Tweaking the colors "as you see them" (e.g, with which you enter the module) is from a signal processing point of view "illegal"; color balancing needs to be performed on linear data (modifying coloring of stretched data aside from perhaps saturation doesn't make sense from a signal/photographic point of view). The Tracking engine applies your color balancing to the linear data and then forward propagates + recomposites the result so it is all mathematically (and photographically) correct.
Okay, that really helped me understanding your point of view and approach and why you don't want this to happen. I will have a look at your suggestions and see what works best for me. Additionally, I will check the recommended DSS settings (thanks for the link, I've already been aware of them - but maybe I changed something during a session and messed things up) and will try to dither harder. Maybe this helps getting rid of the correlated noise.
Thanks for investing your time into this, Ivo. I really appreciate it! Consider this "ticket" closed
admin wrote: ↑Thu Nov 05, 2020 11:28 pm
Please do keep me updated, also on the dithering/noise signature. I'm keen to figure out what's going on here...
After checking all parameters and adjusting them to the recommended settings for StarTools I came up with this stack:
