They did not communicate if they were emotionally in favor or against. They did not conduct comparisons. None of their image experts participated in the discussion. Previously, when WebP was introduced, they had lively, well-justified and in general a professional approach to declining it. This time, the bugs were closed for discussion even when the community had huge interest to discuss these features and the implementation was in rapid development. Result: we don't know what they think deep inside and how/why the decision was made.
I don't really understand. From the discussion, it seems that JPEG-XL is better in every way than AVIF. Or maybe just don't add anything if they don't bring much. WebP is already a pain for no real benefit.
JPEG-XL is better for images, no doubt. However it lacks the political backing that AV1 and hence AVIF has. It’s sad from a technical standpoint but also just a reality of this world - politics decide things, not merit.
Politics trumping technical superiority being how things work should not mean that we accept politics trumping superiority. The only way this changes is with sufficient pushback. Unfortunately with web standards things have gotten so bad that I don't see any solution working out other than real anti-trust enforcement against Google.
There are plenty of use cases where image quality or coding speed cannot be compromised because of competitive reasons. For example medical imaging, professional or prosumer photography, legal evidence, printing cannot always accept small details disappearing or accept any other compression related degradations. JPEG XL, due to its better quality quarantees and great lossless options will surface in these areas first. Later, it will propagate from these uses to operating systems and eventually to browsers, too. Unfortunately slower transition puts some additional work on the web community.
I used to run ‘large’ (millions of images) photo sites before and now I run some small ones.
Once a year for the last 5 years I considered new image formats would affect my practice, last year I finally found webp was an unambiguous winner over jpeg. (The bar is high. If you want to save both storage and transfer costs you can’t keep copies in many different formats.)
I found it was not so easy to do this analysis because I have not seen good studies of quality/size/processing speed and the ones that are there are limited. I’ve seen some where people compress the kind of images you might see in the header of a blog but look close and the images are awful. I take pictures with my DSLR and the photo could be the main content, it has to look really good.
I was unimpressed with AVIF last time, disappointed really, and liked the vision of JPEG XL much better. Maybe we should be lobbying browser vendors to drop AVIF because the same arguments that have against JXL would also work against AVIF.
16 comments
[ 6.4 ms ] story [ 65.6 ms ] threadThey're so neutral about this that they're locking and limiting conversation to collaborators before they post their message.
https://connect.mozilla.org/t5/ideas/idb-p/ideas/status-key/...
They are ignoring their users ... wonder why they are leaving
I used to run ‘large’ (millions of images) photo sites before and now I run some small ones.
Once a year for the last 5 years I considered new image formats would affect my practice, last year I finally found webp was an unambiguous winner over jpeg. (The bar is high. If you want to save both storage and transfer costs you can’t keep copies in many different formats.)
I found it was not so easy to do this analysis because I have not seen good studies of quality/size/processing speed and the ones that are there are limited. I’ve seen some where people compress the kind of images you might see in the header of a blog but look close and the images are awful. I take pictures with my DSLR and the photo could be the main content, it has to look really good.
I was unimpressed with AVIF last time, disappointed really, and liked the vision of JPEG XL much better. Maybe we should be lobbying browser vendors to drop AVIF because the same arguments that have against JXL would also work against AVIF.