Hacker News new | past | comments | ask | show | jobs | submit login
WebM - an open web media project (webmproject.org)
181 points by tswicegood on May 19, 2010 | hide | past | favorite | 46 comments



Mozilla has been working with Google on this for some time now, but we haven't been able to talk about it in public until now.* Some of the developers from the Theora project are also working with Mozilla and Google to implement and improve VP8. Firefox builds are available today:

http://nightly.mozilla.org/webm/

One of the most important parts here is that Google owns the patent rights to VP8, and is licensing them to the public under a royalty-free grant -- but the grant doesn't apply to anyone who files a patent suit against VP8, which means the patents are still useful for defensive purposes.

* This is, of course, part of the outcome Mozilla hoped for from not supporting the non-free H.264 format.


Is there the theoretical possibility that Google could turn around and use the patent offensively? Rescind their grant? Or are we all now guaranteed safe to use it?


The license includes a "perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this section) patent license." - http://www.webmproject.org/license/software/

You'll have to ask a lawyer whether it's truly irrevocable, but clearly that's the intention.


There is nothing a contract can say that a good enough lawyer can't get out of, but there is no reason to expect that Google would try in this case.


Watching this from the audience in Google I/O. This is exciting news - good stuff, mbrubeck.


Do you feel like Mozilla and Opera will be betting heavily on VP8 instead of Theora for the next few years? Or will they take the position that both are equally as good and stay out of the fight?


“The VP8 codec represents a vast improvement in quality-per-bit over Theora and is comparable in quality to H.264.” http://hacks.mozilla.org/2010/05/firefox-youtube-and-webm/

(Little editorial comment: There will be no fight, Theora is already dead. That was already pretty clear yesterday and there can be no more doubt about it today.)


This is absolutely outstanding news. Google gets some flack for privacy (some deserved, some not), but it's undeniable that they care deeply about the health of the Web.

"a container format based on a subset of the Matroska media container"

I wonder what reasons they have for not using Ogg, or at least real Matroska? I notice their container format doesn't support subtitles, and is "undecided" on tags. Why on earth would they drop such important functionality, especially when every other modern container supports them?

It would be nice if Google and Mozilla add support for VP8 in Ogg or Matroska, so users can continue to use existing processing tools for web video.

e: Looks like there's already an Ogg mapping for VP8 at < http://people.collabora.co.uk/~slomo/webm/ogg-vp8.pdf >


"Google gets some flack for privacy (some deserved, some not), but it's undeniable that they care deeply about the health of the Web."

That's ridiculous. Smart competitors commoditize complements. Google's goal is to commoditize content and content players. When cash stops exchanging hands for content it starts exchanging hands for advertising. This is a strategic move by Google, not an altruistic move.

Say hello to your new master, same as the old master.


Basic tech is Matroska container (at least "based on") + Vorbis audio + VP8 video.

Biggest news imho, Adobe Flash Player is listed as a supporting software, but I can't see any details yet. Pretty good list of supporters altogether actually:

http://www.webmproject.org/about/supporters/


On-stage, Adobe just confirmed that they will include VP8/WebM with Flash Player and "push it out to a billion users within a year of its release."


The flash support is absolutely key to this thing being adopted. If I'm a video company, I'd bet the future on VP8 and HTML5 with Flash to support older browsers.


They've been quiet about adding other codecs to the player, but considering Google's recent support for Adobe, I can see how they could easily ask Adobe to add that in support of them. Not to mention that they would need to do it for FP to remain relevant as a rich video player layer.


Ok, they've officially announced support now. No mention of which version of FP though.


Why VP8 instead of h264? VP8 will bring you firefox, which does support flash. h264 brings you i(Phone|Pod|Pad) and other hardware. Am I missing something?


Not having to (eventually) pay royalties for serving H.264 videos could be a factor. This assumes that VP8 successfully stays clear of patent threats.


Well I'd make the bet that Apple will eventually support WebM, namely because of Youtube (because I'm betting Google encodes Youtube videos in VP8).

VP8 has the best chance of becoming the HTML5 video standard since Mozilla flat out refuses h.264, yet Apple and MS have yet to publicly suggest refusal VP8.

The iPad and iPhone won't be the only players in the game. Android is getting more popular, and if Google releases a tablet that plays WebM youtube videos, I don't see why Apple wouldn't try to support it as well.


Flash support is very useful. The other big news is the list of hardware vendors pledging support.


Does it means something that I don't see Apple on this page?

http://www.webmproject.org/about/supporters/


If YouTube goes WebM, it's hard to imagine that Apple won't be forced to cave in on this one. And it's hard to imagine that Google wouldn't re-encode the whole of YouTube (geez, the CPU cycles!!!) But I would expect legacy support of H.264 by YouTube for some years to come. So even if Apple does switch their preference, it could be quite some time.


Over a million YouTube videos are available in VP8/HTML5 today, and that number will continue to increase.

http://www.webmproject.org/users/

http://hacks.mozilla.org/2010/05/firefox-youtube-and-webm/


? I'm not seeing any mention of VP8 there.


Sorry, that wasn't the best URL. Updated it to one with more info.


Apple already supports WebM as much as Microsoft does. You can just install a Quicktime component that adds support for the container and the codec, which gets you <video> support without them having to do anything. Microsoft isn't shipping it either, but they announced that they're adding Google's DirectShow filter to their whitelist for IE9.


I'm very intrigued by Apple not being on this list. Considering that MPEG-4's container shares a history with Quicktime, and the amount of time and effort Apple has spent supporting h.264, I wonder what they're saying internally?


Interesting. With both Chrome, Firefox and Opera supporting it, there will still be plenty of options for viewing WebM on OS X though. I wonder what will happen to Safari if WebM gets popular?


More interesting is Mobile Safari.


Safari and Apple will have to cave in. If WebM becomes a de facto standard for video on the web, then Apple would be fools to limit their browser like that. Especially since it's patent free.


People will just install the Quicktime component, just like they'd have to install the DirectShow filter to get support in IE.


Microsoft and Intel are two other notables that are missing.


The interesting question for me is whether existing mobile hardware can accelerate this in any way - i.e. through the ARM Neon extensions and similar. Otherwise we will have to wait for a whole new generation of hardware if this is going to make any dent in the mobile web.


Both Mozilla and Google funded work to implement hardware-accelerated Theora on mobile devices[1][2]. While I don't know of any similar efforts for VP8 yet, I'd guess that it is inevitable.

[1]: http://google-opensource.blogspot.com/2010/04/interesting-ti...

[2]: http://blog.mjg.im/2010/04/16/theora-on-n900.html


On2 was already developing hardware solutions (iirc they already had hw encoder/decoder designs) for their codecs before G acquired them.


On2 announced some stuff with ARM a couple of years ago, apparently they worked together to both design and optimize VP8 for the ARM platform, including NEON.

And one of the low cost platforms for Android phones had a mysterious "On2" hardware acceleration block on their schematics at least 6 months ago.

A quote from ARM today on the VP8 announcement:

The efficient processing is helped by the fact that the VP8 codec is already optimized for Cortex-A/v7A class processors with the NEON SIMD (Single Instruction Multiple Data) engine. NEON allows the ARM CPU to work on multiple bits of data in parallel, and optimizations for ARM11 processors are also available

http://blogs.arm.com/software-enablement/support-for-vp8-and...


So, what does this mean for Apple? They're betting on H264 being the web video codec. If YouTube is all in WebM and Firefox supports it, that's 60% of the browser market and one of the largest video sites. Is this game over for H264?


So, any bets on how long until litigation from the H.264 patent pool turns up?



Right, but the MPEG-LA’s strategy has always been to hint about possible upcoming litigation while never actually launching a suit.


From the intro page:

a container format based on a subset of the Matroska media container

Am I the only one who couldn't help laughing when reading that part?


If you love the web, the web will love you back.

This is the kind of progress that makes me happy to tears.


Awesome.


We have a saying in Sweden, "the more chefs, the worse the soup". It kinda embodies what I feel about yet another attempt at squeezing something new in through the door in hope of setting the standard.


It doesn't need to be the standard. Chrome now supports WebM, H.264, and Theora. Mozilla's position is that we will support multiple high-quality, open, and royalty-free standards if they are available:

"We believe that it is in the public interest for HTML5 video to be backed by multiple, open and royalty-free codecs available in a way that is consistent with the W3C license standards. We would absolutely consider H.264 if MPEG LA would make it available under open web terms as defined by the W3C standards. We stand by our position on Theora." (Mozilla press statement)


This supposedly has a set of qualities nothing else out there has. Namely its royalty free (excludes H.264) and is high quality (excludes Theora).

If those two claims hold true, VP8 has significantly more value than the current options and should be the standard.


Definitely. This is not "just another video codec" - it is significantly better than any other open source video codec. That's a big deal.

And codecs are really really hard to create, for technical and legal reasons. It's not like 5 new production-ready video codecs come up every year. We might not see another credible open source codec for five years.


You are entirely leaving out one very important factor when it comes to video, and that is the video quality itself. Do you really think any user out there except Richard Stallman will give a damned about the fact that the codec is royalty free when the end-result might look worse than what they are enjoying on the web now due to bandwidth constrains putting harsh limits on bitrates? The answer is of course that it won't matter at all to them - they'll just be annoyed.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: