Slashdot Mirror

On the one hand we have an X264 developer POV. On the other hand we have an On2 POV. I don't want you to believe either of them. They are by their positions biased in obvious ways. I want you to look for yourself and make up your own mind.

In the end though your opinion doesn't matter. Google has many billions to defend their codec. VP8 is open source, patented but a free patent license to all as long as you don't sue. The codec compresses video to an acceptable level, and the decoder presents it well at any relative resolution. Compared to the blocky presentation of low-bitrate H264, I prefer the blurry VP8 because the blur is less distracting to me than the blocks. Check it for yourself. The On2 codecs go back in time as far, or further, than the H.264 licensors (MPEG LA). Google is Google, and patent lawyers are well advised to not mess with a company whose mission is "index the world's information". Google's depth of information gathering is so much deeper than what one would consider "due diligence" that the common usage seems but a shadow of their use. One must consider that they dug down in their trove of information before dishing out $133,000,000 for On2.

So no, I don't want you to believe anything. I don't care what you believe. We'll have our open codec and there's nothing you can do about it. Thanks, Sergey.

While an excellent analysis, it unfortunately confirms all the worst fears I've had about VP8: The quality doesn't match up to H.264

Really? Because they don't have any bias? How about this? VP8 looks significantly better in that video compared to x264.

Here's a contrary point of view, with video.

The spec doesn't document Sorenson Spark

And it shouldn't, that would be the job of Sonsoren Media who have licensed the technologies to both Apple and Adobe. Same thing goes for On2, since it's a On2 Technologies technology and not Adobe. This is not Adobe's fault, it's not like there was "better" technologies at the time they licensed these technologies.

Moreover, Adobe controls the format, not an open standards body

True. On the other hand, your original post did not care about that, only "open up the Flash spec", which Adobe did. However, now I just consider you to be purposely changing your argument because it's convenient for you to keep your stance rather than legitimate reasons.

If Flash were completely open, why isn't there a 100% compliant open-source player out there?

For the same reason new PDF specs aren't supported in most applications - Developers haven't simply done it yet or have no interest in doing so.

With FFMPEG working with these codecs which is essentially part of almost every FOSS media player out there, do you really think projects like Gnash couldn't use it?

Financial objections? What, the price is too low?

You still have to pay people to implement Theora. You have to integrate it into players and browsers, if you offer Theora then you'll likely have to reencode,...

"Vastly"? How can you just make a sweeping claim like that without anything to back it up?

How about using google? There are countless comparisons pitting H.264 and Theora against each other. Basically Xiph says it's more or less on par and everyone else says it's vastly inferior. This comparison is pretty well done and easily reproducible: KeyJ comparison
There is some criticism. The Theora camp complains that the Theora 1.2 alpha version wasn't tested. The H.264 camp complains that insanely slow settings that more than doubled encoding time while giving less than 1% additional quality were used. If you go by metrics Theora needs about twice the bitrate of H.264 High Profile and around 25% more bitrate than H.264 Baseline Profile. If you take visual quality into account the situation looks worse for Theora.

I would say MPEG-2 is vastly inferior to MPEG-4; but Theora is somewhere in the middle.

Depends on the bitrate. For low bitrates Theora is about halfway between H.264 Baseline Profile and MPEG-2. As bitrates get higher MPEG-2 gains on Theora and they come out more or less the same. From the screenshots in the comparison I linked the 1000 kbps MPEG-2 screenshot looks better than the equivalent Theora screenshot. By they way if you read the comments in your bemasc.net link you'll see that there are also links to MPEG-2 beating Theora.

On the other hand, VP8 is supposed to be better than h.264.

This is purely marketing. No settings are given and only screenshots and a reencoded comparison video are available. Since VP8 has never been puplicly available there are no independent comparisons, so we'll have to wait for that until we can really determine how good it is.
From what I've seen VP8 should still be pretty good. It is basically VP7 adaped to be more easily implemented in hardware. I'd guess you could get better quality than H.264 Baseline Profile out of it. It can probably be improved in the future by optimizing the encoders. On2 own codec implementations are notorious smoothing the image too much during encoding thus killing details, a problem the Theora developers have addressed somwhat in their current work.

But I won't pass that judgment until I can see for myself.

You can easily test it yourself. Theora and x264 binaries are easy to find if you just google them.

And that isn't enough? [bemasc.net] Of course, we still have till 2016 to avert this disaster.

This would be the same disaster that only last year people said was inevitable in 2011. Given the MPEG-LAs history it's more likely that nothing will change in 2016.
You shoudn't be too surprised that people won't just switch over to Theora on a whim. If you take Youtube for instance they don't pay a cent for H.264 use so cost isn't a factor. (Although Google pays H.264 licensing fees for the decoder they ship with Chrome.) Switching to Theora would mean loss of wide compatibility, loss of quality or increase of filesizes, switching from a highly optimized and configurable (for speed or quality; speed in the case of Youtube) H.264 encoder, x264, to a slow, slighly buggy Theora encoder while gaining no business advantages.

In other words, the amount of reading between the lines and such required to reach your interpretation is rather excessive.

Things read the same with and without emphasis.

For example, what difference does it make that VP8 is going to be Open Source? And how does VP8 threaten Apple in the first place? If it becomes popular, and if it's superior to H.264, Apple will license it (and if it's fully Open Source in the way people are thinking, Apple won't even *have* to license it).

Who said anything about Apple being threatened? It's the MPEG-LA and their revenue stream that is being threatened; and probably them who are assembling the patent pool. Or are trying to see which of the patents apply to VP8 and Theora

Oh, btw Apple is a part of MPEG-LA as a licensor of patents(they hold one patent) so I guess part of their revenue is threatened too.

It does not take any stretch of imagination at all to see why MPEG-LA would want to try and prevent any MPEG-LA-patent-free video codecs from being released for royalty-free use. The fight is not for a wide-spread free video codec right now. The fight is for 2016.

Apple's objections to Theora aren't strictly financial (outside of the uncertainty of the codec's legal status),

Financial objections? What, the price is too low?

they're practical. In terms of quality, level of adoption and hardware support (specifically as it affects battery and heat)

I wonder how valid that objection is... It turns out that most of the savings are done by using the generic DSP hardware available. So VP8 would also benefit in the same way.

Theora is vastly inferior to H.264.

"Vastly"? How can you just make a sweeping claim like that without anything to back it up? I would say MPEG-2 is vastly inferior to MPEG-4; but Theora is somewhere in the middle. On the other hand, VP8 is supposed to be better than h.264. But I won't pass that judgment until I can see for myself.

The only real advantage of Theora is that it's (presently) free and open.

And that isn't enough? Of course, we still have till 2016 to avert this disaster.

Unfortunately for Microsoft and Apple they actually believe that they control something. Currently there is no h.264 content out there for HTML5 video and Microsoft and Apple have no means to create it.

Tens of millions, hundreds of millions, of cell phones, web cams and camcorders generating H.264 video every minute of every day.

Two fantastically rich corporations with deep penetration into the consumer market space. Partnerships with global content providers and distribution networks.

Out of the game the both of them.

This is what On2 had to say before the merger:

What capabilities does H.264 add to the Adobe Flash Player?

Support of H.264 allows choice for consumers and enterprises, and gives users access to a broader range of content for the Flash Player. Many in the broadcast industry, including content providers for HD DVD/Blue Ray DVD, already encode in H.264. To enable the most efficient consumption of this content on the PC using the Flash Player, supporting H.264 makes sense, and allows users of the new player to avoid delays or other artifacts associated with a transcoding step for a better viewing experience. The already ubiquitous Flash Player has now extended its reach to play back H.264 content across all PC platforms, i.e., Windows, Mac and Linux. Support Center H.264 FAQ

http://www.on2.com/index.php?603 I found them on on2's site. I assume those VP8s are at maximum quality, but if those are real, and this is fully open sourced, Theora AND H264 are in for a beating. I imagine that this will replace a lot of the internet... video if it's really that good.

According to the http://www.on2.com/index.php?605 site, the codec has been optimised, or rather designed, for the ARM 9 processor in particular. This may mean that dedicated decoding hardware isn't necessary if the mobile device is sporting an ARM proc.

I expect there are some programable components, but adding whole a new codec to existing hardware decoders may be asking a bit much.

However, On2 already offered VP6 video decoder hardware designs like this one: http://www.on2.com/index.php?549
And, as I understand it, one of the big factors in the VP8 codec design was correcting issues with VP7 that made it hard to implement efficiently in hardware (or parallel software for that matter). So, I would expect them to be working on VP8 hardware decoders.

Owning VP7 can't hurt.

How would it help? Google is pretty much entirely dependent on other software to get video content to the user, whether it is Flash, video plugins or the browser alone. Owning the codec means nothing if they can't convince the browsers to implement it. As things stand now VP8 is in an even worse position to be adopted by browsers than either H.264 or Theora.

If they do open it up, with a royalty-free transferable patent license, then it has a pretty good chance. Mozilla's and Opera's problem with H.264 was that it was impossible for them to pay patent licenses for each copy they distributed. An opened up VP8 would not have that problem. One of Apple's problem with Theora is that they weren't confident that there were no patent liabilities. Since On2 has been commercially licensing it's codecs for a while now, and no-one else has sued for patent royalties, it is on a bit firmer ground than Theora in that regard. Apple's other problem was that there weren't any hardware implementations of Theora. AFAIK, VP8 doesn't either, although there were(are?) VP6 decoders so maybe On2 has some products in development that they haven't announced yet.

My biggest question is whether On2 holds all the patents for the VP* codecs or whether they are cross-licensing various patents from other folks that apply to many other codecs like VC-1 and H.264. If that is the case then Google will have a very difficult time opening up the On2 codecs any more than the various MPEG codecs already are.

Actually they have vp6 and vp8 http://www.on2.com/index.php?564 which -- surprise, surprise -- on2 claims is better than h.264 -- if google decides to open up vp8 -- it would change the equation radically. Particularly the ogg/vp8 combo. It's also possible some vp3 diffs (theora) would still be useful when applied to vp8 -- although what the chances of this are, I couldn't say. It does solve the h.264 patent license problem for google with android and chrome os. A theora / vp8 release and a move to primacy of vp8 or derivative for youtube would reshape the whole playing field. I'm hopeful, but not gleeful yet.

a codec that looks good with poor PSNR, sure, there's tons of tricks.

*Ahem*

You you believe that one codec can look very good with uncharacteristically poor PSNR, yet you still believe PSNR is useful for comparing different codecs...?

But one that looks bad with great PSNR? It cerrtainly isn't going to happen by accident,

Sure it will. PSNR just doesn't work that way. It is utterly useless for comparing the values of different encoders.

See:
http://www.on2.com/index.php?603

And this PDF:
http://www.on2.com/file.php?185

Or will you just accept, on PSNR figures alone, that On2's proprietary crap vastly outperforms H.264, as well as every other codec on the planet?

a codec that looks good with poor PSNR, sure, there's tons of tricks.

*Ahem*

You you believe that one codec can look very good with uncharacteristically poor PSNR, yet you still believe PSNR is useful for comparing different codecs...?

But one that looks bad with great PSNR? It cerrtainly isn't going to happen by accident,

Sure it will. PSNR just doesn't work that way. It is utterly useless for comparing the values of different encoders.

See:
http://www.on2.com/index.php?603

And this PDF:
http://www.on2.com/file.php?185

Or will you just accept, on PSNR figures alone, that On2's proprietary crap vastly outperforms H.264, as well as every other codec on the planet?

Sun released JavaFX 1.0 today, in a bid to take on Adobe's Flash and Microsoft's Silverlight technologies.

JavaFX, a technology by sun which will let developers write "rich internet applications" using a Java-esq language which has been in development for quite some time now has finally been released, joining rather late to the game. Also it requires it's own plugin which will be a hefty barrier to entry.

It is Sun's first Java release to include standardized, cross-platform audio and video playback code (in the form of On2 licensed codecs).

So, it will include Audio/Video support, and it is using codecs from On2. I don't think that this means the codecs are free in any sense of the term, just that Sun is paying the cost of licensing these codecs for the developers using JavaFX. Nevertheless, good codec support is always a good thing.

The lack of a Linux or Solaris release is a notable absence.

Kind of deflates the "standardized, cross-platform" claim...

The development kit currently consists of the base run-time, a NetBeans/Eclipse plug-in and a set of artifact exporters for Adobe CS 3&4."

So, the download they offer you includes the software to run the software you write, a plugin for eclipse and netbeans which are the two leading Java development environments, and some plugins for a few Adobe products to move stuff from them into the JavaFX world.

So can you please tell me where I can buy Linux versions of Final Cut Studio, Shake 4, Creative Suite CS, Flix Pro, Sound Studio, Toast, and Sony's XDCAM HD transfer apps please?

I cannot believe that you got modded troll for that. It is true, there are not enough commercial apps available for Linux at the moment. What I wouldn't do for Solidworks on my Ubuntu box!

For Creative Suite, write to these folks:
http://www.adobe.com/aboutadobe/contact.html

For Flix, write here:
http://www.on2.com/index.php?373

Write to these folks asking about Sound Studio for Linux:
http://www.freeverse.com/support/

This is the address for the people responsible for Toast:
http://www.roxio.com/enu/company/contact.html

And the infamous Sony:
http://www.sonycreativesoftware.com/corporate/contacts.asp

Please, don't be shy and WRITE TO THEM! If we don't write and let them know that Linux is a viable OS with a strong user base, then they will never port their software to Linux.

Why not use Yahoo Hosting which has unlimited data and transfer?
Couple that with On2's Flix encoder for making VP6 videos (makes Youtube look like a B&W TV with rabbit ears vs HDTV with a true 1080p source)

there's also a few other choices in terms of hosting that has unlimited data storage and unlimited bandwidth for a reasonable price per month...just use our good friend...google.

Flash video (flv) is a container around codecs, like AVI, OGG, and even MPEG is. The codec typically used in Flash is by On2, I believe. I guess Jobs is complaining about Adobe's mobile implementation of the decoder.

However, Adobe recently added support for H.264 in Flash. H.264 is more widespread and there are hardware-accelerated implementations for it in the mobile field. Youtube has started supporting that codec as well (add &fmt=6 at the end of video URL to try, if that video has been converted)

Hell, I worked on a mobile chip which includes MPEG4 and H264 encode/decode acceleration, which has been included in a recently announced Nokia smartphone, and I can confirm that On2 aren't accelerated (and Microsoft's VC1, used in DVB-H, is only partly accelerated), and thus have to run on the ARM core, at the expense of higher power consumption.

just like you can use any number of proprietary image formats and w3c never recommend a patent-free image format... oh wait .. there was the whole png standard. How many png images where used on the web before it was a recommended? MPEG as a "standard" is irrelevant as long as it remains non-free to implement, since distribution costs are inherently incompatible with free browser distributions. Ogg theora is on the other hand perfectly compatible with being distributed in a proprietary system.

A codec agnostic implementation of the video tag is next to worthless. A simple javaScript library could accomplish the same thing.. Codec agnostic video tag resents no significant difference from the object/embed tags that we already have today. If that approach is taken video will remain a second class web citizen wrapped up in proprietary encapsulations. The whole point of the w3c is to promote/develop interoperable technologies, in the current browser environment non-free implementations are not interoperable. The w3c will be obsoleting themselves if they take the codec agnostic approach.

The drive for codec agnostic video tag is simply an effort to put a proprietary wedge into web video distribution platform. The codec agnostic video lets proprietary technology providers squeeze hugely profitable wedge into the web platform. This represents a new, undesirable, untested and unhealthy direction for the web. We are already well down with the consequence of proprietary video being felt worldwide. road of proprietary web with flash video and its network effect that pushes web technologies into service model more so than other web platforms that are based on open standards such as wikis and blogs that have been mostly distributed. of consequences that is only starting to be felt. So far Adobe/On2 has been very lax in enforcing their proprietary codecs with many sites getting away with using ffmpegs flash video encoding for "free". But we should not give away ownership key portion of the web platform to a single corporation. Its antithetical to what the web platform is and why it has been so successful.

http://www.on2.com/company/news-room/press-release s/?id=222

Open Source the player! If Adobe have coded the entire player in-house (which I believe they have), then why not Open Source it...it's a free download after all! Even if they've patented some methods used in the source code, they own the patents and the copyright on the source code, so that shouldn't stop them open-sourcing it surely? Just exactly what is Adobe's objection to open sourcing the player? Sheer bloody-mindedness?

They're using an On2 codec for the video. On2 claims that they're codecs are better than H.264. If that's true it might be just as good as iChat AV.

Looks like the video compression is On2's VP7. Does anybody know about VP7 & what kind of video quality it produces at 512k?

It's using on2's VP7 codec, but it's a custom container of some kind, not avi, wmv or similar: http://www.on2.com/company/news-room/press-release s/?id=254

It's using on2's VP7 codec, but it's a custom container of some kind.
http://www.on2.com/company/news-room/press-release s/?id=254

video direct download

I've got the codecs installed but not even mplayer can deal with it.

The product will be allowed to languish for years, squeezing every last drop of usefulness out of a once mighty product, compressor(ing) a once large user base on2 other software solutions.

dang if I can figure out how to work Divx into that sentence... ;-)

Looks like on2 and their vp6.2 codec didn't make the cut. I guess they could still implement a codec based on the open standards . Time will tell, but MS is positioning itself out of the PC world and into the DRM revenue one.

Just as an example, EVD has been something of a flop.

The Chinese didn't actually invent most of the technology in EVD; they seem to have just taken the existing DVD medium and licensed On2's VP6 video codec (On2 is US-based). They've shipped so little actual EVD units that On2 is suing the Chinese companies involved for not fulfilling their minimum units obligations. As a bit of anecdotal evidence, my Chinese friend claims that he can't even find EVDs any more (there were more several months ago).

TD-SCDMA was also developed in large part by outsiders (Siemen's IIRC), and hasn't completely taken off, though this may change if/when the government decides to require operators to use it. Point is, I believe many of these new "Chinese standards" are really just a way to encourage real competition in the new Chinese economy, and it's actually working extremely well. EVD, for example, might actually be a really great way to stop the HD-DVD mafia from imposing discriminatory patent fees against Chinese electronics manufacturers.

"OGM" is a spin-off of Ogg from some time ago which hacks together Ogg (a great stream container format) and FourCC (the codec identity field from AVI) to easily add proprietary codecs (ie, DivX, XviD, other MPEG derivatives) to Ogg. Obviously, this is not endorsed by Xiph, the creators of Ogg and Vorbis, as they don't support patent-encumbered codecs.

Also, Ogg is not an acronym, so capitalizing every letter is incorrect. This is a common mistake. :-)

For those that don't know, the Theora project is intended to provide an audio/video file format based around Ogg Vorbis and On2's VP3. But while Ogg Vorbis continues to be compettive state of the art audio codec, the VP3 video codec has gotten stale and since been ditched by it's author in favor of VP4, VP5 and now VP6 codecs. It would be interesting to see what the marriage of this new video codec with Ogg Vorbis might provide. Considering after over a year the Theora project has made it to an alpha3 release, I would guess there is definately time to rethink the direction of the project without loosing too much.

As sad as it sounds, I'm not surprised that XviD was left out. After all, alot of these reviewers pay attentions to what is being *marketted*. But I am surprised that On2 failed to get their VP4 Personal Codec noticed by this reviewer. I guess On2's marketting group dropped the ball when it came to make ExtremeTech aware of it's offerings.

As sad as it sounds, I'm not surprised that XviD was left out. After all, alot of these reviewers pay attentions to what is being *marketted*. But I am surprised that On2 failed to get their VP4 Personal Codec noticed by this reviewer. I guess On2's marketting group dropped the ball when it came to make ExtremeTech aware of it's offerings.

This may be true, but why use MPEG4 where it wasn't designed to go? I can't imagine it's that much better than MPEG2. And if it's quality you want, there are codecs available already like VP6 (which is better than H.264). I've already heard talk of using it for a new generation DVD standard.

China licensed VP5 and VP6 for use in the EVD standard, at least according to On2 themselves. However, Ogg Theora is based on VP3 but is not perfectly compatible.

I might be wrong, but I think certain codecs are already "immune" to this type of ickedyness. I think the VP family of codecs by On2 won't pick up on anything that lasts only 1 frame. It would be a waste of time to encode anything that insignificant anyway. If they really want to protect it, they should just make the movies more whiz-bang, with cool special effects that rely on near perfect accuracy.

Sheesh, DivX uses way to much overhead for so little quality, check out VP6, if you want to see real video quality.

Well, first of all, DivX 4 originally had an open source code base. DivXNetworks had a 2 system thing going on, them working on their own code, and also supporting and open source version. They changed however, amid the release of DivX 5. This is why the XviD group was formed. Their original code base was forked from the open source DivX 4 code base. Much of that has been rewritten by now though.

Also, there is an Ogg progect, called Theora, that is an open video codec. It is based off a codec called VP3 that was orignially developed by a company called On2 They gave the VP3 code to Xiph and continue to work on their own proprietary codecs, such as VP6.

Xiph.org isn't only developing Ogg Vorbis, but also Ogg Theora. It's still in alpha stages though. The technology used in Theora is based on the vp3 codec which is covered by patents, but Xiph.org has negotiated an "irrevocable free license to the vp3 codec for any purpose imaginable on behalf of the public".

Xiph.org is also developing the experimental wavelet-based "Tarkin" codec. As I understand it, it's more written from "scratch", much like Ogg Vorbis, but is even further ahead in the future than Ogg Theora, which they are focusing on right now.

You can play back VP5 in software happily, but you can't encode in realtime. I doubt very much you can encode MPEG-2 in realtime, and that's a much simpler algorithm.

So you've laid out $10,000 for an encoder license then?

Shoutcast/icecast will stream VP3, not VP5. I think you've confused the two. You might like to check out On2's page on VP5 system requirements. You'll notice that the realtime VP5 encoder is still "coming soon". I suspect now VP6 is out, it'll stay that way.

You'd better be encoding an awful lot of video and saving an awful lot of bandwidth to make it worth your while.

OTOH, the quality of VP5 was extremely good for a given bandwidth (much better than xvid).

Will the features in VP4/VP5 ever be folded in to the open source version? (a la Aladdin Ghostscript) Are the differences between VP5 and VP3 enough for us to care? Are any On2 developers actively working on Ogg Theora?

Basically, is On2 playing the version game to force customers into upgrading (and getting more money), or are there real improvements in VP4/5 that Ogg users will want?

Anyway, kudos to On2 in any case. I will definitely be using it.

-- Bob

Ogg Theora

Ogg Theora is our new Ogg/VP3 integration project partnership witn On2 Technologies. See the Theora.org website for more information.

Xiph.org is coming up with a video codec entirely on their own called Ogg Tarkin, and it will have been made entirely from scratch.

On the main On2 page, they have links to the codecs.

You've just got to love any company that makes a movie called To Kill a Lawyer.

Great job with the Google search though. Wish I would have thought of that.

• Brand value. mpeg is much better known than the alternatives. This can be a powerful asset. I switched my long-distance from Qwest to PowerNet Global. There is no difference in quality because PNG uses Qwest's infrastructure, but I pay less than half as much as I did, but I realize a lot more people go with Qwest than PNG. In effect, the extra fees go to paying Qwest's marketing, which gives them more market share than the extra price costs them.
  
• Useful features in MP4. I have no idea what features MP4 has over VP3/4/5 that are not visible by viewing the demos on the web, but the article suggests there may be some.
  
• Patent muscle. This surprised me as On2 has been around for some time and once had a market cap of about $1 billion (I miss those days!). They could have sued them when they had the prospect of getting real money in a judgment, where now there is no prospect of getting money from them or the open-source alternatives. I suspect this implies they don't have much of a case.
  
• Negotiating strategy. I have been involved in quite a few enterprise-level business negotiations. Both sides often take extreme initial positions to give them room to negotiate. The belief is that if you begin with your best offer, you'll have to settle on something unacceptable. My experience suggests this is a rational negotiating strategy. The mpeg team is going to face tough negotiating with the major media companies and may feel they need some things they can give away. In this case, they will expect to end up reducing their fees but still be better off than if they had started with a more reasonable offer.
  

I'd be careful with these people and wait for a real Open Source codec like Ogg Tarkin to mature, or contact On2 to get them to fix their license or, alternatively, to remove the "Open Source" references from their site.

From what I recall, there was some discussion about the on2's vp# formats in terms of how free they are. As I remember, there were some limitations and restrictions on using the format (one had to do with keeping all encoders file-compatible with the standard.) and I think they charge for commercial uses of vp4 and vp5 or something....

Does anyone with more experience analysing legalese know how VP3.2 license stands up as "free/open" software?

W

The video quality is too low to be really usable. I find the VP4 codec much more usable. Besides, most mobiles these days still dont have a color display. Perhaps more compression could be achieved by converting the video to grayscale.

On2 web site Seems VP3 is going to die, and I just found out about it.