Canon Eos R Input Color Profile

Firstly, I’ll start off by saying thank you to Aurelien and Boris. I’ve been able understand and utilize the vast offers of darktable thanks to your videos.

However, I had noticed recently that the color change from adjusting values in color balance rgb differed from the videos. I easily ended up with oversaturated photos with small adjustments in perceived saturation. The default color input settings are input profile = standard color matrix and working space = linear rec 2020. Changing the working space to sRGB mellowed things out. I I’m using 3.9.0+1468~g0e64b4646.

The question is, why were my photos so reactive to adjustments in color balance rgb while working in linear rec as compared to sRGB? And when/how do you know when to change the working or input space.

1 Like

I think you should never change working space. Period :).

All the modules expect linear data i assume , so you should always use some sort of linear working space. And (linear) RGB can’t hold all the color information a camera sensor can record , which means you are clamping the gamut range already at the input profile step .

So back to your original problem , what is it exactly ? Color balance RGB suddenly saturates way more then before ?
What is ‘before’? Are you comparing 3.8.1 release vs latest 3.9 dev builds ?

You don’t specify what is compared to what so i find it hard to think about possible issues.

What comes to mind is that filmic changed to a new v6 for new images if you have a dev build , and that it’s made to preserve more color. So it may very well be that other presets in color balance RGB do more than before.

You can try switching filmic back to v5 and see if it makes things better or worse.

There is also a new working method released in dev builds for color balance RGB very recently . There is a toggle somewhere to make it work in the legacy jzazbz mode or the newer ucs22 mode .

The 'add basic ‘colorfulness’ preset also got changed in 3 updated presets , and yes, i find the older legacy one now often does too much saturation, where it did nice things with filmic v5.

If you are comparing two other versions or two other things all together , please be more clear in describing your problem :).

Thanks for the reply. Yes. apologies for the lack of reference. 3.8.1 would be the before. I’ve always shared your opinion on workings space. Mainly out of lack of knowledge on the subject. Initially, I had suspected the new version of filmic may be causing the increased reactiveness to small percentages of increases in saturation. However, after changing the filmic version to v5 I didn’t see much of a change. This led me to explore other possibilities. I’ve also wondered if it could be from other variables such as the ND filter that I was using.

Either way. I’ll go ahead and admit I probably took to the forum to discuss prematurely.

They have been updated and include a couple versions the legacy one and a skin preset

RGB colorbalance has also been reworked…perceptual saturation is using a new colorspace AP was working on…might be what you are noticing

I’ll keep that in mind and look for future updates. Thanks.

No no no no … A forum is to discuss things. Please do so :).

But the more info you give , the more people can help you. That’s all.

I don’t use a canon (or eos r) so i don’t know if anything changed with the parameters of that camera.

Like i said , i do know that I’m the 3.9 branch filmic started to behave more to 'try to preserve chroma even in the highlights ’ and color balance RGB has a brand new color model . Both seem like good candidates for saturation changes compared to what you are used to.

But there can be many more :).