Why on screen label always show text Adobe RGB?

[radioproektor]

Why on screen label always show text Adobe RGB? My source image profile is not Adobe RGB. Is this a bug?

[thetalkingdrum]

It is the current color space selected in the export panel. This is NOT the same as the “working space”. It is the color space your file will be exported to.

[radioproektor]

Nope, In Export panel i set different ICC profile.

[seanharding]

I fixed a bug with this (#976) which is in v0.55.0. Do you still see this issue in that version?

[radioproektor]

In 0.55 bug still exists.
If Color Space set as Same as Source - it always show Adobe RGB.
If Color Space set as some specific Color Space - it show correct Color Space name.
Output ICC profile don't change anything at all.

[thetalkingdrum]

I just tested, it seems like it does not check if there is a custom output space.

[thetalkingdrum]

So a different bug: The HUD color-space label only ever reads the Color Space dropdown value — it never checks the separate custom Output ICC override, so if you've set one, the label keeps showing the dropdown text instead of the profile you actually picked.

[radioproektor]

do we need that hud info at all ? :)
By the way, Color Space is rather confusing title as for my opinion. Why don't change it to Working Space to match all other applications terminology?

[seanharding]

Though, I should say @marcinz606 (who I assume added that to the HUD in the first place) might have a different opinion or more info on why it's useful…

[thetalkingdrum]

@radioproektor I noticed that you are using your input profile as both input and output. I actually tested and it works just fine. But what's the goal with tagging the export in your own camera profile rather than a standard target? Curious what you're going for, might be worth documenting if there's a real use case.

[radioproektor]

It is Well Behaved ICC profile concept. Custom made profile made from IT8.7 chart with linear gamma and manually added D50 white point that could be used as input and working and output profile. Original article about this profile posted https://ninedegreesbelow.com/photography/well-behaved-camera-profile.html
This allow to work with TIFF image like with RAW and keep original image data unchanged from start to end.
I personally use PhotoLine as my main image editor and keep everything in PLD format. So in this scenario NegPy Flat mode could be used for intermediate processing.

Here is my simplified manual from LGG forum https://liftgammagain.com/forum/index.php?threads/home-scanning-negatives-proper-cineon-inversion.13714/page-9 :

I made a small FAQ and package based on Argyll_V2.3.1 with examples to create so called "Well Behaved" Matrix Input Profile for camera/scanner that is specially useful for negatives invert and processing. "Well Behaved" input profile with D50 white point behave and respond in same way as normal working profile, so you can pick film base, invert and pick WB directly inside original unmodified data in linear gamma.

DOWNLOAD HERE https://www.dropbox.com/scl/fo/t2k3fbu2zeu5cq0a3y7w8/h?rlkey=770gxn9x9r64qpyi0m8acegaf&st=s6j0kxq6&dl=0

Based on these articles:
https://ninedegreesbelow.com/photography/well-behaved-camera-profile.html
https://argyllcms.com/doc/scanin.html
https://argyllcms.com/doc/colprof.html

Some background info:

ArgyllCMS folder form ZIP archive should be placed on Desktop
Files and folders names should have no spaces
Don't use TIFF compression. Argyll can't read compressed TIFFs

Test_Chart.tif is a scanned IT8.7 chart image.
Reference_File.txt (or .cie) is a reference file for IT8.7 chart with measurements provided by chart manufacturer.
it8.cht is a grid guide file for patches autodetection. It is provided by Argyll for all supported color charts. You can find it in Argyll_VX.X.X/ref folder downloaded from here https://argyllcms.com/. I use it8.cht file which was extracted from RoughProfiler because its grid better optimized and modified for IT8.7 chart that i use.
diag.tif is a file generated for visual inspection only. It helps to check if all patches where autodetected in correct places.
Test_Chart.ti3 is a file generated by scanin that contains measured data. It is used by colprof to generate ICC profile.

-v Verbose mode
-p Compensate for perspective distortion
-dipnoa different options to adjust the look of diag.tif
-G1.0 Gamma of Test Chart. It helps to improve patches autodetection
-D"" Internal Profile Name
-Zp Profile default intent Perceptual
-qh Profile quality High
-am Profile type Matrix only
-u Auto adjusts Exposure of Test Chart, so white patch became 100% white

---

Prepare your Test Chart
Based on this tutorial https://ninedegreesbelow.com/photography/well-behaved-camera-profile.html#raw-process-set-white-balance
Open your Reference_File.txt in text editor and search for the most neutral LAB_A LAB_B gray patch that have LAB_L value at least 50.00. In most cases it is somewhere between GS1 to GS8 patches.
In Raw editor pick White Balance based on this patch and export Test_Chart.tif in Linear gamma with disabled color management.

Step 1
Generate .ti3 data file and diag.tif file. Type this code in Terminal:
~/Desktop/ArgyllCMS/scanin -v -dipnoa -G1.0 ~/Desktop/ArgyllCMS/Test_Chart.tif ~/Desktop/ArgyllCMS/it8.cht ~/Desktop/ArgyllCMS/Reference_File.txt ~/Desktop/ArgyllCMS/diag.tif

Step 2
Based on this tutorial https://ninedegreesbelow.com/photography/well-behaved-camera-profile.html#add-D50-white-black-color-patches
Open Test_Chart.ti3 in text editor and add additional lines with perfect synthetic D50 white and black color patches. Also NUMBER_OF_SETS should be changed from 288 to 290. So it should look like this:

NUMBER_OF_SETS 290
BEGIN_DATA
00W 96.4200 100.000 82.4910 100.000 100.000 100.000 0.000000 0.000000 0.000000
00B 0.000000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.000000 0.000000

For matrix profiles ArgyllCMS automatically pick WB from whitest and darkest patches in the chart, but in most cases whitest patch is very far from perfect white and also have huge tint compare to other darker patches. So this "hack" helps to bypass White Balance shift of IT8.7 target. Guess it should be used only for simplest matix-based profiles that don't allow to fine tune every WB patch separately. Also perfect D50 white point makes input profile behave and respond in same way as normal working profile, so you can pick film base, invert and pick WB directly inside original unmodified data in linear gamma.
Note that simple matrix-based profile produce decent calibration results only with high CRI light sources. If your light source have low CRI or use RGB LEDs with non uniform peaks, it usually needs way more complex correction than simple matrix input. Guess in this situation it is somehow possible to create additional secondary correction LUT that will adjust colors in complex way.

Step 3
Create Input Matrix ICC profile. Type this code in Terminal:
~/Desktop/ArgyllCMS/colprof -v -D"Profile Name" -Zp -qh -am -u ~/Desktop/ArgyllCMS/Test_Chart

[radioproektor]

And by the way, you don't need to keep debayered image White Balance 1 1 1. WB and any other adjustments in Linear TIFF works perfectly exactly as in RAW if WB set as in-camera. For my opinion WB 1 1 1 is totally useless way to go and only shrinks usable bit depth data for TIFF files.

[radioproektor]

Also here is a small hint chart, i created earlier, that may help to understand how to prepare in PhotoLine scanned IT8.7 chart for Well Behaved Input Profile generation (Note that it will not Photoshop like this because Levels in Photoshop don't work in Linear gamma and use internal gamma management and also can't recover negative (clipped) values in 32 bit depth):

[thetalkingdrum]

@radioproektor 1. Why "Same as Source" = Adobe RGB.

NegPy never converts your buffer into a separate "source" space and back — for the negative/print path, color comes straight from the density/paper model, not a camera→target conversion. The buffer's numbers already are the working space. "Same as Source" just means "don't retarget," and the buffer's already Adobe RGB, so that's what shows.

This is not actually a bug, but a bit confusing.

[marcinz606]

I think we can remove it from HUD completely so it doesn't create the confusion, mode stays but color space gets removed or replaced with another info. Any suggestions?

[thetalkingdrum]

Can't think of anything to replace it with atm, but I think it can be safely removed.

[radioproektor]

Maybe it is more logical to rename "Color space" to "Output ICC"
and rename "Output ICC" to something like "Printer/Paper ICC"

[marcinz606]

feat/confusion branch

[thetalkingdrum]

Great name. Looking at it now. I think it makes sense to consolidate output and color space as they were really doing the same thing.

[thetalkingdrum]

Maybe a real soft proof feature can come at some later point for printing. With the stuff that is needed for that:

Proof profile, independent of the export profile
Proof rendering intent, chosen separately from the export intent
Black point compensation
Simulate paper white / black ink (the absolute-colorimetric leg)

[marcinz606]

thanks for the idea :)

[thetalkingdrum]

That was fast:)

[radioproektor]

Seems all works well 0.56

Tootip suggest to put ICC profiles into C:\Users\username\Documents\NegPy\icc
But when i click to Folder icon it opens C:\Program Files\NegPy
Earlier i kept my custom ICC profiles in C:\Program Files\NegPy_internal\icc

Which location is correct?

Also i am confused why it says now "Same as source (Adobe RGB)"? Where that Adobe RGB came from? But any
way all works well if i set input and output to my custom ICC profile.

[marcinz606]

documents path is the correct one