- Kaunas, Lithuania
- https://aras-p.info/
- Joined on
2022-01-20 -
-
@lichtwerk right. I could actually reproduce this locally too, but I have to make the scope window really small on screen. Will prepare PR for 4.1 asap!
@lichtwerk ok, instead of that line, does adding this help then?
else if (sseq->mainb == SEQ_DRAW_IMG_VECTORSCOPE) {
scope_image = scopes->vector_ibuf;
// ------ START NEW…@lichtwerk yeah I guess (though the original fix is also puzzling as to why that would fix things in the 1st place). A bit hard to fix it without being able to reproduce locally. As a wild guess,…
Just so that I understand what is going on:
Is is that today this would happen: 0. exr files are stored in linear space,
- sequencer loads them, converts from float linear -> float srgb…
@lichtwerk the bug.blend from the linked issue does not include the actual image it uses, does the image matter? I can't reproduce the issue here (I'm on nvidia, but on windows).
@pablovazquez ok, implemented this:
- strip border is 4px wide,
- when selected or active, the selection outline is 2px within the border (i.e. there's 1px of border around the selected outline…
Nice. This might make bilinear (in VSE) be a tiny bit faster too, since now there's less work done for each pixel.
Yes, some OBJ tests fail when SIMD is disabled. I think the proper way ahead would be to make non-SIMD sRGB conversion path use exact same pow() approximation that SIMD path uses. The SIMD one…
Originally I did not change this because I was afraid of what would be implications for other pieces of code that use bilinear sampling functions.
Now, if this is done for the float code, it…