AVIF for Next-Generation Image Coding

h265 is doomed because of the multiple patent pools that have formed and the extreme price hikes compared to h264. It's a risky technology to build on. That's the best ad that AV1 can have, and many members joined the alliance for open media for that reason after it became clear that the patent pool situation wouldn't resolve.

As for your browser support question, yes, Safari has traditionally been anti-ogg, but a few years ago Apple joined the alliance for open media and added opus support to their browsers (although only in the caf container, for some weird unexplainable reason).

Open codecs have won in the lossless audio domain (flac). There is no reason they can't win in the lossy video and lossy audio domains too (not sure about lossless videos but FFV1 seems to have lots of support by archivists who want open technology).

MPEG-LA is indeed well connected to lobbying parties and Hollywood, which established HEVC as the codec for 4K Blu-ray discs. (I personally don't remember the last time I borrowed an optical disc to watch a movie. I actually don't even own any type of optical disc driver anymore!). However, VP9 dominates over HEVC in streamed media.

10 years ago only the hard core capital F Free Software people used them or even considered them

I'm no expert on the new license but it seems the restrictions on HEVC are relatively light. I'm pretty sure it is free to compress and distribute media but the license affects hardware and software encoders/decoders. So maybe your HEVC enabled video card with built-in decoder will be a dollar more expensive. I know that every Apple device shipped comes with HEVC hardware standard for the last few years. I would wager most Android devices too.

Most of them can probably be avoided for the purposes of an AVIF file, but patent US20160232939A1 [2] in my reading seems to be about in-container signalling to express relationships between a "static media item" and "one or more entities" that together "form a group", and "indicating, in the file, a grouping type for the group". The patent appears to be written in a way to allow this definition to encompass, say, a thumbnail and a bunch of frames thereafter, or, say a master image and a set of pictures derived from it, or alternate camera angles of the same thing. Some of these techniques sound like stuff we've seen before, but as is common in patents, the precise wording of claims is often key, and this is where patent lawyers come in.

A thorough look of the AVIF specification [3] and the patents registered with the MPEG LA about this format [1] is likely wise before any widespread deployment that makes use of advanced features of the HEIF container; using it to hold exactly 1 'one-layer' still-image is probably fine.

Additionally, in my reading [5], the HEIF reference software released by Nokia [4] includes a patent grant for non-commercial evaluation, testing and academic research only.

[1] https://news.ycombinator.com/item?id=19874321 [2] https://patents.google.com/patent/US20160232939A1/ [3] https://aomediacodec.github.io/av1-avif/ [4] https://github.com/nokiatech/heif/blob/master/LICENSE.TXT [5] https://news.ycombinator.com/item?id=19874368

Thanks to the propaganda spread by Google and Mozilla. AV1 is "Royalty free". Not Patent Free.

Not to mention Google patents other people's work into AV1. [1]

[1] https://arstechnica.com/tech-policy/2018/06/inventor-says-go...

JPEG XL's current encoder is focused on qualities 0.5 BPP and upwards. Current internet image usage averages at 2 to 3 BPP based on Chrome stats from June 2019. See https://www.spiedigitallibrary.org/conference-proceedings-of... Figure 1.

When you go to 0.1 BPP, every codec will look unacceptable, but the current encoder of JPEG XL looks worse than video codecs.

We will likely be able to extend the scale of handling bad quality down to 0.1 to 0.25 BPP, but it definitely has not been a priority since we don't anticipate that users actually would use it for photographs.

(When an image has a low noise level or less features, then we naturally can compress it down to 0.00x BPP with JPEG XL, too. The 0.5 BPP rule of thumb is for normal vacation photographs and selfies :-)

"Technically open" doesn't help much if each of those formats is primarily pushed by one mega-corp, optimised for their specific use-case, and no industry-wide effort materialises to establish widespread interoperability.

Just last week I tried to convert raw photos to HEIF images on Windows so I could share my images in maximum quality with a bunch of iPhone users. It's basically impossible. The only tool I found was Gimp, but it doesn't properly understand colour spaces and the output looked totally wrong. I'd need OSX or IOS to author files in this format, making HEIC an "Apple thing" for all practical purposes.

It's a safe bet that Google will embed JPEG XL support in Chromium, right? Will I be able to email a .JXL file to an IOS user? Will a Windows PC user with Outlook on their desktop be able to see the image embedded in their email without having to save it and open it again in Chrome? Can I paste a .JXL image into a Word document and have its various features be correctly preserved, or will this be restricted to Google Docs viewed exclusively via Chrome?

Look, here's the thing: All of these post-JPEG formats are nice and all, but are worthless to me and a billion other people unless there's proper effort put into mass adoption. Real industry-wide cooperation, not just token openness.

As I said, it is 2020, and nobody has a viable option for distributing still images in anything other than a handful of 20-30 year old formats.

> The problem is Jpeg XL doesn't do so well in small image size as compared to HEIF or AVIF.

So then why isn't NetFlix working with the JPEG XL group to add the AV1 codec as one of the encoding options? Why are they instead creating yet another standard with no adoption likely outside of their walled garden?

Conversely, why are all the post-JPEG formats such utter failures? Take a look at the list: https://jpeg.org/jpeg/index.html

    JPEG         - The "popular one"
    JPEG XT      - Never heard of it
    JPEG-LS      - Never heard of it
    JPEG 2000    - Heard of it, never used it
    JPEG XR      - Never heard of it
    AIC          - Heard of it, never used it
    JPEG Systems - Never heard of it
    JPEG XS      - Never heard of it
    JPEG Pleno   - Never heard of it
    JPEG XL      - Not yet useful, will likely never use it

This is the core point of what I'm saying: The reason that GIF, PNG, and JPG got widespread support has more to do with luck rather than the standardisation approach. The laundry list of failures and the sad current state are evidence that luck is not a strategy and something different must be tried.

I put this in the same category of failure as IPv6. The IETF got lucky with IPv4 and figured that "more of the same" will work. They didn't get lucky the second time.

I wish I had the "political clout" to bang some heads together at these megacorps, make them sit down together and fix this.

Realistically, I foresee another decade of 8-bit, SDR, sRGB images stretched incorrectly to whatever random gamut the $150 monitor of the recipient is capable of displaying.

If HEIF is free, then fine. But MPEG stuff is not created free by design, so it has to be stated somewhere.