Hi,
Still working on getting the best results since the ink change.
I’m finding that printing with an ICC (the older one or the new one) result in a “big” loss of details.
Everything was fine before but I have hard time to keep my details when rgb values get under 15 (that’s was not a problem before).
Both Cool and Warm curves (blended in Print Tool) are perfectly linear.
I don’t know where to look to resolve this.
Best,
Frankie
Make sure you are printing with Relative Colorimetric Black Point Compensation if this is a new style ICC. Print with Perceptual if this is an old-style ICC.
best,
Walker
I double checked yes, everything is ok for this but it give me nearly the exact same result.
Here is the test image I’m printing.
I know this is not an “easy” one but I long as I remember I never had problem with it.
Best,
Frankie
Hi Walker,
I re-linearized Cool and Warm, blended them in Print Tool, printed a target, measured it (perfectly linear), rebuild an ICC from it, re-print a test image with the new ICC, I get the same results.
Everything below RGB=15 is so dark that I have almost no details in it on the print.
I know I could play with the curves so darkest tones prints lighter but I never had to, everything was fine before, the image didn’t changed.
Can you help me with some advices ?
Best,
Frankie
Please let me know if anything has changed in your system. Have you changed OS? etc. If linearity is verified and you are using the same ICC this is a software issue and not a hardware or liquid issue.
-Walker
As far as I can remember I did no change in the system, no OS update or anything.
Only change I did is that before I was used to work with QTR Linearize Quad for the linearization and now I use the Piezo Pro Tool but I’m getting perfectly linear curve in both case so I don’t think it’s a problem.
Do we agree that the texture of the wall (mostly the upper part) in this test image should be clearly visible on the print ? (I’m beginning to be paranoïd ahah)
EDIT : This is printed on Photo Rag 308.
So you _have_changed the process by moving to a new ICC profile system. This system has a different “transfer” curve compared to QTR. Please shared the screenshots of what your print rendering intent is for printing with the new ICC. It should be Relative Colorimetric w/ BPC.
And also please share the screenshot of the measurements you used when building the ICC.
best,
Walker
Here they are.
Note that I’m using “,” separator because Excel don’t want to read the values otherwise.
Then I copy/paste the ICC data and replace the “,” by “.” to make it work with the PiezoICC Profiler.
I see the issue. You are not measuring the correct target or you are not allowing for the proper measurement of a correct target.
There is most likely not any (even red) line in your HOME sheet. Print and measure the proper target and place in measurements tab. You will most likely see a wavy red light and will need to smooth it a bit, etc. Essentially you should be seeing what goes on in this video: https://piezography.com/about-piezography-professional/
best,
Walker
I do have a red line in the Home tab, that I smoothed a bit before exporting data.
And here is the file that I printed for building the ICC (with NO color management), is it the wrong one ? Using an i1io1 and i1Profiler for the reading.
Please PM (private message) me a .zip of your tool. I’d like to take a look and debug.
thank you,
Walker
I’m following this thread with interest since this seem like it might be relevant to my observations about the lack of deep shadow detail that I wrote about in my PiezoDN thread in the beta-tester forum using an ICC made from a linearized quad with the new dye inks.
I described it it terms of % and based on my visual assessment of a target print (not a picture) as well as a graph of the measurements, but it was very similar to this:
I am concerned about the lack of any meaningful density in the bottom 5-6% of the deep shadow steps though.
I haven’t yet tried the possible solution that we discussed, Walker.
Keith
That would be different. I don’t soft-proof at all.
Can you link me the thread you talk about @jkschreiber ?
Edit : I’m not used to soft proof either, or rarely, while working on a very difficult image.
Sure - there are actually 2 threads. The first was going off on too many tangents (all interesting in their own way) so I started a second. If you’re not fluent in “alt-process” darkroom work some of it may seem like incomprehensible babel but the digital side uses the same Piezography software so that is relevant.
https://community.inkjetmall.com/t/definition-of-a-master-quad/6547/8
https://community.inkjetmall.com/t/piezodn-dye-curves-for-pt-pd-dop/6575
This has the graph that shows Linear, PPE ICC, and QTR ICC together.
It also has my proposed solution for creating a custom-shaped ICC.
Cheers,
Keith
Thanks, I’ll give a look at it.
Is there a way to actually “see” the curve that is applied by QTR ICC and PiezoProfilerICC ?
I would like to compare them and maybe make a custom one.
Should be ok if I just print a target with the ICC and measure it, is there a simpler way ?
Frankie
Auto reply for myself after some tests to give some news.
I first assign the Gray Gamma 2.2 profile to the 256 steps target in order to have a proper managed workflow (just for the test, I know that we don’t assign the profile while creating ICC).
Here are the three test prints I did :
First one is No Color Managed = Perfectly Linear in Gray Gamma 2.2
Second is printed with a QTR ICC from a 51 steps target I did = Slighly darker in midtones, little bump in the last shadows to preserve details.
Last one is printed with an PiezoProfiler ICC from a 256 steps target = Highlight a bit lighter, midtones spot on linear, shadows completely crushed into the deep black universe.
It is clearly obvious that something is wrong in the shadows.
Is it a choice of the ICCPieezoProfiler-app developpement, trying to match something I’ve never seen on any screen : total block out in the shadows or is there a bugg or user error somewhere ?
Best,
Frankie
First one is No Color Managed = Perfectly Linear in Gray Gamma 2.2
Second is printed with a QTR ICC from a 51 steps target I did = Slighly darker in midtones, little bump in the last shadows to preserve details.
Last one is printed with an PiezoProfiler ICC from a 256 steps target = Highlight a bit lighter, midtones spot on linear, shadows completely crushed into the deep black universe.
Why are you assigning GG22? It’s already hard coded into the target. Just load the untagged target into print-tool and print the target for validation. This could be your issue right there because if the incorrect target was printed for the ICC, than the ICC is unable to produce the correct data for your print environment.
Wait, so you assigned when testing with the print icc but not when printing the target?
Ok. Hmm.
-Walker
