Are you or have you been an MPEG member?
No, but I have worked closely with folks at member companies.
What do you think makes MPEG special?
MPEG’s collaborative environment brings the best innovators from across the golbe together to move the industry forward.
What do you think is the most important MPEG impact?
By providing a universal codec upon which new products, services, and communications tools can be built.
Do you think MPEG is a good conduit for research?
MPEG provides a competitive environment for some research ideas to be compared against each other and to determine their maturity level. The time window for standardization is too short for getting an idea all the way from concept to “product worthiness”.
Can you comment on your MPEG experience?
As a close observer of the MPEG process since 1990, it seem that the emphasis has shifted from solving really hard problems to making small incremental improvements in order to cash in on a revenue stream. Many company devote vast resources to develop and push for their ideas to be considered essential. The return on investment drives the work, and then the licensing mess afterward (the money grab) kicks in and dramatically reduces adoption.
Are you satisfied with MPEG standards?
Mpeg 1,2,4(AVC) were all AWESOME. HEVC has some nice features, but it seems to be mostly lots of small improvements but nothing really dramatically innovative. H.266 has some new modes and use cases, but they are not core to the codec and might not get into chips, so those extensions might not get wide adoption. I think the Whole MPEG process could be getting stuck in a local minimal, and the push for profit is making the process too risk averse. AOM is focusing on solutions that make economic sense given the realities of the internet, etc. And MPEG is still stuck in the “licensing royally subsidizing research” economic model.
Do you think MPEG standards are the right choice?
AVC, yes, until AV1 gets widely deployed. HEVC is to risky, and there are no signs that VVC will be much better. AV1 will be good enough for 10 years, just like AVC is(was).
What do you expect from MPEG in the future?
VVC has some interesting PCC modes and a few other things, but if they don’t fix their licensing model they won’t be competitive. Maybe require all essential core patents to be licensed in a single pool, at RAND rate that are agreed to in ADVANCE so there is clarity to the participants AND the potential licensees.
