FR: Ability to shift color eq nodes about darktable HOT 8 CLOSED

If nothing else, one could rotate a pixel's hue before applying the adjustments, then rotate it back.... ISTR that option having been mentioned in previous discussions.

from darktable.

Note that we also want to keep compatibility with hardware like Loopdeck which comes with 8 controls for the colors.

from darktable.

So presumably as long as they're allocated to the nodes rather than the colours those nodes represent that should be do-able (since we would be keeping the fixed number of nodes but just shifting their position)?

from darktable.

Unfortunately, this would require us to come up with another 16 colour names! Would probably be ok to do 22.5 degree shifts instead if we wanted to

If there was only one shift that would be just 8 more names which wouldn't be too difficult. For example, below are colour names, with their hue angle in brackets (HSL) according to colorabout.com (there are many variations there, I took the most likely one). I know color eq doesn't use HSL, but it's good enough to collect ideas.

Group A (current)

1. Red (0), 2. orange (30), 3. lime (75), 4. green (120), 5. cyan (180), 6. azure (210), 7. Violet (270)/blue-violet (271), 8. magenta (300)

Group B (shift)

1. Red-orange (16), 2. yellow (60), 3. chartreuse (90), 4. Spring green (150)/aquamarine (160)/turquoise (174)/teal (180), 5. cerulean (196)/sky blue (197)/celeste (205), 6. blue (240), 7. lavender (275)/purple (275)/mauve (276), 8. pink (350)

Goal being for the hue angles of colours in group B to be between those of group A.

Converting hue values from color eq space to hsl/hsv/etc would give exact values/names, if that was necessary.

from darktable.

it might sound trivial but since the sliders aren’t independent and there might be issues properly blending the changes.
that was discussed already in the original wip pr - and I doubt, mathematics changed that much ;)

from darktable.

I thought the limitation was that the nodes had to be equally spaced by the same amount. In my head I thought/assumed that the maths would be the same regardless of where you chose to be the "zero point" (for want of a better term)

from darktable.

Let's see ...

from darktable.

Go and test #16279

from darktable.