Problem linearizing curves

Did we take the red pill or the blue pill??

That all said, I suggest the 21x16 i1Pro 3 target and workflow.

And set your contrast intent to the following for a perfect screen-to-print match (non-linear) that does not require any softproofing in Photoshop:

Screen Shot 2022-11-28 at 11.01.15 AM

The “Calibrate and Validate” workflow.

  1. Print the master.
  2. Hair dry.
  3. Measure and paste measurements + master curve into tool
  4. Calibrate the master with the contrast tweak settings above.
  5. Save new curve and install.
  6. Print same target again.
  7. Dry and measure this new target.
  8. Paste the first calibrated curve (not master, but what you made out of it) into starting curve.
  9. Paste new measurements into measurements.
  10. Keep contrast intent same as above.
  11. Make sure green line matches red line (ish). It should be close.
  12. Go to New Curve sheet, copy and paste from that over the first calibrated curve (you have it open in Sublime Text already). This is an iterative calibration!
  13. Save and re-install.
  14. With this new final curve, print an image that is gray gamma 2.2 encoded and looks good on your screen without softproofing. Should match screen just fine if screen is around 1:200 contrast range, you have a white background around your image in photoshop, and you have a bit of light behind your monitor.

best
-Walker

Walker: OK, that’s good news, but I’m curious obviously failing to understand: should not that 256 patch chart I"m using have a patch for every RGB grayscale value from 0 to 255? When I run the eyedropper Photoshop over the patches, that is certainly not what I get…

… and thank you for the new process. I’ll give it a go right away.

Thanks once again Keith. (And to Walker for his detailed reply using a different chart altogether).
To answer your questions.

M0 is simply “measurement one” and it is indeed the lab value, and is output from using the spectro.

The numbers in columns N and 0 are the RGB values as read in Photoshop’s info window, when rolling over each patch in the column. -IF- there are supposed to be 256 separate values in that chart, well, that’s not the case.

Yes, I’d suspect that at some point that chart got corrupted, but the one I’m using is straight out of the download, and shows up as “untagged”.

All that said, I’m going to give Walker’s plan a go right now!

T

Is it possible to save the spreadsheet with those new contrast intents?

sure

file, save as (as long as you excel app is licensed)

-Walker

MO is one of the ISO 13655-defined instrument measurement modes. Mostly a color measurement, thing. We are really just needing the L* value. The modes impact the A and B values.

Check out these links…

Excellent! Thank you for the info. I was never able to find a definite answer, and just knew it gave me the L values.

Save As…

I know that probably leaves the impression I’m a newbie. In fact, I was the first independent developer ever hired by Apple Computer, back in 1978. I’ve made my living programming and working with Apple products for 44 years, now, so “save as…” was the first thing I tried.

Cutting to the chase, with the Mac Studio and Ventura, you must disable SIP to properly install the Piezo software, or so it appears. Without doing that, the permissions belong to an unknown user, and thus prevent changes being made to the files and folders.

With my original install, instead of a simple permissions error, I started getting long, convoluted error messages when trying to save (something to do with the VB macros). [Sorry, we couldn’t find /var/folders/0f/5rq5k8ys76jd4n8080_y4d080000gp/T/com.microsoft.Excel/TemporaryItems/NSIRD_Excel_fxYEJx/E7688000.MACTF. Is it possible it was moved, renamed or deleted?]

So I reinstalled both editions, and it cleared up those issues, but because of the permission problems, I had to save as to my user documents folder.

I’m going to reboot; disable SIP; reinstall the Piezo files; enable SIP again, and see if that fixes the issues.

The tool is installed into >Applications>…

So you either have to change perms on the Piezography folder (not exactly secure all the time) or save to your user folder (I suggest Save As to User folder)

best
-Walker

OK, I’m lost again.

I loaded up the “Piezography-21x16step-i1Pro3.tif” in PrintTool, and printed it.
Went to i1Profiler, and clicked on the workflow “Piezography-21x16step-i1Pro3.pwxf”

It does not look remotely like the .tif image file. So I directly loaded the .pwxf from my desktop. Same thing

So, should I use that .pwxf generated image, and print it? Or is there some other .pwxf I should try?

Meanwhile, I’ll go disable SIP and try this whole thing over from scratch…

--------- LATER-----
Nope: I was wrong. Not a SIP issue. The main user/owner for Piezography is not correct, and unknown on my machine. A get-info shows “fetching…” for the owner. The user ID for the owner is recorded as “755” (whereas it is usually “501” or “502”.) Looks like a permissions code instead…

Here’s how to fix the “fetching…” issue, from my own blog. I just did it, and the “fetching…” issue is gone.

https://valleau.art/blog/2021/05/08/permissions-fetching-the-fix/

Unfortunately, there are apparently other issues preventing me from accessing the Piezography folder, and directly saving (aka “save as”) the spreadsheet to the tools folder.

I was able to save a copy to my own documents folder, and then drag it back to Tools, however.

Could easily have been a regression in my debugging here.

Correct workflow:

Piezography-21step-i1Pro3.pwxf.zip (8.7 KB)

Yes! Thank you, Walker. That seems to match the image, and I’m going to see how the 21 x 16 pans out.

I’ll report back, and maybe we can close this thing. I know I’m ready…

BINGO!!!

Changing the target and pwxf finally got me a match with my screen! HALLELUJAH! :slight_smile:

My thanks to Walker and Keith for your patience and courtesy and for following along on this journey.

My best wishes to one and all for the holidays.

Tracy
www.valleau.art

cheers and happy TG
-W

Just to make sure I am following the plot, it was an issue related to the i1Pro 3 target and workflow, correct?

Yes, that is correct.

Probably, I will validate the bug when I have a moment.

21x16 target is most accurate IMO.

-Walker

Or perhaps simply change the Piezo Pro documentation to recommending the 21x16 i1Pro 3 target and workflow for i1Profiler users… Thanks again, Walker. I’m spending today building the new curves for the 9 high-end papers I use here.

I will so it matches the video doc (that uses 21x16)
-W