your deluted if you think google isn’t big enough to make people use their codec. their becoming larger and larger every day and they support the public good since they support open standards.

So it’s okay for Google to force their codec on to people, but if Microsoft bundles IE with Windows it’s anti-trust?

want2bfree, have you taken the time to try and play back 720p and 1080p WebM content? It’s basically impossible without a high horsepower machine. Even 360p doesn’t play back properly unless you’ve got a fairly quick Core 2 machine, or better.

Here’s a few seconds of Big Buck Bunny encoded using WebM, available in 1080p, 720p and 360p: http://www.youtube.com/watch?v=jdqvZ2dP6AQ

And you have to be part of the HTML5 trial to get it to play using WebM, if you’re not, go here: http://www.youtube.com/html5

WANT2BFREE, VP8 is NOT an open standard. It is controlled 100% by Google. In terms that you can understand, it’s akin to Microsoft using *.doc for files. H.264 actually is a standard, which is used essentially everywhere. (The MPEG group is an actual ISO standards body.)

WebM is technically inferior to H.264. http://www.compression.ru/video/codec_comparison/h264_2010/vp8_vs_h264.html or also http://x264dev.multimedia.cx/archives/377

Plus they’re not telling that Windows, OSX, iOS and Windows Phone 7 ship with H.264 codecs licensed with the OS. Google would require H.264 licenses for Chrome, Android and Chrome OS, obviously. Yet this explains the actual costs for Google quite well: http://www.zdnet.com/blog/bott/a-closer-look-at-the-costs-and-fine-print-of-h264-licenses/2884

Personallly, I am with the freetards on this one.

Suppose I make a beautiful media player, and i want to distribute it for free. In the US. In packaged form. Does this mean i have to pay an “H.264 royalty” with EVERY COPY sold? Why?? Who do these MPEG LA guys think they are anyway? And from what i heard, they didn’t even invent the standard, they just came afterwards and formed the patent pool (same with VC-1).

Instead, with WebM, i can implement any encoder and decoder i want, and give it away for free. As a hobbyist programmer, this matters to me a lot. And i dont really care who controls WebM, as long as he puts the related patents in public domain.

Also, in costs nothing for Apple and Microsoft to add support for webm in their browsers. Instead, firefox cannot add support for h.264, because they may have to take it off in 5 years (2016).

Sure, all this codec war is going to slow down html5, but it’s better to wait some months more, than have conflicting standards between browsers (microsoft and apple will add support for webm soon, trust me, they will not give up the youtube crowd, who loves the idea of no more flash, to competitors firefox and chrome).

P.S.: WebM is actually better than h.264 in low bitrates.

P.S.: Hardware acceleration for webm is on the way, and the ffmpeg guys just released a faster decoder

P.S.: I ll vote down. Patent pools and “royalties” suck.

Does the YouTube crowd honestly care about Flash? No. Consumers don’t know what the hell it is, let alone care. They just know something that has “Flash” written on it pops up on the screen every once in a while, they click “Next” a few times, maybe an “I Agree” here and there and they’re done. Until they can manage to get widespread hardware support, it will be Google who doesn’t give up users, and keeps Flash, not the other way around.

And the MPEG-LA is actually a very good thing to have, it is the precise opposite of patent trolling. Whereas trolls seek out patents to buy up and sit on in order to bring legal suit against people with valid innovations at some point in the future, licensing associations like MPEG-LA simply act on behalf of all the patent holders who are involved with a technology. Then, you have a one stop shop to get all the necessary licenses. Sure, patents suck. Patent lawsuits suck worse. MPEG-LA helps you avoid them.

And yes, WebM/VP8 IS an open standard. It may not be backed by an ISO or IEEE group yet, but almost all the IP is already available under a license similar to BSD license which is easily one of the most industry-friendly ways to go open. ISO and IEEE aren’t the only groups who can set standards and we certainly shouldn’t get stuck thinking they are. That’d be a dangerous affront to innovation.

The problem with organizations like MPEG LA is that they act as an excuse for the ISO and IEEE groups so that they can adopt patent-encumbered stuff as universal standards. “Don’t worry about the patents, there is a friendly licensing organization right there, it ll just cost you a certain royalty fee”. Of course, this fails if you are a hobbyist or a open source project.

If licensing organizations didn’t exist (or the ISO and IEEE groups refused to adopt license encumbered stuff as standards), people would actually be motivated to make formats that go around patents, like the webm.

“Also, in costs nothing for Apple and Microsoft to add support for webm in their browsers.”
Wrong. It is not guaranteed that WebM is not patent encumbered. You mentioned VC-1. VC-1 was originally attempting to not to create a patent pool. See where it’s today. Many industry experts agree that VP8 is doing too many things to similar to H.264 to not to be encumbered. That recent i4i vs Microsoft patent trial costed $200M. H.264 licensing is capped to $6.5M per enterprise. Do you really think Apple and Microsoft are going to risk that? Definitely not, especially there could be more than one entity holding patents here.

“Does this mean i have to pay an “H.264 royalty” with EVERY COPY sold?”
Of stunning $0.02, that’s capped to that value. You may want to read the article of Ed Bott. And only if you ship your own decoder. If you use Media Foundation, you’re not paying. Plus you need to ship a big amount of software.

“Instead, firefox cannot add support for h.264, because they may have to take it off in 5 years (2016).
Except OSX and Windows ship with licensed implementations. If they implement H.264 that way, they don’t need to pay a penny. ( yes, that is possible: http://blogs.msdn.com/b/interoperability/archive/2010/12/15/html5-video-and-interop-firefox-add-on-provides-h-264-support-on-windows.aspx )
Any licensing would be a matter of Microsoft or Apple. That’s really low quality HTML5 croud FUD. Also Apple is not dropping MP3 or AAC support. Both are patented, too.

“P.S.: WebM is actually better than h.264 in low bitrates.”
I’d like to see sources for that before I’m buying that.

“microsoft and apple will add support for webm soon, trust me”
That means they’d need to add a WebM codec system wide? You really believe that will happen? Microsoft and Apple hold according to NMS about 70% of the browser market. YouTube needs to keep H.264 anyway (and it is not costing them a single cent for encoding), so there’s no advantage here.
Plus WebM guys said they’d build codecs for QT and MF soon, so that would work.

“If licensing organizations didn’t exist (or the ISO and IEEE groups refused to adopt license encumbered stuff as standards), people would actually be motivated to make formats that go around patents, like the webm.”
And licensing would be about impossible.

“And yes, WebM/VP8 IS an open standard.”
As much as PHP or OGG Vorbis (on which WebM relies, too) are standards. I wouldn’t call it standard, as they format is not written down anywhere, but rather WebM is a de facto standard only.

“P.S.: Hardware acceleration for webm is on the way, and the ffmpeg guys just released a faster decoder”
“NVIDIA has stated that they support VP8 adoption, but they have no specific plans to provide hardware support.[39]” (from Wikipedia)
AMD and ARM announced support, but that doesn’t meant we’ll see any kind of real high quality acceleration as it is available for H.264 now.

First of all, if MPEG LA was able to make a patent pool for VP8, they would have done it by now. But they can’t, because special care was taken during the development of VP8 to avoid all the MPEG LA patents. Such care wasn’t taken during the development of VC-1, and that’s why it quickly got patent encumbered. We could play the “will VP8 become patent encumbered?” game forever, but trust me, all the MPEG LA has against Theora and VP8 is vague threats.

Also, how does media foundation helps multi OS browsers like Firefox? Or multi OS players like VLC? It doesn’t. So what you are saying to me is that open source projects should either be frontends for Media Foundation and other licensed decoders (ex from Apple), or pay $2000 for every 10.000 copies shipped (last time i checked, VLC had shipped about 11 million, thank they are based in france were software patents dont apply). This kind of thing sucks! You ve just validated my points! Long live webm!

P.S.: Sorry for trying to make predictions in the previous post. Lets wait and see.

*100000 copies shipped

“We could play the “will VP8 become patent encumbered?” game forever, but trust me, all the MPEG LA has against Theora and VP8 is vague threats.”
Mistake is, you’re equaling the MPEG LA with ANY possible patent holder. Second mistake is, WebM was launched last May, that’s about 7 months ago. For Android, it took Oracle until recently to sue them for having copied almost the same source code. The MPEG LA is controlling tons of patents, plus it’s not granted any holder has stepped up. Also, yes you are correct, this game is playable forever. Only H.264 is out for almost 7 years now, so it happening there is unlikely, especially because of the wide popularity. Even if there’s only the slightest risk for this to happen, they won’t do it. One patent trial outweights any licensing costs by far. And it’s not only that. They simply have no reason to implement WebM.

“Also, how does media foundation helps multi OS browsers like Firefox? Or multi OS players like VLC? It doesn’t.”
Now the really harsh truth kicks in. It does. Media Foundation and QuickTime are available and market dominant. Let me show you some figures: http://en.wikipedia.org/wiki/Usage_share_of_operating_systems#Web_clients Here, let’s take the Wiki medians between fosstard and commercial analysis. Of these listed OSes, every single one, except mainstream Linux has licensed H.264 decoders. (OSX and Windows even with licensed encoders) (Note: You need OSX 10.6+ or W7+, however considering that these platforms as well as HTML5 are definitely adopted before 2015 – XP support ends 2014 for instance -, there is no problem) Now before you say: “But this will require them to ship with at least MF, QT and GStreamer”. It does for propery playback anyway. Otherwise you can enjoy non-HW accelerated playback.

Lastly, something else comes into play. That is that you can’t sue any FOSS project as an entity. VLC can ship H.264. Firefox – due to being Mozilla Corporation backed – can’t. Plus Mozilla is concealing something else: Remember why the IRS inquired them? That $43.1M net income, where did that go again? Even if they had to pay the maximum available amount, it would only make up $6.5M of the money they get from Google for their ads.

“I wouldn’t call it standard, as they format is not written down anywhere…”

Interesting…

http://static.googleusercontent.com/external_content/untrusted_dlcp/www.webmproject.org/en/us/media/pdf/vp8-bitstream.pdf

http://www.webmproject.org/code/specs/

““NVIDIA has stated that they support VP8 adoption, but they have no specific plans to provide hardware support.[39]” (from Wikipedia)
AMD and ARM announced support, but that doesn’t meant we’ll see any kind of real high quality acceleration as it is available for H.264 now.”

Any video card which supports OpenCL can potentially provide hardware acceleration, albeit not as efficiently as having dedicated decoding IP on die. It is indeed entirely a matter of whether or not NVIDIA, ARM (or their licensees like Samsung, TI and Qualcomm), Intel, et. al. incorporate the RTL code into their own architecture’s IPs.

But the kicker is that all the code is available in your choice of VHDL or Verilog. Outside of chip layout, adding hardware support is almost a simple drag and drop operation with some of the FPGA tools out there. And the provided IP is relatively small (as in # of gates) and power efficient. The decoder uses 50mW to decode 1080p, 10mW for SD and can pull off decoding 10 SD streams simultaneously. That’d be amazing for set top boxes. And the codec only uses slightly more at 80mW and 20mW, respectively.

Details here: http://www.webmproject.org/hardware/

It is actually quite impressive. After researching it more, I’d be amazed if it doesn’t see widespread adoption. Unfortunately, that’s going to take a lot of time.

That being said, the quality issues still remain. Anyone who’s seen WebM 1080p content alongside H.264 will attest to that. I think I’d honestly rather watch 720p H.264 over 1080p WebM! ;)

@ChrisXT

See, this is the big difference. There is a big difference between not wanting to program for linux because its a stinky OS with a minimal user base, and not be ALLOWED to program for it, or having to require the user to download gstreamer, which is technically illegal in the US.

If you haven’t understood it by now, there is no point in trying to explain it further.

P.S.: I am too interested in the H.264 vs VP8 debate, especially in the 720p and 1080p resolutions. Its inferior in high bitrates, we already know that (the videos randallagordon saw were probably high bitrate). But the interesting bit is in the low bitrates youtube uses. The “720p” H.264 videos youtube currently serves are actually worse than a PAL DVD. Let’s wait and see if VP8 will do better or worse.

“The “720p” H.264 videos youtube currently serves are actually worse than a PAL DVD. Let’s wait and see if VP8 will do better or worse.”

That actually has nothing to do with YouTube and everything to do with the source content. A lot of the “720p” content on YouTube comes straight off Flip cameras and their ilk. Check out videos that are sourced from a RED ONE or 5DmkII, etc. and you’ll find the quality is top notch.

“If you haven’t understood it by now, there is no point in trying to explain it further.”

I have understood that, that’s not my point. ( Why is downloading GStreamer illegal in the US, by the way, afaik it does not boundle the [patented] codecs, the ffmpeg plugin does? )

However, my point is that the world is not for free. Free software patent argumentation is usually taking out a particular case out of the view. Sure, having to pay for H.264 is an issue. But then so is that BluRays ship with VC-1 or H.264, and not Theora! Anyone heard complaining about that?

“and not be ALLOWED to program for it”

Another common FUD point. That’s not the case. You’re only not allowed to ship it for free (without licensing that is). Hey guess what, so is MP3, AAC, MPEG4, MPEG2, MP4, WMA, WMV, RTSP, H.263 and so on. Technically you’re also not allowed to ship the Linux kernel. An FSF analysis alone found 280 patents violated by it.
Nobody will sue a goddamn FOSS project as it’s not a legal entity anyway.

So basically, your point is that the 99% of the world who paid for having a decent H.264 de/encoder are now supposed to use a codec of by far lower quality just because there’s 1% (note: this isn’t even the case, you can buy licensed Linux en/decoders from example the Ubuntu Store) who aren’t willing to pay for what the other 99% did?

Innovation must be protected. And as stupid as it sounds, if there were no software patents, there was no R&D in the video encoding market either. VP8 is also only where it is today because somebody made money from it – using patents.

I think we should all return any BluRay we own and request a new media format we can all play for free.

Furthermore, you should maybe think in another direction: Google paid $130M+ for On2 – just to make VP8 open source? (they shut down the rest of On2’s products -> http://www.on2.com/ )That’s be the H.264 license fee maximum for 20 years – which they still have to pay anyway, as Android ships H.264, by the way… So why exactly would Google spend that much money to make a codec open source?

I don’t know, but there’s something rotten in the state of Denmark, if you ask me…

“Its inferior in high bitrates, we already know that (the videos randallagordon saw were probably high bitrate).”
Says who? I have provided multiple links which state otherwise, you didn’t provide any.

“Interesting…
http://static.googleusercontent.com/external_content/untrusted_dlcp/www.webmproject.org/en/us/media/pdf/vp8-bitstream.pdf
http://www.webmproject.org/code/specs/”
VP8 is defined there. WebM also uses Matroska and OGG Vorbis which are developed as de facto standards. Anyway, that’s not the point, if there was a need for a real standarization, Google could go to ECMA, Oasis or a similar entity and standarize it. So let’s not argue about that.