Photoshop with gray background RGB 209, then selected 2 squares, 1/8 of full square each. Filled them with Black color RGB 38.
> 
> Now with analysis menu measurement I've got following:
> 
>   black area size  1.42  RGB 38
>   gray area size    4.25  RGB 209
> 
> Smoothness: (1.42 * 38)/(4.25 * 209) = 0.0607
> 
> Std dev shows: 74.18
> 
> Then I've converted 2 squares 1/8 of the big square each into black ones. So now I have a half of square black, a half is gray.
> 
>   black area size  2.85  RGB 38
>   gray area size    2.83  RGB 209
> 
> Smoothness: (2.85 * 38)/(2.83 * 209) = 0.183
> 
> std. dev shows: 85
> 
> So calculated smoothness increased 3 times which is about right, standard dev increased 10% which is not even close.
> 
> So you can't trust standard. dev, it makes inks looks too smooth relatively. 
> 
> I'm not sure what the right term should be for "smoothness", the best I can come up with is "relative specularity".
> 
> Jacob Mann
> http://www.photo3dart.com
>

Please don't worry about it, try your UT14 and enjoy it.

Please remember to remove colored plugs from the top of the cartridges.

Jacob

Fine! It's just that I'm trying to learn all this stuff about partitioning, linearization and so on, so I can extract the most out of the system, and sometimes I get stuck in all the technicality ...but I'll keep walking...

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Jacob" <flashinet@...> wrote:

>
> Please don't worry about it, try your UT14 and enjoy it.
> 
> Please remember to remove colored plugs from the top of the cartridges.
> 
> Jacob
>

The interesting part would be relating this data to people reactions to
prints, although doing that well would be an awful lot of work.  Smoothness
is just one quality that we might want in a photograph. For example, some
people really like BO printing because the white between the dots can lend a
nice sense of luminosity to the print.


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

"andrevallejo" <vallejodasilva@...> wrote:
>
> ... I'm trying to learn all this stuff about partitioning, linearization and so on, so I can extract the most out of the system, ...

With UT14 as well as most of my inksets, I recommend the following approach to learning the various printing workflows or approaches: 

First, just plug in the carts (removing the air intake plugs first) and print with the Epson driver and no other ICC or profile.  See page 2 of my PDF at http://www.paulroark.com/BW-Info/UT14.pdf for the settings.  An inexpensive matte paper like Epson Ultra Premium Presentation Paper is not a bad learning paper (but it's acidic).  I use ink densities and placements that allow quick and easy positive results, even if not perfect.

Second, to "linearize" the system, which in this case more accurately means to make it "color managed" in the sense that the print will match a calibrated monitor, try using QTR's Create ICC to make a grayscale ICC.  This involves making a 21-step test print, reading it with a spectro like the ColorVision PrintFixPro or even a flatbed scanner (see link on page 4 of the PDF).  Basically all you do is drag and drop the text file output of the spectro into the Create ICC icon on your desktop.

Third, one can control the print tones by using Photoshop curves.  Basic starting curves are in the Zip file that is linked to the PDF.  These should be close enough to be re-linearized with QTR's Create ICC-RGB.  With this approach you print a 21-step test file using the curve and then drag and drop the curve and text Lab spectro data into the Create ICC-RGB icon at the same time.

These ICCs in combination with the Epson driver make for very good prints.

The final stage for those who really want to get into the details of profiling is to use the QTR rip.  I don't get into that in detail in the PDF, but there are a number of tutorials out there both in the QTR download and on my web page.

In general, simply following this forum is the best way to learn digital B&W printing, particularly with dedicated B&W inksets.   Ask questions and you'll get good advice from a lot of different people.

Enjoy the journey.

Paul
www.PaulRoark.com

Yes I would like to do it if I knew how. We can measure smoothness of the inks but how relate it to image quality? 

There should be some impact of non-smooth inks on image quality, but I've printed on my epson 1400 number of images letter sized - predominantly black images, light images and mid-tone images - and I don't see any difference between BO and EB6 in smoothness just luminosity for BO and warmer tones for EB6.  

I don't know what to think - may be it's just me? Or my printer is too good? Or may be I need to print bigger size prints?  Anyway what I don't see and I can't come up with the way to measure. 

Jacob Mann
http://www.photo3dart.com


--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "pdesmidt tds.net" <pdesmidt@...> wrote:

>
> The interesting part would be relating this data to people reactions to
> prints, although doing that well would be an awful lot of work.  Smoothness
> is just one quality that we might want in a photograph. For example, some
> people really like BO printing because the white between the dots can lend a
> nice sense of luminosity to the print.
> 
> 
> [Non-text portions of this message have been removed]
>

Tks Paul, I surely will.All these years with the 1160/FSN I usually used your curves and that was it. Now I think about going depper with the new ombo, and I think I go the QTR way. I also experimented with PrintFixPro and the Spyder3Print, also with good results. I might try it to and compare results...I'll post them here.
Best,
André

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "pr_roark" <pr_roark@...> wrote:

>
> "andrevallejo" <vallejodasilva@> wrote:
> >
> > ... I'm trying to learn all this stuff about partitioning, linearization and so on, so I can extract the most out of the system, ...
> 
> With UT14 as well as most of my inksets, I recommend the following approach to learning the various printing workflows or approaches: 
> 
> First, just plug in the carts (removing the air intake plugs first) and print with the Epson driver and no other ICC or profile.  See page 2 of my PDF at http://www.paulroark.com/BW-Info/UT14.pdf for the settings.  An inexpensive matte paper like Epson Ultra Premium Presentation Paper is not a bad learning paper (but it's acidic).  I use ink densities and placements that allow quick and easy positive results, even if not perfect.
> 
> Second, to "linearize" the system, which in this case more accurately means to make it "color managed" in the sense that the print will match a calibrated monitor, try using QTR's Create ICC to make a grayscale ICC.  This involves making a 21-step test print, reading it with a spectro like the ColorVision PrintFixPro or even a flatbed scanner (see link on page 4 of the PDF).  Basically all you do is drag and drop the text file output of the spectro into the Create ICC icon on your desktop.
> 
> Third, one can control the print tones by using Photoshop curves.  Basic starting curves are in the Zip file that is linked to the PDF.  These should be close enough to be re-linearized with QTR's Create ICC-RGB.  With this approach you print a 21-step test file using the curve and then drag and drop the curve and text Lab spectro data into the Create ICC-RGB icon at the same time.
> 
> These ICCs in combination with the Epson driver make for very good prints.
> 
> The final stage for those who really want to get into the details of profiling is to use the QTR rip.  I don't get into that in detail in the PDF, but there are a number of tutorials out there both in the QTR download and on my web page.
> 
> In general, simply following this forum is the best way to learn digital B&W printing, particularly with dedicated B&W inksets.   Ask questions and you'll get good advice from a lot of different people.
> 
> Enjoy the journey.
> 
> Paul
> www.PaulRoark.com
>

"Jacob" <jacob@...> wrote:

... 
> There should be some impact of non-smooth inks on image quality, but I've printed on my epson 1400 number of images letter sized - predominantly black images, light images and mid-tone images - and I don't see any difference between BO and EB6 in smoothness ...


We've become so picky that we're looking at these images at very close distances and objecting to a noise (grain) level that would have been considered very smooth in the old film world.

I tried to compare the dots a 1.5 picoliter printer makes in a black only print with some Kodak examples of film grain in prints.  
See http://www.paulroark.com/BW-Info/Grain.pdf

The 1800 3-MK print would be about like a 1400 black only print.  Actually, I think the 1400 is a bit smoother.

I've included my methods in the PDF.  Maybe there is an error there.  If so, let me know.  

Paul
www.PaulRoark.com

Paul,

I see, because ink smoothness impacts image quality so little you wanted to show that it's no worse than film grain. It sounds about right, with only difference that film grain becomes bigger with enlargement, ink roughness becomes less visible with bigger size of prints, physically it's still the same size but it's less visible because viewing distance bigger for bigger prints.

So BO with 1 ink which is less warm and more luminous is fine.

Now tell me please somebody - when couple of weeks ago I was asking if dedicated b&w printer will give better image quality, I was told by many people "Yes!" and now I'm confused - why 3 black inks (K3) in my Epson 7900 were suggested to give worse image quality than 6 Eboni inks when we see that even 1 ink used in BO only mode gives comparable  with 6 inks quality?

Actually I like to play with this stuff and I know why - all these inks, bottles - I feel like in old good times - I can play with chemicals again.
Just don't tell me about better image quality, I'm still trying to make print on dedicated printer with MIS inks which will have the same quality as print from 7900, may be some day...

Jacob Mann
http://www.photo3dart.com

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Jacob" <flashinet@...> wrote:
>
>[...] Then I calculated:
> roughness = (white area size * white area RGB)/(gray area size * gray area RGB)

Just out of interest a highly academic question: what does "gray" exactly means? I assume "non-white". If so, one very light dot against a white background contributes as much to "roughness" as a black one, which goes against my intuition. But as said, I guess it is of academic relevance only...

Joost

"Jacob" <jacob@...> wrote:
>
> ...
> So BO with 1 ink which is less warm and more luminous is fine.

Black only printing can make fine prints in a 1.5 picoliter printer.  Still, however, I prefer to blend it with Eboni-6 even in the 1400 for better smoothness.  While my film grain comparison shows, I think, that compared to the old film days our expectations have become very high, that does not necessarily mean that our drive for maximum smoothness is not worthwhile.  Given the grainlessness of good digital files, why not go for a totally smooth output?  

 
> Now tell me please somebody - when couple of weeks ago I was asking if dedicated b&w printer will give better image quality, I was told by many people "Yes!" and now I'm confused - why 3 black inks (K3) in my Epson 7900 were suggested to give worse image quality than 6 Eboni inks when we see that even 1 ink used in BO only mode gives comparable  with 6 inks quality?

The 1400 black only is not as smooth as the Eboni-6.  Also, in a wide format printer the 3.5 picoliter drops do not, in my view, allow photo quality black only printing.  I wish there were a 1.5 pl 24" printer, but I gave up waiting and bought a 7800 for Eboni/Carbon-6.

As to the quality of a K3 B&W, I think they satisfy most people.  I, personally, don't want any color inks in my images.  However, if the 3 levels of black-gray were neutral and if the printer did not microband with only 3 inks, I do think a k3 system could make an excellent print.  Those are big "if's" however.

 
> Actually I like to play with this stuff and I know why - all these inks, bottles - I feel like in old good times - I can play with chemicals again.

Ah, a true darkroom worker -- there is a satisfaction in making your own.  You'll have to get into the C6 trip.  
See http://www.paulroark.com/BW-Info/Ink-Mixing.pdf

> Just don't tell me about better image quality, I'm still trying to make print on dedicated printer with MIS inks which will have the same quality as print from 7900, may be some day...

I don't have any hesitation in saying I think my 100% carbon prints are better than K3 prints in terms of image stability, including lack of metamerism.

Paul
www.PaulRoark.com

Joost,

by using RGB as number from 1 to 256 we automatically take into account color of dots, measure tool under analyze menu calculates area and mean RGB of selection, so we need only select by range of color and measure.

Jacob


-- In DigitalBlackandWhiteThePrint@yahoogroups.com, "horstenj" <j.h.j.h@...> wrote:

>
> --- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Jacob" <flashinet@> wrote:
> >
> >[...] Then I calculated:
> > roughness = (white area size * white area RGB)/(gray area size * gray area RGB)
> 
> Just out of interest a highly academic question: what does "gray" exactly means? I assume "non-white". If so, one very light dot against a white background contributes as much to "roughness" as a black one, which goes against my intuition. But as said, I guess it is of academic relevance only...
> 
> Joost
>

Now I'm happy to agree with you Paul, finally I've got nice prints using combination BO and UT14 cool and warm curves. 
Also you are probably better printer than I am, I would take lessons from you if you were in New York. 

Jacob


--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "pr_roark" <pr_roark@...> wrote:

>
> "Jacob" <jacob@> wrote:
> >
> > ...
> > So BO with 1 ink which is less warm and more luminous is fine.
> 
> Black only printing can make fine prints in a 1.5 picoliter printer.  Still, however, I prefer to blend it with Eboni-6 even in the 1400 for better smoothness.  While my film grain comparison shows, I think, that compared to the old film days our expectations have become very high, that does not necessarily mean that our drive for maximum smoothness is not worthwhile.  Given the grainlessness of good digital files, why not go for a totally smooth output?  
> 
>  
> > Now tell me please somebody - when couple of weeks ago I was asking if dedicated b&w printer will give better image quality, I was told by many people "Yes!" and now I'm confused - why 3 black inks (K3) in my Epson 7900 were suggested to give worse image quality than 6 Eboni inks when we see that even 1 ink used in BO only mode gives comparable  with 6 inks quality?
> 
> The 1400 black only is not as smooth as the Eboni-6.  Also, in a wide format printer the 3.5 picoliter drops do not, in my view, allow photo quality black only printing.  I wish there were a 1.5 pl 24" printer, but I gave up waiting and bought a 7800 for Eboni/Carbon-6.
> 
> As to the quality of a K3 B&W, I think they satisfy most people.  I, personally, don't want any color inks in my images.  However, if the 3 levels of black-gray were neutral and if the printer did not microband with only 3 inks, I do think a k3 system could make an excellent print.  Those are big "if's" however.
> 
>  
> > Actually I like to play with this stuff and I know why - all these inks, bottles - I feel like in old good times - I can play with chemicals again.
> 
> Ah, a true darkroom worker -- there is a satisfaction in making your own.  You'll have to get into the C6 trip.  
> See http://www.paulroark.com/BW-Info/Ink-Mixing.pdf
> 
> > Just don't tell me about better image quality, I'm still trying to make print on dedicated printer with MIS inks which will have the same quality as print from 7900, may be some day...
> 
> I don't have any hesitation in saying I think my 100% carbon prints are better than K3 prints in terms of image stability, including lack of metamerism.
> 
> Paul
> www.PaulRoark.com
>