On2 VP8

[Sagittaire]

also Sagittaire you should redo your comparison vs VP8 again with --no-psy

I obtain OPSNR result with insane optimized OPSNR setting for x264 and even in this case VP8 is in par with x264. Really impressive result for VP8.

[nakTT]

I obtain OPSNR result with insane optimized OPSNR setting for x264 and even in this case VP8 is in par with x264. Really impressive result for VP8.
Any trial version that we can download and have a try? I just like to run a test to confirm this with my own eyes.

[IgorC]

I obtain OPSNR result with insane optimized OPSNR setting for x264 and even in this case VP8 is in par with x264. Really impressive result for VP8.
It's not surprise. At first time on2 was always good.
The problem of on2 is that they always stoped to improve their existed codecs after few months from release date.

Good example is VP6/7. They were good but now (don't need to say where they belong).

The situation can be different now with acquisition.

[mysoogal]

It's not surprise. At first time on2 was always good.
The problem of on2 is that they always stoped to improve their existed codecs after few months from release date.

Good example is VP6/7. They were good but now (don't need to say where they belong).

The situation can be different now with acquisition.

" The situation can be different now with acquisition. "

and you can take that to the bank ! i hope google opens vp8, hope youtube somehow starts using VP8, can we say monopoly :devil: google are smart with money they keep buying new technologies, shame they don't really invent their own :confused:

[iwod]

The problem is there are no VP8 hardware decoder. So i cant see how VP8 will take off?
H.264 already has widespread Hardware support.

[IgorC]

The new SDK VP8 is released
http://on2.com/index.php?id=439&news_id=689

[iwod]

The new SDK VP8 is released
http://on2.com/index.php?id=439&news_id=689

What's new? Sounds like a PR to me...

[nakTT]

Perhaps we should have a duel between the latest x264 and the latest VP8.

[the Mad Duke]

Google now Owns On2's codecs.

[Dark Shikari]

Google now Owns On2's codecs.I thought the merger was still up in the air? Last I recall On2's shareholders were complaining that $100m was too low.

[the Mad Duke]

We tried fighting off the merger but On2 kept extending the voting deadline until they finally got the votes they needed at todays meeting. I expect you'll see an official press release very soon.

[the Mad Duke]

On2 shareholders approve Google merger. (Not all of us did)

http://www.timesunion.com/AspStories/story.asp?storyID=901822

[RNiK]

We tried fighting off the merger ...
Can I ask you "why"?

[Astrophizz]

From the article: some people got a bad deal with the On2 shares -> Google shares transition.

[iwod]

Cant wait to see what google do with VP8.....

[iwod]

A Stupid Question, is it technically possible to encode VP8 with an Modified H.264 encoder if VP8 is 90%+ similar to H.264

[Dark Shikari]

A Stupid Question, is it technically possible to encode VP8 with an Modified H.264 encoder if VP8 is 90%+ similar to H.264It's not 90%+ similar to H.264. In terms of open standards, it probably has more in common with MPEG-4 SP or VP3 than H.264. But given that nothing since VP6 has been reverse-engineered, it's a bit hard to say (VP7 still had the stupid "golden frame" thing though).

[LigH]

On2 VP8 is said to be free of third-party patents. If that is true, and VP8 will be released as OpenSource, then it will possibly become a serious competitor of the "free of charge for web video until 2015" standard H.264 currently used in MPEG4-AVC and Flash Video 10, but also of the still developed free codec Ogg Theora, which is based on On2 VP3.2 and already the top candidate for HTML 5 video.

Let's sit back and watch the battle for the biggest piece'o'cake.

Or rather ... get a grip on the slicing knife too. ;)

[IgorC]

I will jump and throw my point of view too.

Untill now pre-H.265 was able to achieve ~20% of improvement over H.264. It will be reasonably to have doubts if VP8 will achieve at least 5-10% over H.264.
Even if VP8 will be slightly better than H.264 how much time it will take to VP8 to have comparable hard+soft support as it has H.264 now. Remember it takes quite long time for H.264 to become enough popular.
What has happened to VP6? It was superior to ASP. It's forgotten forever.

I think H.264 was adopted in right time and it's late for VP8-like codec.
Too many advantages: blu-ray, HDTV, mobile devices, etc. It's becoming universal like MP3.
And Google just protects itself.

It's not about doom9's ideals but all people's entire love to play videos everywhere.

[Keiyakusha]

I will jump and throw my point of view too.

Untill now pre-H.265 was able to achieve ~20% of improvement over H.264. It will be reasonably to have doubts if VP8 will achieve at least 5-10% over H.264.
Even if VP8 will be slightly better than H.264 how much time it will take to VP8 to have comparable hard+soft support as it has H.264 now. Remember it takes quite long time for H.264 to become enough popular.
What has happened to VP6? It was superior to ASP. It's forgotten forever.

I think H.264 was adopted in right time and it's late for VP8-like codec.
Too many advantages: blu-ray, HDTV, mobile devices, etc. It's becoming universal like MP3.
And Google just protects itself.

It's not about doom9's ideals but all people's entire love to play videos everywhere.

Hmm, I thought the whole point in (possible) opensourcing VP8 is to have nice format for web, out of the box support in browsers and of course it should be free. This is good since it should be superior in quality compared to theora. Thats it. Who cares about hardware support and stuff?

Vp6 was forgotten because h264 came to its place in web, or better to say in flash player. If vp8 will be opensource and used by google, it will be with us for a long time. EDIT: it even doesn't need to be better than h264.

[IgorC]

Who cares about hardware support and stuff?

In theory nobody does. In practice? What has happened to Vorbis? It was open source+high quality. The history is repeating. Everybody promised soon dead to other not-free or closed schemes of audio compression.

Open source and high quality aren't enough. Without wide compatibility (include hardware support) VP8 won't go any far.

[Keiyakusha]

In theory nobody does. In practice? What has happened to Vorbis? It was open source+high quality. The history is repeating. Everybody promised soon dead to other not-free or closed schemes of audio compression.

Open source and high quality aren't enough. Without wide compatibility (include hardware support) VP8 won't go any far.

Well, I'll try to expand my point:
I'm not followed the story with vorbis. Maybe it claimed to replace something, or claimed wide hardware support and turned out it was fail, but VP8 is another story. Now that Google owns all its patents, all that we should expect from it is to became open standard for web video. Also if VP8 will be opened, popularity of vorbis also may be boosted (it will be good idea to use open audio).
I agree that VP8 won't go far, it will be just a format for web. For that it doesn't even need to be better than h264 in terms of compression...
If it won't be open... yes, it will be even more dead than vorbis

Open source and high quality are enough. Why? Because for web we anyway don't have anything better.

[Astrophizz]

The h.264 spec is open, just not free. I'm not sure what you mean by open source (as in editable?) but the theora bitstream is locked and pretty much can't be changed. If you mean the encoder, well there's an open source, high quality h.264 encoder... For VP8 to get any support, the bitstream would also have to be frozen (with the possibility of future expansion).

[Keiyakusha]

I mean something that can be used by anyone without paying anything and with quality better than theora.
EDIT: I didn't actually mean unfrozen bitstream. And free opensource encoders by google - why not?

[Astrophizz]

Yeah I can see it. I'd expect they'd have their own version and then an open source (a la Chrome/Chromium, etc).

[the Mad Duke]

Looks like Sorenson may be jumping on the VP8 bandwagon.

http://newteevee.com/2010/05/10/is-sorenson-media-betting-on-vp8/

[CruNcher]

What a surprise ;) hehe not really anything is better Technology wise then what Sorenson presented over the last years :P

[MfA]

The h.264 spec is open, just not free. I'm not sure what you mean by open source (as in editable?) but the theora bitstream is locked and pretty much can't be changed. If you mean the encoder, well there's an open source, high quality h.264 encoder... For VP8 to get any support, the bitstream would also have to be frozen (with the possibility of future expansion).
H.264 is frozen for patent reasons, the patent licenses are only valid for compliant use. Theora and presumably VP8 decoder specs will be frozen for practical reasons in most uses, but if you want to you can still use a variant.

[Dark Shikari]

H.264 is frozen for patent reasons, the patent licenses are only valid for compliant use.Wrong. Many, many companies have variants on H.264 that they sell (e.g. SVQ3, RV30, RV40, etc). The patent licenses are nondiscriminatory and they're required to sell you one if you ask for it, even if you're not using it for H.264.

[MfA]

Well you have the advantage over me by having up to date information about the existing licenses, while I have to go with web leaks ... so tell me is this no longer part of the MPEG-LA AVC pool license?

Scope of License Grant. Notwithstanding anything to the contrary herein, all licenses granted under this Agreement are limited to a field of use to comply with the AVC Standard. No other licenses for any other purpose or use are granted herein nor are any licenses granted to any portion or segment of any product or thing except those portions or segments of such product or thing that comply with the AVC Standard.

Or if it's still in there where can I find information about the separate RAND license you allude to?

[the Mad Duke]

and thus it begins.

http://www.earthtimes.org/articles/show/vp8-video-from-anystream-avalon-and-agility-2g,1305194.shtml

add Zencoder to that list.

http://newteevee.com/2010/05/18/zencoder-excited-about-open-source-vp8/

[Dark Shikari]

On Wednesday, I will be coming out with the first in-depth analysis of the VP8 standard, along with the On2 encoder and decoder. This will cover every major part of the spec, along with the features of the encoder, its performance, quality, and patent risks.

Watch my blog.

[the Mad Duke]

Awesome. Looking forward to that.

[IgorC]

Google conference will be available online as well http://www.youtube.com/GoogleDevelopers
This goes beyond just videocodec VP8. Google pushes TV platform. Serious deal for global communication systems.

[weasel_]

http://x264dev.multimedia.cx/?p=377

Dark Shikari GJ :)

I dont see single point where vp8 is better then h264...
I cant see it`s better then baseline ... comapre to main profile vp8 looks a lot closer to Mpeg 4 ASP(xvid etc)

[weasel_]

I agree that VP8 won't go far, it will be just a format for web. For that it doesn't even need to be better than h264 in terms of compression...
.
Hmm , evereybody is on dailup, HD is baned from web, or what else is reason why picture quality isn`t important for web ?

[Keiyakusha]

Hmm , evereybody is on dailup, HD is baned from web, or what else is reason why picture quality isn`t important for web ?
Where did I said that picture quality is not important for web? I said VP8 doesn't have to be better than h264. This is two different things.

[weasel_]

U have right ;)
and still vp8 doesn`t have to be worse then xvid and h264 baisline profile
Google claim something else : 50% batter then h264, broadcast QUALITY 720@ 1200... both is joke ...

[LigH]

The demo video on the On2 VP8 website (http://www.on2.com/index.php?599) tries its best to make H.264 look worse than VP8. But it also shows clearly that the "slightly more convenient look" could only be achieved by massive post processing. As long as there is no full-featured and detailed configurable encoder, there is no objective comparison with fair encoder option balance...

And comparing against "x264" instead of "H.264" is indeed a bold challenge! Other AVC encoders (like MainConcept) would have been more harmless competitors.

[MfA]

Nothing inherently wrong with post-processing filters, wanting to keep everything in loop is a bad habit ... what is good for the eye is not necessary good for the encoder. Of course for it to be really relevant the codec should have sideband information to indicate the postprocessing should be done.