VP8 and the future of Video in Circa 2015 or RIP H.264 ?
This blog post is neither endorsed by Mozilla nor represent their position. ..
Few days ago I , Asa and a QA manager were having a spirited discussion during a coffee break. It was few days before the Google I/O and before the world knew when and how Google is going to make the open source announcement of VP8 codec.
So, I asked Asa why should VP8 be successful other than the fact .. that it is backed by Google.... ( God knows, should it chose, Google can make even my dead Grandma the most famous person in the world !! ).
We had a long technical discussion on how Theora / VP3 is a better codec at crappy bandwidths and how VP8 gives a better cost per bit performance over H.264 under majority of pro-sumer video/web-video scenarios etc.,
But the point that caught my attention is that there is an excellent business reason why VP8 would become a dominant player and take over the lands ruled by H.264... (This is in addition to the technical and vendor support reasons ... )
The H.264 is a technology based on a bunch of patents owned by many companies including Apple,Microsoft etc., so for an individual to acquire H.264 license, it would a pain in all places to obtain licenses from multiple patent holders. So, all these companies patent pooled and delegated the job of H.264 licensing to a single vendor, MPEG-LA, who would become a one stop shop to dispense H.264 license and then distribute the proceeds on a pre-agreed basis to all patent holders minus its service fee.
Even though Apple is part of MPEG-LA consortium, it still needs to get a license to use H.264 in products it sells. So, when it pays a dollar to acquire a license, it goes to the MPEG-LA and say, Apple gets 10 cents as profit sharing from MPEG-LA when they do their books.
In effect, Apple payed a total of 90 cents to acquire H.264 license instead of the full dollar. So, it still fraking losing money every time it need to acquire a license. Same goes for all other companies in the consortium that use H.264 in their products portfolio. Only those companies that hold patents in the H.264 pool and do not make any products using H.264 get a true profit from the MPEG-LA bounty distribution.
So, what would be the reason -- why Apple or Microsoft would not switch to VP8!! Nothing !!!
Or, they can force MPEG-LA to come out saying they don't charge any thing for H.264 for next 20 years or some thing like that ... effectively making it a free codec. This would be a questionable outcome if the only-profit-taking-partners object to it.
So, if I can look into the crystal ball ... come year 2015, MPEG-LA may not have lot of chips on hand to play any video codec game. Either VP8 would become the most dominant player in video codec world and makes H.264 a bit player or hopefully, saner heads prevail at MPEG-LA and they would make H.264 an equally free and open source codec by that time.
Couldn't think of a better recent example of market economy creating a winner in the field of Web Video technology.
Few days ago I , Asa and a QA manager were having a spirited discussion during a coffee break. It was few days before the Google I/O and before the world knew when and how Google is going to make the open source announcement of VP8 codec.
So, I asked Asa why should VP8 be successful other than the fact .. that it is backed by Google.... ( God knows, should it chose, Google can make even my dead Grandma the most famous person in the world !! ).
We had a long technical discussion on how Theora / VP3 is a better codec at crappy bandwidths and how VP8 gives a better cost per bit performance over H.264 under majority of pro-sumer video/web-video scenarios etc.,
But the point that caught my attention is that there is an excellent business reason why VP8 would become a dominant player and take over the lands ruled by H.264... (This is in addition to the technical and vendor support reasons ... )
The H.264 is a technology based on a bunch of patents owned by many companies including Apple,Microsoft etc., so for an individual to acquire H.264 license, it would a pain in all places to obtain licenses from multiple patent holders. So, all these companies patent pooled and delegated the job of H.264 licensing to a single vendor, MPEG-LA, who would become a one stop shop to dispense H.264 license and then distribute the proceeds on a pre-agreed basis to all patent holders minus its service fee.
Even though Apple is part of MPEG-LA consortium, it still needs to get a license to use H.264 in products it sells. So, when it pays a dollar to acquire a license, it goes to the MPEG-LA and say, Apple gets 10 cents as profit sharing from MPEG-LA when they do their books.
In effect, Apple payed a total of 90 cents to acquire H.264 license instead of the full dollar. So, it still fraking losing money every time it need to acquire a license. Same goes for all other companies in the consortium that use H.264 in their products portfolio. Only those companies that hold patents in the H.264 pool and do not make any products using H.264 get a true profit from the MPEG-LA bounty distribution.
So, what would be the reason -- why Apple or Microsoft would not switch to VP8!! Nothing !!!
Or, they can force MPEG-LA to come out saying they don't charge any thing for H.264 for next 20 years or some thing like that ... effectively making it a free codec. This would be a questionable outcome if the only-profit-taking-partners object to it.
So, if I can look into the crystal ball ... come year 2015, MPEG-LA may not have lot of chips on hand to play any video codec game. Either VP8 would become the most dominant player in video codec world and makes H.264 a bit player or hopefully, saner heads prevail at MPEG-LA and they would make H.264 an equally free and open source codec by that time.
Couldn't think of a better recent example of market economy creating a winner in the field of Web Video technology.
Comments
Regarding your cost argument against h.264: The licensing fees are capped at 5 million dollar. Corporations like Apple, Microsoft, or Google don't care about that. 5 mio is peanuts for them,
VP8/WebM is absolutely not based on H.264 Baseline Profile. This is simply wrong information.
http://x264dev.multimedia.cx/?p=377
It seems highly credible to me and easily verified. Yes, the author is invested in x264, which might make you question his motives, but that doesn't necessarily make him wrong. his opinions are backed up specific cases where vp8 uses the patented algorithms of h264.
The other bad news is he thinks the compression quality is going to be worse than h264. Again, he gives specific point by point reasons why.
Your rebuttal here was no more than "not it doesn't!"
No idea if that's true, but he hasn't even attempted to go through the actual patents and sort out which apply and which don't. You'd need a team of lawyers to do that sort of analysis, which one hopes Google did (though it would be great if they'd publish their work so the rest of us could understand their reasoning).
Anyway, the original point about business reasons for adopting WebM isn't going to pan out. Both Microsoft and Apple have strong business reasons for trying to shut out competitors, and enough market power to make things very tough for WebM. But Google and Mozilla have a lot of power too - so it will be an interesting fight.