3 ms·
AVIF is worse than JXL both for high fidelity compression and for lossless compression, which is what Apple is using JXL for here.
by lonjil 2y ago
AVIF is worse than JXL both for high fidelity compression and for lossless compression, which is what Apple is using JXL for here.
- trompetenaccoun 2y ago"Worse" in what sense? There are of course always trade-offs when compressing. In terms of size though even the reviewer from the above video - who says he much prefers JPEG XL - claims his AVIF compression lead to a very significant reduction in file size when testing his family photos. At a loss of detail that's negligible, or basically invisible for practical every-day purposes. This seems to track with other comparisons I've read as well. Some claim even larger size reductions, depending on the source files. The main issue with AVIF seems to be slow loading, which granted is a real problem for web use in places with slow connections (as is file size!). Like implied I'm not an expert, I'm just wondering why AVIF can't have faster loading with a preview function like JPEG has. Is there a technical reason?
- lonjil 2y ago> "Worse" in what sense? Worse as in not compressing as well. As I said, at high fidelity or lossless. At lower quality levels, AVIF files are smaller and JXL files are bigger. At higher quality levels, JXL files are smaller and AVIF files are bigger. At lossless, AVIF is worse than PNG, while JXL is much better. > Like implied I'm not an expert, I'm just wondering why AVIF can't have faster loading with a preview function like JPEG has. JPEG is fast to load regardless of its progressive loading feature. So there are really two things at play, speed of decoding, and whether partially loaded files can give you a preview. The slow decoding of AVIF is inherent to the design of the format, but different encoding choices can lead to different decoding speed. IIRC, the lower the encoding quality, the faster the decoding can be. And as for progressive display, that simply wasn't considered when AV1 was designed. You can do something funny and store two AVIF frames, one low quality and maybe lower resolution, followed by the full image, but then you're storing two images just to get a preview.
- TD-Linux 2y agoYou don't need to store two whole independent images. The high quality image can predict from the low quality image, and the low quality image can be a lower resolution, too. It is less efficient than storing one image, but more efficient than storing two independent images.