I finally got this to work, so I thought I’d share. The trick is to print and measure the 21x16step Piezography chart, not the 256-step chart. I’m using i1Profiler with an i1Pro 2. I saved the measurements to a CGATS text file with CIELab data, then dragged that file onto QTR-Create-ICC-RGB, then dragged the resulting ICC file into /Library/ColorSync/Profiles.
It was easy to create a grayscale ICC print profile using the Piezography-ICC-Profiler droplet as shown in the Pro manual’s video. That ICC works great when used with Print-Tool, but has problems with softproofing. If you try to Simulate Ink and Paper in Lightroom or Photoshop (2025 versions) with this profile it gets extremely dark / black, as described in this post. The suggested fix in that thread is to use QTR-Create-ICC-RGB instead, which makes an RGB rather than gray profile. It kept failing with 256-step data, saying my L values were out of order, even after I corrected them with the PiezoPro spreadsheet. I finally found a post where someone said this works with a 21-step target and indeed it does – verified with 21-step and 21x16-step.
Now I can print with the gray ICC created by the PiezoPro droplet, and softproof with the RGB ICC created by the QTR droplet. I get a very good match with a calibrated monitor and D50 viewing station at similar brightness in a dim room.
The ICC that comes with the Piezography installation called “Piezography Gloss Neutral Softproof” is a pretty good match to what I’m seeing on my PiezoPro neutral prints on Canson Baryta Photographique II, using a quad I re-linearized with the Pro tools. But the hue is off slightly, so I wanted to try making my own softproof ICC.
Previously I have always printed flat (no ICC in Print-Tool, just using re-linearized quads). But I’m liking the flexibility the PiezoPro tools give me to apply a contrast curve tweak via a custom ICC at print time.
Gosh… I’ve was close to giving up, but then I found your posting. Your experience was exactly the same as mine. Your steps were absolutely identical to what I tried!
So, it sounds like you are printing with a profile and a curve. (I assume you are printing with Print-Tools managed.) I want to be sure I got the right… Under Print Color Management, you specify Print-Tool managed and specify the PiezoPro ICC generated profile. When it comes to the Quad-Tone-RIP dialogue… you put the curve for the particular paper… then print.
Is that correct?
A million thanks! A big thank you for documenting what you found…
Yes, what you described for the printing process is correct, sorry I left out a few critical details.
In Print-Tool, under Color Management choose Print-Tool Managed rather than Unmanaged, select the print ICC you created, and set Rendering Intent to Perceptual, which automatically includes Black Point Compensation. You don’t need to do anything with the softproof settings in Print-Tool. Then when you get to the Quad-Tone-RIP dialog you select the correct paper profile, either a built-in Piezo paper profile or one you have calibrated yourself. I happen to use a calibrated profile for my paper, but using the print ICC makes this less important.
Currently I don’t need to add any extra contrast curve tweaks into my print ICC to get a very good match, sorry if that was unclear. I was just saying I like that the PiezoPro tools have that capability if it ever becomes necessary. Search for “contrast tweaking” in the Piezo Pro manual online, it’s done with some settings in the spreadsheet tool. As discussed somewhere in the PiezoPro documentation (maybe a video), one reason for making a special print ICC with a contrast curve tweak might be if you have many images to print that were adjusted by somebody else in a differently-calibrated environment – could be much faster to just apply a single correction curve to them all at print time via a custom ICC, rather than re-adjusting each image individually in Photoshop or Lightroom. I thought I would need to do that with a batch of images, but it ended up not being necessary.
Good luck!
No issue, I just wanted to be sure… thanks!
I’m trying to research why we can’t generate a 256 point profile. I printed out a 256 point target from both the color inks (Cone inks for color) and for the piezo. I then collected spectro data for both, When I compare them, I see some large variance in both the a and b variables. That’s why I think we are failing with 256-step data, They use all three variables, not just the L*. I think when we figure out what is going on there, that may solve the root of our problem Just a hunch… but it’s worth pursuing…
1 Like