New tonecurve features

Some work on tonecure in preparation:

  • added an rgb norm to RGB linked channels. The idea is that some enhancements brought by filmic can be applied on tonecurve too. In particular about chrominance control with a norm based on max(rbg). Maybe other norms more or less scientific (more or less creative ?) could be interesting too (like weighted yellow power).
  • added an LCh independent channels mode with 3 tabs L (as for Lab) plus C(L) and C(h), letting the user correct the chrominance based on luminance and/or hue.
  • added an RGB independent channels mode with 4 tabs: L, R, G & B.
    All histograms are shown in Lab color space to be consistent with the other modes, even if the tonecurve is applied on a different color space (linear prophotorbg for the RGB modes).
    Apart of that histogram issue with log scale is fixed.

Comments are welcome.

rgb norm on RGB linked channels:

LCh independent channels mode (here the histogram is related to hue while curve is applied on chrominance)

RGB independent channels mode

8 Likes

Awesome, channel independent RBG will be welcome by many!

Nice progress. I do what you are describing in G’MIC all of the time. Good to see it in dt.

Really nice!
I make heavy use of the preserve chrominance option in filmic, and it is awesome that you put something similar in the tonecurve :slight_smile:
Thanks :slight_smile:

:thinking: … you may disappointed here. Perhaps preserve the chrominance is not the right wording here. Preserve color could be more accurate.
Despite the maths are the same in filmic and in tonecurve for that option, I get different effects on my images.
Applying the norm on tone curve, beyond preserving the color, limits the increase of saturation brought by the curve transformation.
On filmic side, when I enable preserve the chrominance I get more saturated color.
I don’t know where this difference comes from.

I don’t mind the different effect on saturation :wink:
I may have expressed wrongly what I like, I like precisely the fact of keeping acurate colors :slight_smile:
So I am perfectly happy haha :wink:
Maybe the difference in saturation comes from the logarithmic transform in filmic, or to the middle grey setting?