As many of you know I’ve been using Lr before moving to Darktable. The later is free, sure but the UI is not easy as reported lot of time on this forum. Those last months I have been less involved in Darktable because I was working on a new UI layer looking like Lr. All this is coming together quite well now, the UI is very close except for the scope that I have keep as in Darktable as more advanced (vectorscope & waveform).
The sliders are the one from Lr, yes the same and I have trained the models with a specialized AI to match as much as possible the Lr result. This is almost working now. A preference can be used to the original or lr-like UI. I do expect a release of all this end of the week. For now just a screenshot for teasing:
Thanks a lot Pascal, we all have been waiting by far too long for this significant step But - please bear in mind the environmental burdon when you train the model, energy costs can sum up quickly and darktable’s new monthly licence model is not finalized yet
That’s a great step forward! Just a minor thing: could you, with your AI, add an enhance slider, that would automatically tune the others? The low end could have the value meh, the high end wow. I suggest setting the default to wow, and then disabling/greying out the slider to avoid accidental modification.
I am by now used to the idiosyncratic UI of darktable, but this Lr UI will make darktable easier to adopt for many who would like an open source alternative. Great work!
I recommend we remove the image from the darkroom view. Instead just add big curve graph to edit the image. Also, let’s remove the d&s module. While it works really good, we don’t understand the sliders or their math. It’s unacceptable to have something working if we can’t explain it.
Did you already reveal that with version DarkTable(R) 95[TM], the import of images will be removed and replaced by an image generation AI? This will save us so much time, we don’t have to go out and do photography anymore, we can just sit and wait what the AI gives us.