Commit 3290711d9cb8ccf184855f06fa558310cf4d5676
2024-06-22T17:45:31
cjpeg: Only support 8-bit precision w/ GIF input Creating 12-bit-per-sample JPEG images from GIF input images was a useful testing feature when the data precision was a compile-time setting. However, now that the data precision is a runtime setting, it doesn't make sense for cjpeg to allow data precisions other than 8-bit with GIF input images. GIF images are limited to 256 colors from a palette of 8-bit-per-component RGB values, so they cannot take advantage of the additional gamut afforded by higher data precisions.