Open age1988 opened 2 years ago
Since there seems to be debate on whether we should aim for pleasing results or theoretically correct results, this needs to be settled first imo. I'm completely ambivalent on the issue atm.
I'm interested in knowing why this works. If we know why, then we don't have to worry about this being purely anecdotal.
Demosaicing in gamma corrected space instead of linear gamma.
Smoother gradients and it fixes some issues at pixel levels, mostly black dots and white dots.
[https://discuss.pixls.us/t/port-the-lmmse-gamma-step-to-the-other-demosaicers/28233]()