"Whenever it's on it's like having somebody in my house that I want to get rid of and they won't leave. I hate the sound of it. All that noise and light coming from a piece of furniture."
- John Waters -
        Syndicate

* Atom 1.0
* RSS 2.0

Videopia the Book for only $9.99! 236 pages, all original content on EVERY topic from light to copyright.

"In my opinion, the content is worth much more than you are charging for it..."
                            - David B.

"This booklet is a great resource, and worth many times the price. Heck, it's worth four bucks just for the entertaining read alone..."
                            - Shane G.

"This is the best explanation of
Internet video I have ever read."
                            - Mark H.

Google's Video Codec PDF Print E-mail
Written by D. Eric Franks   
Wednesday, 19 May 2010 08:45

While Apple and Adobe have had their horns locked in a very public battle over the next video codec to take over the Internet, Google has been quietly plotting its own path... to world domination. Google, of course, has a very, very large stake in the matter - otherwise known as "Youtube" - so when they officially announced[1] the completely open source On2 VP8 codec, complete with supprt from big-league players like Adobe, Mozilla, AMD, NVidia, Skype, freescale and more,[2] it's a big deal. A very big deal.

The name may sound familiar: the commercial On2 Technologies' On2 VP6 codec was (and continues to be) a very impressive codec and was the only widely available distribution codec to support alpha channel transparency, which made it the codec of choice for folks wanting to create Walk-On Web Avatars[3] that stroll out across a Webpage to say hi/annoy the crap out of you. And how, exactly, did the On2 VP6 codec get so widely distributed? Well, because it was supported by... (wait for it)... Adobe Flash, of course.

On2 Technologies was purchased in 2009 by Google for $106.5 million dollars.[4] At the time, I figured the reason for this was simply so YouTube could use On2 codecs for free, which makes a lot of economic sense when you are serving two billion videos a day.[5] But maybe there's something more to it than that. Google is rolling out and releasing the code to the On2 VP8 codec today, making it a true and free open source standard, unlike Adobe Flash or even the LA MPEG owned H.264, no matter how friendly their terms.[6] More than that, what would happen if Google announced the On2 VP8 codec was going to be the default codec for YouTube? Everyone would have to support it, from Adobe to Apple. That'd make last week's spat between those two "giants" look like a silly schoolgirl slapfight.

References:
[1] http://code.google.com/events/io/2010
[2] http://www.webmproject.org
[3] Walk-On Web Hosts. The Definitive Guide.
[4] Google to Acquire On2 Technologies. August 9, 2009.
[5] At five years, two billion views per day and counting. May 16, 2010.
[6] Adobe vs. Apple: Flash vs. World. April 30, 2010.

< Prev   Next >
 
Discuss (3 posts)

Jeff Whitley
Google's Video Codec
May 20 2010 02:15:12
This thread discusses the Content article: Google's Video Codec

This is down the same lines as your article, this time Google is wanting to use HTML5 and open source. www.pcworld.com/article/196670/google_an..._of_html5_video.html

Should be interesting how things play out.
 #3264

D. Eric Franks
Re:Google's Video Codec
May 20 2010 15:10:46
And some quality tests on VP8 this AM (and later today too):

www.businessinsider.com/google-has-a-pro...-good-as-h264-2010-5

Generally consensus might be "not as good as H.264" - which I've heard before during the "Theora" phase. Also seen some internal tests that supposedly show "Theora" is better. Having performed a bajillion codec render tests in the last dozen or so years I can say that objectively measuring the quality with PQA tools is fairly easy, but also completely meaningless in terms of human perception of quality. So maybe the consensus WILL end up saying H.264 is better than VP8 (that'd be my bet), but it's gonna be so close as to not matter. The two keys that'll determine VP8's success or failure will be

(1) Is it really more efficient for comparable quality? Now we are really talking insanely difficult comparisons, but we're looking for numbers like "A VP8 file with 80% of the size of an H.264 file has 95% of the quality." Yea. Good luck with coming up with real meaningful numbers like that, but...
(2) If someone at Google can be convinced of the better efficiency/quality ratio and they roll it out as the default format for Youtube, the battle is over. Even a 10% bandwidth savings would be HUGE for Youtube, since a good guess (based on old numbers) is that bandwidth costs are likely a half-billion dollars a year and maybe even a cool billion. I'll take 10% of either number, please.
 #3265

D. Eric Franks
Re:Google's Video Codec
May 21 2010 03:41:33
Holy crap on a stick, here's an end-all-be-all analysis of On2 VP8:

x264dev.multimedia.cx/?p=377

This is from a coder's point of view, not from PQA, PSNR or even visual/perceptual tests, which makes the analysis both more objective and more speculative. The authoritative takeaway?

(1) VP8 has patent issues with H.264. Patent violations? I didn't say that. It is apparently very similar in a lot of ways (which isn't entirely surprising since Microsoft and Xvid codecs are very similar as well). Will this be a problem? Who knows.

(2) Quality-wise and from looking at the code (it is open source, remember), it is hard to see how it could possibly be significantly better and it might just be worse.

(3) Efficiency-wise (file size/bandwidth), it looks like it's impossible for it to be better.

Yikes! Pretty negative. I guess the only upside (and the author is NOT optimistic in this regard at all) is that it is open source and, theoretically, the "community" could fix it. I wouldn't hold my breath.
 #3268

You need to login or register to post comments.
Comments and Discussion: (3 comments)