Firefox gobbles H.264 to serve up vids to mobes, slabs Mozilla is doing a deal with the patent devil to serve video to users via Firefox on smartphones and tablets. The freedom-luvvin’ web shop will allow Firefox to play video on existing decoders on devices, including the patent-encumbered H.264, licensed by the MPEG-LA patent pool. In a project here, the decoding would work on …
Reminds me of an issue of the satirical comic "Transmetropolitan" in which there is some reminiscing in France about the loss of their national language. They had clung on stubbornly but eventually lost when the truth set in: that no one really wanted to pay attention to their prized creations until they were in English, plus all the most popular must-see programs were only available in English (NOTE: This is just the account in the comic).
I see a parallel here. Google and Mozilla had been trying hard to push for an unencumbered codec, but they face the "English" of Apple. iPhones and iPads all speak H.264 (ONLY...and WANT it that way--after all, they're part of MPEG-LA). Most people don't care about this or that codec. All they'll do is complain when a site doesn't work on their iPhone. Considering the level of attachment people have to them, they're more likely to switch sites than switch phones, so Google's caught between the proverbial rock (H.264 is an encumbered codec...) and hard place (...but everyone else insists on it anyway).
The problem is much larger than iPhones and iPads, three examples come to mind:
1) Windows doesn't come with WebM codecs, so users would need to install the WebM components to see it in browsers like Internet Explorer, H.264 comes ready to go.
2) Many devices (e.g. GPUs or media players, even the Raspberry Pi) don't have hardware support to play WebM (Computer such as the Pi can decode it in software, but it's slow)
3) Most streams are already in H.264, no consumer camera supports WebM. Everything would need to be re-encoded, however VP8's quality has been demonstrated[1] to be poor when re-compressing video previously encoded by other codecs.
[1] http://compression.graphicon.ru/video/codec_comparison/h264_2010/appendixes.html#Appendix_8
Rather disingenuous statement there, ath0
All lossy codecs show demonstrable loss of quality when re-encoding something that's already been compressed using another lossy compression method.
That's true of MPEG2, of VP8 and of H.264.
Even the original CCDs (or film negatives!) are lossy compared to the original light.
It's a fundamental of information theory - once the data is gone, you cannot get it back.
Richard 12, if you had bothered to follow the reference I included in my post you'd find:
"One issue we noticed in the [VP8 compression] test is that most input sequences were previously compressed using other codecs. These sequences have an inherent bias against VP8 in recompression tests. As pointed out by other developers, H.264 and MPEG-like encoders have slight advantages in reproducing some of their own typical artefacts"
So, according to this, MPEG encoders deliver better quality when re-compressing content previously stored as an MPEG format than the VP8 codec does.
@ath0
I repeat, supremely disingenuous. (That means true but phrased in a way intended to give the entirely the wrong impression, in case you were previously unaware of the word.)
Perhaps I could rephrase that precise statement you quoted:
H.264 and other MPEG-like encoders are better at screwing up the source material in the way that MPEG-like encoders screw up the source material.
Or perhaps another example would be clearer:
The Pope is better at being a Catholic than at being a Muslim.
One significant problem is that there does not seem to be a proper implementation indepedendent specification of VP8. It appears to exist as nothing more than some poorly documented program, complete with bugs.
Google could have allowed time to produce a proper specification, perhaps with hooks to allow future versions to incorporate improvements, and fixes for known flaws (e.g. poor fine-texture detail), and to incorporate other methods as they fall out of patent. They could also have fixed the more obvious bugs in the implementation, and added documentation. Instead, they rushed things, and naively decided to treat the messy program they aquired as a substitute for a specification, whose bugs have become "features" that others are now obliged to emulate.
This kind of undocumented hackery with non-existant specifications may be fine for Google's in house software, but it is not what you should to do if you are serious about encouraging others to incorporate a codec in their own software and hardware, and for content creators to adopt it in their production work-flow. They really should have worked with people who know how to create proper specifications of standards which others can use.
(Still, there is always Dirac with low-loss compression, if you really need a patent-free high quality codec.)
Not because it doesn't support H.264 but because it doesn't support Flash.
I know, Flash is going the way of the Dodo but at this very moment in time it is still used extensively on the web, and not just for ads mind you.
So until Firefox supports Flash or the web converts to HTML5, I won't be using it.
Since buying Motorola, Google now owns some H264 patents but are not part of MPEG-LA, I wonder how this plays out.
If Google is going to continue with Motorola’s patent licensing strategy - and they've said that they will, Google is going to find themselves patents to enforce regarding H.264, the very video codec that Google itself is trying to kill with WebM.
This post has been deleted by its author
I don't see how it's any different from learning to install DivX codec on default Win XP. That didn't seem to stop avi files from becoming popular and standard for a time... in fact as avi files became popular, people learned they had to install codecs (or get someone to do it for them). What changed?
Why is it assumed that the average computer user is thick as pigsh*t and must have a complete uninterrupted and magical experience with their computer; that installing drivers and codecs are akin to installing a new engine, rather than simply changing of a tire?
The aim should be to educate people on these types of low-level support issues, not hide them. Anyone who operates a computer in this day and age has some idea about installing software and the occasional driver for a specific piece of hardware; codecs are no different and should be treated as such.
A simple dialogue box informing them of the need for the codec should be all that's required.
Divx had a perfectly working application/ codec installer downloaded from a very clear Divx.com address. They also had quicktime and avi export codec which is a must in any professional, semi professional environment.
Is there an officially supported quicktime export/ import codec (that means avid/fc) for webm? Anything avi based for premiere? Did they bother to talk to codec implementation power houses like 3ivx or core codec guys?
If there's one thing history shows, time and time again, it's that principle gets you nowhere.
The world revolves around compromise, that's why Intel's architecture is crap, along with Microsoft's APIs, and the protocols and basic methods of the Internet are so shite and have no place in 2012. Yet they're what we have (thanks to legacy and the fact that progress is a gradient not a gear change) so we have to just make do, invent innovative solutions which add more layers of crap to the underline problems and get on with life.
As a rule of thumb, if there's any dispute and one side puts forward a moral, idealistic or common sense argument, run a mile and invest in the opposition. They'll make the world a more angry, unhappier place, but they'll end up in the FTSE-100 whilst the moral crusaders and sensible techies go bankrupt.
Sad, and enough to make you a jaded gin swilling ginseng pill popper, but true. Deal with it.
With Google failing to promote WebM and continuing to support h.264 in Chrome and Youtube, Mozilla have no choice here.
The Firefox browser has been a total failure on mobile (for a number of reasons, not least terrible performance, lack of flash support and non-standard UI) so naturally to make it at least part way relevant in the future Mozilla need to embrace h.264 otherwise they will be completely forgotten thanks to trusting Google to do the right thing on WebM, and coming a major cropper when they didn't.
Boot2Geko is Mozillas only remaining hope to achieve some relevance in the mobile space for their browser, and launching it with the ability to show only the least popular HTML5 video content was never going to work.
I'd like to have seen WebM supplant h.264 as the standard for web video, but Google totally failed on making that happen. The WebM vs. h.264 battle is lost - time to move on.
Back in day, Mozilla had a working Symbian port. I really don't know who to blame but it disappeared. This is way before Apple iphone or android.
If they actively maintained that port, they could have a great start on way more powerful devices. Just look to Opera mobile. These guys started back in Nokia 7650 days and maintained their Symbian port no matter what Nokia idiots did. In fact, their fuzzy logic link selection via touch with automatic zoom first appeared on Symbian uiq Sony p1I after it died, seriously.
The development of Mobile Firefox on Windows Mobile back in the 2005's was a total joke - the guy behind it, Doug Turner, spent years seemingly making zero progress.
To think Mozilla have now been developing their mobile browser for something like 7 years and it's still awful, and has absolutely no chance of replacing the stock browsers on iOS, Android or Windows Phone means their only hope is to create their own mobile operating system otherwise Mozilla and Firefox will be forgotten.
Mozilla made a catastrophic error of judgement by not taking mobile more seriously, and now they're nowhere in that space.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
