Testing the coverage, density and accuracy of the colour dictionary we’re building for the GH2 colour picker. Only greys, magentas, reds, oranges and yellows have so far been done, so the blues and greens are covered by large regions of adjacent colour categories.
This is just a test, but it does nicely show the progressive texture rendering for the sphere.
Nice. One time I make a color picker and I was proud of it. I share here the idea to you and Rhino so that you can pick from. Just in case.
https://connect.unity.com/p/ui-color-picker-eyedropper-tool
In 2015 I was trying to combine all my color selection ideas in a simple to understand UI tool
That was at the same time traditional.
Horizontally there is a saturation color slider
Vertical is the intensity or brightness
Alpha is on the bottom
on the right you can compare actual color selection with last starting color.
Is much faster for selecting the color than other tools.
The hue circle looks too smooth for a standard HSL/HSV progression. Did you create a custom equation to map angle to hue, or am I seeing things?
The intention was to maintain the same hue luminosity. The circle is a custom texture made in photoshop. Is working and I do not notice color shifting.The user can choose the color also far away from the center.
I do not remember now exactly the code. So I presume yes. I was using C# and some help from the Unity UI. Create a custom equation to map angle to hue and then via screenshot past that to photoshop. In photoshop I make it darker in the blue-green transition sector. So is not an accurate traditional hue but works better for the artist.
The horizontal saturation is made with two overlapping images. The pure color and the other constant gray + alpha to make a fade gray across it.
It was a working prototype and need calibration and improvement. For example, the vertical slider background needs improvement. But shows up the idea.
For my H*SP colour sphere model I created a hue adjustment equation (along with its inverse):
I also picked a better luminance metric, although there is no simple equation for that. So the inverse has to be solved using a binary search. The graph below is somewhat non-nonsensical, but you can see how my own Hue+Lum behaves along the bottom vs. the standard HSL/HSV along the top. It sadly does tend to rather dull out the yellows and blues, as no doubt you noticed with your model as well.
Looks like yours is more correct and my was made manually. The only thing I was able to find was this pallet:
_Gradient Ring Mask.rar (2.0 MB)
In my arc, right now, I use the color-picker eyedropper itself to check if hue is shifted. And it is.For example orange-yellow and magenta sectors.
In your case the bottom gradient need more blue push. And the top gradient looks correct but need a mix of green that can be taken out interpolated from the bottom. You can make a third gradient interpolating the them using a curve?
