And how will Fraunhofer Institute not be dicks like they were with MPEG?
The Joint Photographic Experts Group's image compression standard, JPEG XS, has started to turn up in implementations ahead of its slated April 2019 completion. The still-under-development standard is pitched as targeting VR, 5G, and high-resolution image formats like 8K. The group says it's working on a low-latency image …
Wednesday 11th April 2018 09:40 GMT Charlie Clark
Indeed. Pity the article makes no mention of whether the format will be patented or not.
As it stands it may be too little too late with other bitmap formats based on more recent codecs (WebP and HIFF) having the advantage of being around already with hardware support (via Google and Apple respectively). I guess we can expect a slew of benchmarks over the next year.
Wednesday 11th April 2018 05:46 GMT Kevin McMurtrie
Is this solving a lack of fresh patents?
Didn't the 1990s already have a ton of research into fixed ratio, low overhead, lossy codecs? The idea was that of all the possible permutations of a small data block, not all of them were significantly different so they can be removed. Prediction and state of the compression were trivial so latency and effort were low as well. Primary uses were boosting multimedia throughout on 1x DVD-ROM and LANs. Some of them are still around today in various telephone bits.
Wednesday 11th April 2018 20:31 GMT Anonymous Coward
Patents: Biz as usual or try something new?
The MPEG group has never really had an interest in free and open standards. However, due to the pressure of good quality codecs that are not encumbered with active patents, it will be interesting if they try to push for business as usual, or try a less restrictive approach.
I think they should consider limiting royalties to the hardware implementations, (Pay for the things you are already paying for) not software implementations, which in addition to blocking many Open Source implementations, also can slow adoption and encourage format wars. This can be accomplished by tying the payments back to devices like AV hardware, and video cards. Allow the hardware license to act as a blanket to cover any software implementation. Then an open driver and player can be built to implement support for the licensed hardware, and once the patents are expired, allow fully open implementations.
This is also more fair then making an end user effectively pay royalties on multiple implementations of the protocol on the same device(for example if Chrome, Windows Media Player, Adobe Premier and a dvd/bluer-ray player all supported the codec.)