Closed szali closed 3 years ago
Acknowledged. It isn't necessarily supposed to be an exact match, but I don't think there's any reason for the numbers to be different in this case.
The pal8*.png images should be fixed by commit 8e9503e6eea2.
There might still be a few other images with the same type of problem -- I haven't reviewed them all yet.
Are you still interested in fixing these?
rgb16-3103.png
and rgb32-7187.png
also have low-bit differences compared to my expected output.
I'm comparing against Chromium's output, and I will defer to MS Paint if it renders differently. I don't have a Windows machine handy to verify against. (But I suspect they probably agree.)
I don't know. It'd be nice to review them to see if they're off by more than I'd want, but it's a lot of trouble for something so minor.
For channels that are not exactly 1, 2, 4, or 8 bits, some rounding or dithering is necessary when you convert them to 8 bits. Not everything is going to do that the same way.
When opening the generated g/pal8.bmp and the reference html/pal8.png (both with MS Paint which is quite an authority in BMP), and taking a screenshot of the magnified files to compare pixel values in an editor, some pixels have different green levels (e.g. pixel (7, 63) has green=42 in the png and green=43 in the bmp).