Digital BW, The Print

Hello IJC/OPM users,
 
I was initially very thrilled at the ability to create quad tone profiles very quickly using IJC but after printing one of my favorite image for comparision with one using the IJC profile and one using Paul Roark's curve for MIS UT-FS for EEM paper I noticed that the one using Paul's curve is significantly more brilliant.
 
It's not the the IJC/OPM image is flat, it's very nicely rendered. Smooth gradients with good details in both shadow and highlights. But measurements of darkest patches show the one printed using Paul's curve has the expected high DMax of 1.67 or somewhere about there. The one rendered by OPM yields about 1.59. This is enough to show the first one has significantly more punch when they are placed side by side.
 
Can someone explain?
 
I played with IJC to push maximum ink limit, increase the black ink setting, cut off the medium gray earlier, force maximum overlap, etc. but there are no changes. The maximum dmax remained the same.
 
Thanks,
--nick


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

Nick,

I use IJC/OPM and QTR and haven't had a problem reaching Dmax with 
the papers I use.  I haven't used Paul's curves, though, so cannot 
compare directly.  Still, Dmax is Dmax, and is goverend by the paper 
and ink combination (plus any other post printing treatments used to 
enhance it).  I can achieve Dmax on my papers using OPM with my IJC 
created profiles.  

Are you able to achieve Dmax on step 26 of the 26 step linearized 
grayscale wedge when printed after creating the profile?  If so, you 
should get it with your prints.

If step 26 does not reach the Dmax of your paper, it may be the way 
you built your profile.  Too little or too much ink, strangly enough, 
can sometimes reduce Dmax, exceeding the paper's ink limit.  Or, 
adding other inks to your black at the dark end of the scale can also 
reduce Dmax.  My experience suggests that I should use as little of 
the other inks as possible as I approach jet black.  I only use as 
much as required to get the coloration required.  And near Dmax, I 
find that I have a harder time distinguishing colors anyway.

Punch can also have to do with things other than Dmax.  Another 
thread on this forum was talking about that.  If you are showing a 
lot more detail in the shadows in one print than in another, the one 
with more shadow detail will appear to have less contrast and punch, 
since there is less solid black than in the image with blocked up 
shadows.  

I'd probably print out my image with a 21 or 26 step grayscale as 
part of the image, measuring the targets to get more objective data.  
But, first you have to find out why OPM is giving you only 1.59 while 
the other method is giving you 1.67.

Part of it could be the proofing issue.  Without any proofing, I do 
find my prints print lighter than seen on my screen.  If I soft 
proof, using Roy's or someone elses procedure, I get to see 
approximately how the image will look when Printed.  I find I have to 
darken the image somewhat so my soft proof looks good, then all is 
well.  Perhaps Paul's method shows a more accurate proof??

I will say that using the same inks and paper, either method should 
be able to hit the Dmax of the paper.  So, it is probably a profile 
or workflow issue.  Hope something here triggers a solution.

Lou

Nick,

you don't mention the printer or OS you use, so the following may be 
somewhat generic troubleshooting advice :

-First you want to isolate issues that are built into the 2 drivers you are
comparing vs the profiles/curves you use with them.

- Then you may explore the profile further (though you seem to have done
a lot already). 

An efficient way to go about it is to make a half-inch square file of 100% black.
You can quickly print this on a sheet and keep repositioning it as you go. 
Keep notes to make sure you know what settings produced what print.

-  To find out if OPM produces less black than the Epson driver, you may print
through both drivers using only black ink. Using the profile you have for EEM in IJC,
turn off the other inks, then go up or down the limits while leaving the shape
of the black-ink curve as it originally was in the profile. Make several prints and
read them to see if the limits had any effect. Verify that the black ink curve is in
fact all the way up to the top right corner of the graph.
Likewise, turn the Epson driver to black only, leaving other settings 
(where possible) as you used them. Compare this to a similar print with the
Epson driver set to color and the black patch going through Paul's RGB
workflow - etc etc

You can further explore this with a 26step grayscale file (such as the one 
that I believe ships with IJC) and using a profile that has a straight diagonal 
"curve" for the black ink only (and a limit of 26), read the patches and determine 
which is darkest. This is the same as the built in Test Patches, but without the 
black squares meant to offer a visual way to determine the darkest patch. 
This way you can take actual readings and compare. You can also try the 
same exercise but with the black ink having an actual curve as it would 
in the final profile. You won't get a full scale to print, but you'll see
exactly what the black ink does in the profile.  Is the 100% patch still the same ?.... 

You pretty much get the idea of what I'm suggesting here: Try to isolate the
variables using a test file, not a real photo, and taking each component by itself.

It's possible that you'll find that the Epson driver produces a better black under
certain circumstances and there is nothing you can do with profiling. Or you may
find that OPM in fact could produce a black equal to or better than the Epson
driver if only the profile was tweaked. This may vary from printer to printer.

As has been suggested in this thread, a classic reason for loosing black density is 
too much liquid hitting the paper. This typically happens if too much of the dark 
gray (usually "cyan" position) or other inks print alongside the black. The same
can happen if the black level is too high.  

Not uncommon is the possibility that there is a difference between the dmax
you can achieve with a straight diagonal curve vs the one that really prints when
you apply the typical steep curve of the black ink. In that case you have to cheat
the limit that was determined by Test Patches (or the diagonal ink "curve"), 
but only enough to produce a "readable" improvement in density. You go any further 
and you start loosing density again because you flood the paper.

Antonis

 



--- In DigitalBlackandWhiteThePrint@yahoogroups.com, Nick Nugent 
<nick_h_nugent@y...> wrote:
...

> It's not the the IJC/OPM image is flat, it's very nicely rendered. Smooth gradients with 
good details in both shadow and highlights. But measurements of darkest patches show 
the one printed using Paul's curve has the expected high DMax of 1.67 or somewhere 
about there. The one rendered by OPM yields about 1.59. This is enough to show the first 
one has significantly more punch when they are placed side by side.
>  
> Can someone explain?

Hi Antonis,

Thank you for a very helpful reply. I just sent a reply to Joe 
Berndt of Bowhaus who had asked me for more information just like 
you did.

--- In DigitalBlackandWhiteThePrint@yahoogroups.com, "Antonis" 
<antonisphoto@y...> wrote:
> 
> Nick,
> 
> you don't mention the printer or OS you use, so the following may 
> be somewhat generic troubleshooting advice :

It's a 1160 running MIS UT-FS quadtone inkset. The paper I'm testing 
now is Photo Rag. I created this profile from scratch because I 
couldn't find anything close enough to use.

> ...
> Not uncommon is the possibility that there is a difference between 
> the dmax you can achieve with a straight diagonal curve vs the one 
> that really prints when you apply the typical steep curve of the 
> black ink. In that case you have to cheat the limit that was 
> determined by Test Patches (or the diagonal ink "curve"), 
> but only enough to produce a "readable" improvement in density. 
> You go any further and you start loosing density again because you 
> flood the paper.
> 
> Antonis

I think you may have hit the nail right on the head here. I did look 
at some profiles provided by IJC/OPM and notice that the ink limit 
for black tends to be 22 with the black ink curve stop short at 
hitting the max value of 255 according to the coordinate system of 
IJC. This may be the reason. My black ink limit tends to be set at 
24 or 25, but mostly 25 because I thought the more the merrier. But 
this according to your experience maybe the reason for my getting a 
low dmax.

I did try cutting back on the dark grey ink so it won't protrude far 
into the darkest black region, but didn't notice any improvement 
probably because I already had a black ink overload.

I will follow your advice in finding out what my black ink limit 
should be. Other than this dmax issue which I think can be resolved 
using your tip the profile created by IJC is excellent. It took me a 
while to create an initial set of curves to create a target for 
linearization.

In doing this I learned to let IJC create me a nice set of numbers 
by providing it with values for the darkest and lightest patch, then 
using these I place the peaks of each ink at appropriate location on 
the 26 zones, then change the Q or the width of the bell shaped 
curves so I got distinct wedges throughout the test strip.

I got an excellent profile alright, except I was a little naive in 
setting the ink limits. Thanks again for your valuable advice.

--nick