ECI RGB v2

[dariusz b]

Hi,

I am not sure if the RPP behaviour is correct - pardon my ignorance, but I am quite new to colour spaces.

When exporting to ECI RGB v2 (preprint), resulting files are saved with L_star_RGB_v4.icc profile. Is that correct or a glitch in RPP?

Photoshop then interprets it differently from ECI RGB v2 - when I attempt to convert to ECI rgb v2 inside Photoshop, I can see clear differences, both in lightness and colour in some areas.

Is the L star RGB v4 identical with ECI RGB v2 and if so why Photoshop has problems with it? If theses are not identical, why L*RGB used in RPP when saving for ECI RGB v2?

[RPP]

Dariusz,

L star RGB v4 is identical to ECI RGB v4. v4 stands for ICC profile format version.

This was added over 12 years ago, so I just don't remember why it has a different name, but content of profiles is the same, so they should be interpreted identically.

Can't really answer photoshop related questions - it's neither my product nor used by me :)

Regards,

Andrey

[dariusz b]

Thanks a lot for your answer!

I think I am getting better in using your progam and in fact it slowly becomes my favourite raw processor. The color accuracy and control seems to be one of the best, although it takes quite a lot of time to tweak the figures right way.

Thank you very much for your amazing software,

Dariusz

[Iliah Borg]

Dariusz,

> when I attempt to convert to ECI rgb v2 inside Photoshop, I can see clear differences, both in lightness and colour in some areas.

With conversion, that shouldn't be the case. Are you trying assigning a profile instead?

Conversion between profiles by definition preserves the look as Lab/XYZ colour values do not change.

Assigning preserves RGB / CMYK numbers, but reinterprets the the colour.

There should be explanations with examples, try googling "assign convert profile"

--
Best regards,
Iliah Borg
LibRaw, LLC
www.libraw.org
www.rawdigger.com
www.fastrawviewer.com

[dariusz b]

Hi,

yes, the visible color and lightness shifts (especially in magenta) are seen when I am using photoshop option "convert to profile", when the source profile is "L-star-RGB-v4.icc" (as exported by RPP) and when I set the target profile to be "ECI RGB v2" (installed and present by default in photoshop).

I have tried to download the ECI RGBv2 and v4 profiles and I also found "L-star-RGB-v4" profile online and added it to photoshop:  in all cases the behaviour is exactly the same - for some reason default RPP icc profile seems to be a little bit lighter and with more pinkish tones in magentas.

Was the profile maybe updated by the ECI since the first profile implementation in RPP? That would explain slight difference.

Or maybe I am doing something wrong, but I tried to follow numerous guides and with different profiles, too, to the same effect, which puzzles me a lot.

The fact that interpolation to external L*RGB and ECI v2 from ECI website gives exactly same effect made me think that maybe there was indeed slight specification change to ECI rgb v2 profile sometime during last 12 years.

Thanks,

Dariusz.

[Iliah Borg]

Dariusz,

Whatever the specification, "convert" changing the rendition is abnormal. Open the file, make a copy, try converting to a different RGB working spaces with rel. colorimetric intent and use eyedropper to compare Lab readings on the original and on the copy, if convert goes correctly, the Lab values will be the same (apart of colours where gamut clipping occurs).

> --
> You received this message because you are subscribed to the Google Groups "Raw Photo Processor" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to raw-photo-proce...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/raw-photo-processor/3789c159-c97a-4075-89c3-6473d8db2aa1o%40googlegroups.com.

[dariusz b]

Iliah,

Thank you very much for your response!

You are right, the LAB values are exactly the same before and after conversion on all points tested.

It turns out that the issue with color shift is only present in the preview mode and not affecting actual files.

I do not know why the Adobe product shows different hues (with almost exclusively magentas affected) on the preview, but thankfully it is only Adobe's preview glitch and not actual problem with profiles.

I feel so much better now, thanks for your great piece of advice!

Dariusz

> To unsubscribe from this group and stop receiving emails from it, send an email to raw-photo-processor+unsub...@googlegroups.com.