Digital BW, The Print

I posted this on photo.net, too, but the chances of an answer are much
better here.

I'm having a devil of a time getting a decent negative out of an Epson
4000.

I make a 101 square "step wedge" as part of my curve building process.
My first negative was on Pictorico OHP film, with the printer set for
Epson Premium Glossy Photo Paper (same way I set the 2200 and the
7600). The step wedge is printer with color management turned off on
the printer driver. I'm getting a density crossover in the darkest
parts of the negative, from 75% to 100%. The curve increases in
density, properly, from 0% to 70% black. At 75%, it levels off, at
80%, density decreases until it's back to the same level it was at
70%, then it begins increaseing from 85% to 100%.

I tried a red negative (this is for multigrade paper, sensitive to
green and blue light, red blocks green and blue) and got a monotonic
increse in density from 0% to 100% (as it should be) but the negative
density range from 0% to 100% isn't enough to get a full DMIN to DMAX
range out of the Ilford multigrade, even at contrast grade 4. I don't
even want to be at 4, the high contrast makes the digital negatives
look "gritty". On the 2200, I have enough contrast range (using more
conventional black negatives, not red) to get a full DMIN to DMAX with
a curve I built for contrast grade 1.5. That's butter smooth, with no
grit.

Anyone have an idea for the next step? Is there another media setting
that eliminates the crossovers (I don't want to make 12 test negatives
and print them all trying to find out).

QTR really isn't an option, the Epson drivers have a dither that makes
a much better digital negative.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "koloshor"
<wiz@n...> wrote:
 
> I make a 101 square "step wedge" as part of my curve building process.
> My first negative was on Pictorico OHP film, with the printer set for
> Epson Premium Glossy Photo Paper (same way I set the 2200 and the
> 7600). The step wedge is printer with color management turned off on
> the printer driver. I'm getting a density crossover in the darkest
> parts of the negative, from 75% to 100%. The curve increases in
> density, properly, from 0% to 70% black. At 75%, it levels off, at
> 80%, density decreases until it's back to the same level it was at
> 70%, then it begins increaseing from 85% to 100%.

Let me add a couple of things:

First, it is definitely the way the Epson 4000 is dealing with black
generation. I'm used to making digital negatives on the Epson 2200,
7600, and several other machines. I've built curves for a lot of
different silver and alternative process (Pt, Pt/Pd, gum, cyanotype,
Azo). The 4000 is an entirely different animal.

It appears to have a lot to do with differences between the reflective
and transmissive density of the Epson PK, light black, and whatever
colors are mixed in by the Epson drivers.

The crossover problem extends to every paper type that I've tried, so far.

I'm beginning to think that the answer is a RIP. Has anyone ever made
digital negatives using one? Our experiments with QTR show that the
dither pattern is just too obvious for digital negatives. Our Onyx RIP
also has an ugly dither. Has anyone tried ImagePrint on digital negatives?

Thanks for any help.

Joe

Hi

I probably can't offer you much help. But, I've been messing with making digital 
negatives with a 2200 for silver printing for a couple of years now. And I have encountered 
enough pitfalls that I've already quit several times. So, I have some comments.

With regard to the density crossover you are seeing. First make sure that the reversal isn't 
related to the positioning of the steps on the printed negative. I have found that I get 
significant variation in ink density with location on the page. It's possible that the 
arrangement of the steps on your 101 step-wedge is such that a an area on the page of 
improper ink densities makes it look like a crossover. For example, I am currently using a 
129-stepwedge that goes from the highest ink density step in the upper right corner of 
the 8x10 page. The steps reduce in density in twelve steps from right to left with 12 steps 
per row and proceeds in 11 rows to lowest density near the bottom center of the page. My 
printer apparently produces a vertical band of improperly reduced densities (a dip in 
density of about 0.04 at the band center). The band is about 2-3 inches wide and affects 
the medium to low densities the most. The result is I get a plot of densities with a very 
patterned oscillation. And a different arrange of the steps in the wedge could lead to some 
very wrong conclusions.

The above observations were made when I was workiing with color or spectral density 
negatives on 
Pictorico Glossy Film paper. The color I was using was R=255, B=G=0. About a year ago 
I also tried Glossy Film with Black ink negatives. I quit after finding an even more 
pronouced ink density variation (about .08 density). I didn't pursue the problem but as I 
remember the density seemed to vary from corner to corner in both x and y. I have since 
read that the Epson black ink is a real trouble maker, so I decided to try the spectral 
density approach (again) to avoid the use of the black ink.

After all that, I do have a recommendation.

 --- If you are going to do silver prints only, try the glossy film paper with spectral density 
negatives. I have found the glossy film apparently takes the ink more smoothly than OHP 
and the spectral density negative (at least with the colors I am using) yields a much more 
well behaved adjustment curve. I think that "Adjustment Curve" is Dan Burkholder's 
terminology. Mark Nelson uses "Process adjustment Curve".

I have tried QTR as well as the Bowhaus OPM rips and came to the same conclusion that 
you did.

I would certainly appreciate hearing from you about the success or failure of your use of 
the 4000. I have been considering replacing my 2200 to solve the ink density variation 
problem, but have been reluctant not knowing if the 4000 might have the same problem.

Doug Nielsen




--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "koloshor" <wiz@n...> wrote:

> 
> I posted this on photo.net, too, but the chances of an answer are much
> better here.
> 
> I'm having a devil of a time getting a decent negative out of an Epson
> 4000.
> 
> I make a 101 square "step wedge" as part of my curve building process.
> My first negative was on Pictorico OHP film, with the printer set for
> Epson Premium Glossy Photo Paper (same way I set the 2200 and the
> 7600). The step wedge is printer with color management turned off on
> the printer driver. I'm getting a density crossover in the darkest
> parts of the negative, from 75% to 100%. The curve increases in
> density, properly, from 0% to 70% black. At 75%, it levels off, at
> 80%, density decreases until it's back to the same level it was at
> 70%, then it begins increaseing from 85% to 100%.
> 
> I tried a red negative (this is for multigrade paper, sensitive to
> green and blue light, red blocks green and blue) and got a monotonic
> increse in density from 0% to 100% (as it should be) but the negative
> density range from 0% to 100% isn't enough to get a full DMIN to DMAX
> range out of the Ilford multigrade, even at contrast grade 4. I don't
> even want to be at 4, the high contrast makes the digital negatives
> look "gritty". On the 2200, I have enough contrast range (using more
> conventional black negatives, not red) to get a full DMIN to DMAX with
> a curve I built for contrast grade 1.5. That's butter smooth, with no
> grit.
> 
> Anyone have an idea for the next step? Is there another media setting
> that eliminates the crossovers (I don't want to make 12 test negatives
> and print them all trying to find out).
> 
> QTR really isn't an option, the Epson drivers have a dither that makes
> a much better digital negative.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Doug Nielsen" <dougnielsen@c...> wrote:
> 
> Hi
> 
> I probably can't offer you much help. But, I've been messing with making digital 
> negatives with a 2200 for silver printing for a couple of years now. And I have encountered 
> enough pitfalls that I've already quit several times. So, I have some comments.
> 
> With regard to the density crossover you are seeing. First make sure that the reversal isn't 
> related to the positioning of the steps on the printed negative.

It's not. It shows up when I do a 101 step chart (10x10 with one extra square) in linear order, or a 300 step chart, with 100 values repeating, each repeated on the page three times, in random order, so it can be read by the EyeOne.

>I have found that I get 
> significant variation in ink density with location on the page. It's possible that the 
> arrangement of the steps on your 101 step-wedge is such that a an area on the page of 
> improper ink densities makes it look like a crossover. For example, I am currently using a 
> 129-stepwedge that goes from the highest ink density step in the upper right corner of 
> the 8x10 page. The steps reduce in density in twelve steps from right to left with 12 steps 
> per row and proceeds in 11 rows to lowest density near the bottom center of the page. My 
> printer apparently produces a vertical band of improperly reduced densities (a dip in 
> density of about 0.04 at the band center). The band is about 2-3 inches wide and affects 
> the medium to low densities the most. The result is I get a plot of densities with a very 
> patterned oscillation. And a different arrange of the steps in the wedge could lead to some 
> very wrong conclusions.

That's the beauty of the EyeOne and a nultiple pattern where each step appears multiple times. If the 23% step on the upper right corner doesn't match the 23% stel near the left side of the middle of the page, you get an instant flag that something is wrong. That's how I learned to even out my contact print light source, and turn the frame several times while printing.
 
> The above observations were made when I was workiing with color or spectral density 
> negatives on 
> Pictorico Glossy Film paper. The color I was using was R=255, B=G=0. 

That's sort of what I'm trying right now. Except I'm using a reddish black. I do the mix in PhotoShop by going to "levels" and raising the lower red limot from 0 to 128 (for 50% red), or 96 (for 37% red). You can mix red and black pretty freely for Ilford (oe Kodak, for that matter) multigrade. Red has both magenta (blocks green) and yellow (blocks blue). 

> About a year ago 
> I also tried Glossy Film with Black ink negatives. I quit after finding an even more 
> pronouced ink density variation (about .08 density). I didn't pursue the problem but as I 
> remember the density seemed to vary from corner to corner in both x and y. I have since 
> read that the Epson black ink is a real trouble maker, so I decided to try the spectral 
> density approach (again) to avoid the use of the black ink.
> 
> After all that, I do have a recommendation.

I should have one soon. I've tried a few more things, and have gotten some success...
 
>  --- If you are going to do silver prints only, try the glossy film paper with spectral density 
> negatives. I have found the glossy film apparently takes the ink more smoothly than OHP 
> and the spectral density negative (at least with the colors I am using) yields a much more 
> well behaved adjustment curve. I think that "Adjustment Curve" is Dan Burkholder's 
> terminology. Mark Nelson uses "Process adjustment Curve".

I just use the term "curve". I'm a bit weird, I make a curve that does the linearization and converts the image to a negative. I don't want to see the "corrected" positive image, and two conversions are risky in PhotoShop, they make the final result gamma dependent.

> I have tried QTR as well as the Bowhaus OPM rips and came to the same conclusion that 
> you did.
> 
> I would certainly appreciate hearing from you about the success or failure of your use of 
> the 4000. I have been considering replacing my 2200 to solve the ink density variation 
> problem, but have been reluctant not knowing if the 4000 might have the same problem.

I don't see any ink density variation on the 4000, or my 2200. Do you always print in unidirectional mode, and align the printer for the OHP or white film you're using?

I'm one big step closer to a solution on the 4000. I printed paper negatives (Kirkland glossy) for four paper settings on the 4000. And I learned that the inl kaydown is dramatically different on the ones I tested:

Premium Glossy Photo Paper
Premium Glossy Photo Paper 250
Premium Luster
Premium Luster 250

My first curves were done on the Premium Glossy (not 250) setting. I've discovered that even the "250" and "not 250" versions of Premium Glossy Photo Paper vary dramatically from each other. 

Luster 250 seems to be the most predictable. No actual crossover, but a big flat spot on the curve that is really annoying, when printing a black negative (like the 2200, a black negative on the 4000 has greenisn mid tones). Wien using the 37.5% red mix, the negatives are ugly brown, but fit the tonal range of Ilford Multigrade pretty well.

The problem with the 4000 is that it has about 40 basic paper types, while the 2200 had 6. So checking them all out is going to take some time.

Ciao!

Joe

Hey

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "koloshor" <wiz@n...> wrote:
> 
> I don't see any ink density variation on the 4000, or my 2200. Do you always print in 
unidirectional mode, and align the printer for the OHP or white film you're using?

I am a bit red-faced. Even before I saw your comment, I pretty conclusively discovered that 
my current problem is a non-ununiform light source for exposing the print - not a printer 
output density problem. I am using an enlatger with a cold-lite head for exposure. And it 
appears that I am getting a hot spot that led me to my bad conclusion. A new lite source is 
next.

That said - the hot spot explains my current problems using a "red" (cyan and magenta 
inks) negative. The corner to corner printer density variation that I observed a year ago 
using black ink (Pk) still looks real. I didn't own a densitometer at the time, but now that I 
do, I have been measuring everything that isn't nailed down. I still have one of the test 
prints that I did a year ago that I have just recently measured. It is a direct 2200 output of 
a complex gray scale pattern on Epson Premium Semi-Gloss. And the corner to corner 
density variation is as much as 0.08.

I am going to do some more testing, but my current conclusion remains that the Epson UC 
black ink is problematic. Mark Nelson in his Precision Digital Negative book states that 
that is the case. However, I don't think he provides any direct evidence to support that 
conclusion.



>  I'm using a reddish black. I do the mix in PhotoShop by going to "levels" and raising the 
lower red limot from 0 to 128 (for 50% red), or 96 (for 37% red). You can mix red and 
black pretty freely for Ilford (oe Kodak, for that matter) multigrade. Red has both magenta 
(blocks green) and yellow (blocks blue). 

I am quite pleased with the results I have been getting with a basically red (R=255, G=B=0 
at dmax) negative color. The way I achieve applying the red is to create a new layer, fill it 
with R=255 and then setting the layer mode to screen. 

On Kodak Polymax/Polycontrast paper I have no problem producing near-dmax while still 
maintaining near-paperwhite. And, as I said before, the required adjustment curve for 
liearization is much more gentle (easily and accurately produced with the 16 points 
available in PS-curves). That wasn't the case when I used black ink and gray scale 
negatives.

> 
> I'm one big step closer to a solution on the 4000.
> 
> The problem with the 4000 is that it has about 40 basic paper types, while the 2200 had  
6. So checking them all out is going to take some time.

Please keep us posted on your 4000 results.

Doug