Thanks for the suggestion
Nothing analogue on this edit. But I think I failed with my attempts anyway…
20230824_0182.RW2.xmp (23,7 KB)
In RawTherapee at the original challenge:
RT: Auto Levels; Fujifilm Provia Simulation.
GIMP: downsized 25%.
G’MIC: added Kodak Tmax 400 grain.
20230824_0182.RW2.jpg.out.pp3 (11.9 KB)
Very late answer but here is my version, using my film simulation tool. Zoom in for the grain ![]()
I am doing a first edit in darktable keeping linear rgb values, without filmic or sigmoid. Here is the xmp 20230824_0182.RW2.xmp (7.5 KB). I export in png 16bit (sRGB) to retain the large dynamic range of linear images.
Yes definitely, I am planning to make a post showcasing what I am doing and releasing the code repo. It is very experimental and by no mean perfect math- and code-wise. I am really just playing around with brutally minimal models trying to reproduce some chemical magic that Kodak and Fuji developed in decades, and that is not really present in film stock datasheets. It is all Python and slow for now because it is the language that I am faster with.
very exciting!
i’d be interested to help port to a fast gpu version once you think it’s ready.
That would be awesome! I will keep you posted on the progress here on the forum, and I’ll try to converge on something. I’m not the best programmer so I am very glad about this. ![]()
Here is a simulation result of Ektar 100 at Supra Endura. IMHO it looks awesome:







