And the fallout from Google’s decision to drop H.264 support from its Chrome web browser continues to fall. Opera’s Haavard – speaking on his own behalf – slammed the article which appeared on Ars Technica earlier today, while Micrsoft’s Tim Sneath likened Google’s move to the president of the United States banning English in favour of Esperanto. Also within, a rant (there’s no other word for it) about the disrespect displayed by H.264 proponents towards the very open source community that saved and invigorated the web.
We already featured Peter Bright’s article in the news elsewhere column to the right, and I originally planned to write a thorough response to it. In my opinion – and judging by the comments at Ars, I wasn’t alone – it was a very poorly written article, full of disingenuous remarks and blatant twisting of facts. Luckily, thanks to Haavard, I don’t have to.
Then there’s this Tim Sneath, the head of ‘Windows and web evangelism for Microsoft’, whatever that means. He likens Google’s move to the president of the United States banning English in favour of Esperanto, which, of course, makes about as much sense as gobble hikky hikky daf gorot. If this is indeed somewhat reflective of Microsoft’s stance in this matter, than the company is even more out of touch with the web than we already knew.
There are still a few things I’d like to add, most importantly the apparent hypocrisy displayed by many of the more vocal H.264 proponents. My biggest gripe – one that actually slightly infuriates me – is that of the parallel between Internet Explorer 6 and H.264.
When Firefox was gaining popularity, it also faced an uphill battle against Internet Explorer’s popularity. Many people who claimed to be of a practical nature, stated that supporting Firefox was nonsensical, since Internet Explorer was the de-facto standard, and by coding your website to at least work on IE, you’d be covering the web more than enough.
Geeks railed against this mentality. Web standards were important, and IE ignored those to a great degree, while Firefox adhered to them – this way, IE hindered the development of a standards-based web. The people who rallied against coding for Firefox sound suspiciously like the people currently promoting H.264: why code for anything else than H.264 when it is the de-facto standard? Why think of the open source world (which is the very motor of much of the web as you see it today) from which we profited so much?
As anyone reading OSNews regularly can attest, I’m no die-hard open source fanatic. However, die-hard or no, it is undeniable that without the vigour of the open source community, the development of the web would have been a lot slower. Firefox is the single-biggest thing to have happened to the web, and now that the Apples and Microsofts of this world have reaped the benefits of Mozilla’s hard work, we should just kick them to the curb?
To me, demanding they pay up millions of dollars per year just to throw away the very ideals that made the web what it is today feels a lot like breaking the arms of the paramedic who saved your life only moments ago.
I’m rarely this passionate about a subject related to technology. Most of the time, I try to bring a light tone to OSNews’ subjects, to bring some much-needed glitter and unicorns to this rather dry industry. However, when it comes to the web, we’re dealing with so much more than just ones and zeros. The web is probably on the same level of importance to the development of human kind as the printing press before it, and handing over the video aspect of it to a known patent troll is one of the most monumentally stupid things to possibly happen to the web.
The web is changing the very world around us, whether it is as a tool for oppressed people to organise themselves, or for westerners to question and hold accountable their governments. The web is not owned by any one of us – not by governments, not by Apple, not by Google, and most certainly not by the MPEG-LA.
Let’s not forget, Mr Gruber, that without the web and open source, the resurgence of your precious company probably would not have taken place. Let’s not forget, Mr Bright, that without the impact Firefox has had on the web, the site you write for would probably load even slower. It is thanks to the open source Firefox that the web is what it is today, and it’s thanks to the open source khtml that the mobile web is what is today. Do not forget that.
In the end, it appears this discussion will be moot. Within a few months, browsers able to display H.264 video will be in the minority, while browsers able to play WebM will make up the majority. With VP8 support coming to Flash later this year, and YouTube inevitably defaulting to WebM, the open source world and Opera will get HTML5 video with WebM. Internet Explorer and Safari users can either install the WebM codec, or enjoy VP8 wrapped in Flash, which, in the case of Safari, is delightfully ironic (and I’m sure a side-effect Google is not too unhappy with).
The lines have been drawn, and at least on the web, H.264 is going to lose.
What a joke. Corporations and Governments created and funded the work.
End of story.
Not that it is much, but I have contributed at least a few thousand lines of code in various open source endeavors… Where is my check Uncle Sam?
While many corporations have contributed a lot to open source, its still the individuals that make it better than the competition. There is a lot to be said for meritocracy and open processes. The web as we know it exists for two reasons: Open source software (apache, firefox, linux, BSD) and open standards (html, tcp/ip, ssl). Before that, the web was just a giant network with relatively nothing on it.
Not sure the web existed before HTML. No, actually, I am sure. It didn’t. Did you mean the Internet? If so, that didn’t exist before TCP/IP. Should not read “as we know it exists” but simply “exists”.
\end{pedantry}
Oh, please. Internet did so exist before TCP/IP.
Okay, my mistake. But before TCP/IP is certainly wasn’t ‘giant’.
Before html, there WAS no web. Sure, there was the Internet, with ftp, smtp, and so on, but there was no “web”.
The roll of RFC editor was done by a graduate student in his spare time: the very first RFCs were edited by hand, in his bathroom.
TCP/IP was developed by academics and students, with some input by corporations such as BBN.
BSD unix was developed by graduate students.
The HTTP protocol and HTML was developed by Tim Berners Lee largely in his spare time.
The NCSA HTTPD, from which Apache derives, was Open Source.
The languages which drove the early web, namely PHP and Perl, are Open Source.
No matter who funds it, the Internet was, and always has been, about open collaboration. That is what makes it one of the greatest communication systems ever created. The moment people start to forget that is the moment the web becomes a passive medium where you and I are nothing more than end users who can only consume what we are given, and only those of us with a couple of million dollars are allowed to play. No more innovative new websites that catch peoples attention. No new protocols that let you and I communicate in new ways. No small yet insanely useful websites. No unfiltered, raw information and opinions from every possible position you can imagine (and many which you can not).
It becomes passive. Just as television is passive. It would destroy the very basis of the thing that made the Internet great, and be as big a backward step as grounding Concorde.
Still, as long as we can watch a video of a cat fall off a chair which has been encoded in a format that people perceive to be ever so slightly better than another one, who cares?
That should be engraved on 50ft letters of stone somewhere. My hat is off.
Sounds like the days of Gopher, where the “Internet” was just a giant encyclopedia used for reading and research.
Really hope we don’t go back to that.
It wasn’t just research, there was porn on gopher too!
TCP/IP was paid for by XEROX PARC. CERN paid Timothy Berners Lee to work and that money afforded him to work on pet projects using NeXT Workstations he didn’t buy.
Oh… you are that guy that thinks that since I or anyone else works for a company they belong to that company at work as well as at home? Good job for being in the pro-corporate slavery camp.
CERN should pay Tim Bernes Lee for the overtime… With interest!.. over 20 years!
All US Graduate Programs, especially Berkeley are Land Grant Universities, paid for by Governments. In this case, the US Government and California State Government.
The Infrastructure of the networks again was Corporations and Government Funding.
HTTP and it’s advances with the W3 are all sponsored by Corporations who draft specs all by professionals working for Corporations.
There is no Global Network by a bunch of monkeys living off of Free and Open Collaboration because they love Richard Stallman.
Oh I see. So what you’re trying to say is that the only “real” Open Source developers are those living rough sleeping under a pile of cardboard?
Are they still Open Source developers if they pan-handle for small change? It’s possible a guy in a suit might have given them a dollar, and then they’d be funded by corporations.
Can I have some of dat shit, you’re smoking?
Open source is still open source no matter who pays for it.
Had Mozilla not been open source AOL would’ve killed it along with Netscape. Had KTML/KJS not been open source (and copyleft) Apple and Google would’ve had to make their own rendering engine from scratch. FLOSS code keeps software alive when profit no longer justifies it’s development/distribution.
Okay.
Ironically, it might be the end of story for the open/collaborative web.
What a stupid comment, really.
If it wasn’t for people having the freedom to create content for the web, then those corps would be nothing. It’s not the paintbrush and paint maker that made Picasso famous!
Oh and thankyou Thom for speaking my mind, you put it perfectly
Edited 2011-01-14 09:59 UTC
webM is open, as in anybody can implement, use and distribute it without asking anybody for permission or paying up anything.
Now it is time to ask apple and microsoft to add support for the codec out of the box in their browsers and all users will be happy. It wont cost them anything. h.264 is a bad choice on the web because some browser’s simply cant support it natively because of licensing conflicts. There are no licensing conflicts with webm making it a better choice.
Yes. webM is subject to submarine patents, so does h.264 and both are backed up by heavy weights. It seem disingenuous to me to always bring up submarine patents when talking about webM but not at all when talking about h.264.
I said it before here, “it aint over until its over”. This was when h.264 was gaining a bit of momentum and it looks like it was going to win over but now it is the one that is on overdrive trying to stay relevant and some of the arguments its supported use? ..shameful, shameful, shameful..
A ‘submarine’ patent is one where the applicant delays the patent being issued, typically to let damages accrue. This problem was addressed by changing the patent right from 17 years from the date of issue to 20 years from the date of first filing.
As for companies owning essential patents that are not part of the H.264 patent pool, they have two choices: (1) Join the patent pool, or (2) file suit against individual companies. The fact that companies have reliably chosen option (1) over option (2) demonstrates that the patent pool has reduced uncertainty for companies wishing to sell products incorporating H.264.
Forming a patent pool for VP8 creates the equivalent option (1) for VP8 essential patent rights holders, and hence reduces risk for companies wishing to sell products incorporating VP8.
This debate is a bit like the reactions to Wikileaks’Cablegate : all the actors who pretended to be in favor of freedom and openess, suddenly spin as hard as they can, blabbering lenghty, convoluted, menacing, self-contradictory “arguments” against those who dare to do something for freedom for all, who dare to put freedom before the private interests of a self-serving clique.
Edited 2011-01-13 21:20 UTC
Who are you talking about?
Engadget, Ars Technica, Techcrunch who passed as “freedom fighters” defending net neutrality, Gruber and all the “web designers” who hailed Apple’s move against flash in favor of an open web, and now rail against Google’s move (or alignment with Mozilla) on h.264, those kind of people.
All of the ones you mention are Apple fanboys and fansites. What did you expect?
Ars Technica is certainly not an apple fansite, and Techcrunch is mostly about web tech, but whatever.
Techcrunch is about how Apple is going to save the world from doomsday. It is also a fanboy website
There are people who think both Apple refusing to support Flash AND Google refusing to support h264 is a good thing.
I have a few questions, which are basically the only issues I have with this.
Is WebM a Google controlled codec?
If users of the codec are sued by other companies, would Google step in to protect the users?
“WebM is an open-source project sponsored by Google.”
http://www.webmproject.org/about/faq/
I doubt seriously that anyone would sue anyone but Google over WebM. The reason? You usually target the person with the deepest pockets, as they have the money to pay.
You would only do that if you’re certain you will win the case with little resistance.
In the case of a patent dispute such as this, you might be better off going after the middle-to-smaller-sized targets who are more likely to settle than pay exorbitant lawyer fees for a lengthy drawn-out battle that might end up bankrupting them before it’s over. If you rinse->repeat this process across enough of the smaller guys first, you gain revenue through settlements and set a precedent that will cause others to become afraid and avoid the entire technology at all.
Right, but if Microsoft or Apple were to adopt WebM, then those companies would be better targets than Google. It’s little wonder they’re resisting in favor of a licensed format where the probability of being sued is much lower.
That’s ridiculous. A plaintiff doesn’t need to just choose one entity to sue. THey can see ALL of them with pockets to spare.
Googl released the VP8 codec. It is left for all to use under a BSD-like license. You hardly get more open and free that that (public domain, maybe).
On the other hand, Google doesn’t protect you from lawsuits.
On the other other hand, as Google intends to use massively VP8, anybody suing you for your use of VP8 will be perceived by Google as a menace to themselves…
No. Google released the code under a BSD license with additional rights granting a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable patent license. The only restriction is you can’t sue them – suing them invalidates the whole deal.
They have effectively handed control over to the community. I’m sure they’ll steer the direction of things to a high degree, but based on the license it will be by the will of the community – they don’t have any choice in the matter.
I don’t know. Maybe? Probably not… That is a loaded question to be honest, it entirely depends on the circumstances. Regardless, there is certainly no guarantee, so the only honest answer to that question is no. However, you have no such guarantees with h.264 either…
“However, you have no such guarantees with h.264 either…”
You are incorrect. The license does in fact have a hold harmless clause and the will defend and represent that to the best of their knowledge h.264 does not infringe on any patents. On the other hand Google refuses to make that claim.
A hold harmless clause is not a guarentee, nor is it indemnification. Do you know what a hold harmless clause is?
http://www.mpegla.com/main/programs/AVC/Pages/FAQ.aspx
Q: Are all AVC essential patents included?
A: No assurance is or can be made that the License includes every essential patent.
If a patent suit is filed against you based on your use of h.264, and the patent is valid and upheld – there is not a damn thing MPEG-LA can or will do about it. Sure, they will likely try like hell to either squash the patent or convince the holder to join the pool – but that is simple self interest. They do NOT promise to cover your legal defense and do NOT guarentee to cover your loses.
Google? your pretty much in the same boat I would expect.
WebM as a format is specified by Google. It comprises various components, the video codec component being wholly owned, specified, and controlled by them. It incorporates the public domain Ogg Vorbis audio codec, Google’s VP8 video codec (owned by them, but open-source and provided both royalty-free and without license cost), and the Matroska container format (a portion of which is licensed under the LGPL, and a portion under the BSD license).
While the individual parts of WebM might evolve, the reference format will always defined by Google (e.g., if Vorbis evolves, the Google fork will be the browser standard, not the one maintained by Xiph).
Per Google, they will not protect users/developers. Their license to use the codec is very clear and explicit; it states that they feel that they probably aren’t infringing on any patents (at least with regard to decoding) and that they grant you an irrevocable license to all the patents that they own that are associated with it. Anything part of VP8 that isn’t covered by their patents is your responsibility.
Note that “royalty-free” means that you don’t pay to distribute/produce content in the format. License cost-free means you don’t pay for implementations of the codecs. WebM aims to be royalty-free, h.264 is royalty free until 2014, then royalties kick in if you earn more than a certain amount ($100,000?) from h264 content distribution. There are h.264 license costs for producing implementations of the codec, but none for WebM. The one caveat being that VP8 may well actually have some patent encumbrances as it seems to apply a number of strategies common with h.264 – but none of that’s really clear and won’t be until a lawsuit provides clarification.
Wrong, the Matroska specs are not LGPL, BSD or whatever. It’s public domain, available to anyone. In fact Matroska was very much designed in the spirit of W3C standards, even though video in web browsers was out of the question at that time. It can be extended just like XML, without affecting older implementations.
Also H.264 is only free for servers sending free video files. It’s not free if people pay for the video (like ads?) or on the player side. So it’s not even half of the equation that is free.
Disclaimer: The following is my own opinion as someone who has experience in the video industry, and is now working as a filmmaker.
So, in my own opinion, Google is right to push WebM over the less open h.264 (you might remember my article about the crazy h.264 license found on ALL video h.264 cameras).
However, I believe that Google with fail with WebM. And while they’re failing, the web becomes fragmented.
The problems with WebM are too many:
1. The codec is not better than h.264. If history has taught us anything (e.g. Lotus 1-2-3 vs Excel in the ’90s), a new format must be substantially better than its predecessor in order to succeed. WebM is not.
2. Google is down-playing the importance of video *editors* being able to export in WebM. Without content encoded in WebM, the format will never get the push it needs. Google seems to think that as long as they have Youtube, Vimeo, and a few phones on their side, they won the war. All I can say about this is: WRONG.
3. According to Vimeo, WebM encoder is prohibitively slow compared to x264. So much slower, that it made no business sense for them to support it, so they removed their short lived support for WebM last year. Vimeo WANTS to support WebM, they’re friendly to Google for this cause. But the format makes no business sense.
And since WebM “is done”, and not just a 0.9 version of the codec awaiting fixes, it means that unless Google puts a BUTTLOAD of engineers working towards WebM 2.0, and THEN releasing the SDK to manufacturers, AND making all the work themselves for FCP/CS/Vegas to support their format, the format simply goes nowhere.
And all we will end up doing, is being fragmented on the web for the next 10 years. Personally, as a content creator, this is a problem for me at the professional level.
1. If history has taught us anything, it’s that quality isn’t the factor that determines what wins. See MP3
2. That’s step two, no doubt. YouTube is step 1.
3. Demand caused by YouTube / browsers switching to WebM will see to the improvement of the encoder. The encoder won’t improve without demand for the format
I’m not disagreeing with you, but that this is a long term thing that started a couple of years ago and will still be going on in a couple of years.
On2 didn’t just stop working when Google purchased them. They’re working on VP8. And VP9. And VP10… WebM is not the only video codec for all time, but Google are picking their battles in order.
1. It doesn’t work like that with content authors I’m afraid. It’s the reason why FLV died: it was too bad compared to h.264. Video people have an eye for pixalation and artifacts, and they get irked when that happens. WebM must become better than h.264, or none of the content creators will touch it. I know a lot of filmmakers myself who would prefer something that’s not h.264, but WebM isn’t what they’re looking for. In the video case, the format must be faster to encode and decode, and have better quality at the same bitrate.
And even when FLV used to rule the web world via Flash, only Adobe’s tools supported it! No other video editor could edit/export to FLV. And even if that was such a popular format online, Apple and Sony didn’t care one bit to add it to their codec support for their video editors. For these editors to add editing/exporting support for it, CAMERAS must be released with that codec! And as it doesn’t seem to be in the immediate plans for anyone to create WebM cameras, editors won’t support the format, and content creators won’t use it. And that would be the end of it. Because at the end of the day, content creators matter, since they supply the actual content.
And don’t think for one moment that Hulu, FOX, ABC etc will ever switch to WebM online, because they ALREADY pay for the h.264 license! There is no reason for them to support yet one more format!
How about saving 6.5 million USD per year?
You misunderstood. They already pay a bulk licensing for other products and services that require h.264. So the license is already paid for completely other reasons, non-web related. So all they have to do is reuse that license for their online videos too. They gain nothing by spending R&D to switch to WebM.
Edited 2011-01-13 21:45 UTC
Totally right. WebM did cost money to Google. But the end goal is to free the web from the H.264 patents. A noble cause.
Once Adobe supports WebM in Flash, the war will be over. Only WebM will need to remain on the web. Browsers where it’s native, fine. The others fall back to Flash, voila.
The providers of web video have to pay license fees for h.264 video. The yearly fees increase per the number of views. It is like a rent, rather than a one-time fee.
There are no such fees for WebM video. Web pages can be written and promulgated on the web to millions upon of viewers year after year for no charge.
Even if we grant you (debatable) that content producers see no reason to switch away from h.264 video, this will undoubtedly change if their web-page-provider customers start asking for WebM.
So, if content providers use professional cameras and work in raw video:
http://en.wikipedia.org/wiki/Red_Digital_Cinema_Camera_Company
… then why not simply provide the final output file to webpage editors in WebM rather than h.264?
Where is there any R&D required for that?
Isn’t HDTV broadcast in h.264? So, they’re already paying the royalties?
Not to mention all their video equipment used for producing video is designed for H.264
In the US it’s still mpeg2, but they’re working towards h.264, fully moved to it sometime in 2015. So everything is already paid, and hardware is already designed. WebM needed to happen in 2005 if they wanted to be a glimpse of support for it in the TV world.
Edited 2011-01-13 21:48 UTC
Uh, hello? WebM is a format for the web!
This does not fly anymore. Professionals need convergence. One good codec as an intermediate codec (e.g. Cineform), and one good codec for delivery (e.g. h.264 or WebM). Having a gazillion codecs and incompatibility between services and hardware is a thing of the past. The industry mandates compatibility, so thinking that WebM is “just for the web” is stupid and naive.
Not at all. Everyone is already processing and converting even h264 videos when they are uploaded to online services, so nothing will really change.
That subthread was about TVs, remember. Please reply to the actual thing discussed, you’re jumping left and right. Now you went back to web.
The story is about video on the web, not on TV.
Dude, don’t piss off the discussion. READ THE THREAD you were replying to EXACTLY. The original subthread you replied to was about TV.
yes, but WebM is about video for the web. Different market, in case you hadn’t noticed. Mostly this market is ads, and silly personal videos about the antics of people’s cats.
Thats just it though… The web is not part of “the industry”. The industry can go to hell for all I care – the industry is the problem – its built on mountains of patents and consortiums and what not and you can never escape the racket…
Having the internet’s video format dictated by “the industry” is stupid and naive. The internet should not have to adapt itself to make “the industry” happy, it should be the other way around…
You like it or not, the industry will mandate the next big format, just like it did with h.264. Besides, the industry is not just billionaire institutions, but normal videographers. I just got an email from a friend telling me that “I could never send my draft editing video to my client in Webm, because i know he wouldn’t be able to play it back on his Windows Media Player or quicktime”.
This little things are what are mandating the future.
By send I suppose you mean putting a WebM file on a website and sending the HTTP link to a client to visualize in a web browser. Yeah, that’s totally not feasible…
I suppose any writer/journalist also has to write their literature directly with an HTML editor, otherwise their work will never end up on the web. Too bad they don’t use a Kindle editor too!
I’m not saying webm will be the “next big format”, you interjected that into the discussion by basically claiming it had to become that to be successful.
You look at this from an industry point of view. Fine, I get that. What I am saying that maybe webm doesn’t cut it for you, but that doesn’t mean it won’t serve as a good, default, works everywhere that matters, open, and free video format. It doesn’t have to be better than h.264 to do that, and it doesn’t have to penetrate the editing world either.
No one is trying to conquer the world with this – we just want an open format…
>No one is trying to conquer the world with this – we just want an open format…
You already have that. Webm is free and open. You also have OGV. But that’s not what Google seems to be trying to do. Google is trying to impose webm on the web by removing it from Chrome (a 25% market browser), and that’s when all that industry comes into place, and as an industry, no, we are not happy with webm for the variety of reasons I mentioned. In fact, it’s for the same reasons why we weren’t happy with OGV either. Not good enough, no support from manufacturers/editors. Being free is GOOD, but it’s simply not the only thing that matters to us. And yes, I hate the h.264 licensing, but at some point as a filmmaker I have to weigh and pros and cons, and h.264 wins so far. If I was to send my music bands their music video draft in webm, I would be fired, or at least screamed at.
As my last comment:
1. I need a WebM 2.0 of sorts. I need Google to put lots of engineers, create an implementation that is both fast to encode and decode IN SOFTWARE (similar levels of speed to x264 for encoding, and CoreAVC for decoding), and is of similar if not better quality than h.264 at the same bitrate. And only when they have that new version frozen, only then they should release the format to hardware manufacturers.
2. I need Google to put yet more engineers and work FOR FREE for Apple, Sony and Avid to implement the codec for them. Because there’s no way in hell these people will add support for webm on their own accord.
Until this happens, to me, h.264 will be doing a better job for both my personal needs, and my professional ones.
Edited 2011-01-13 23:41 UTC
It was the same problem when IE6 dominated the web: you had to forget the standards and conform yourself to MS design choices. During the rise of Mozilla, it was a pain: you had either to choose one browser XOR the other, or to know and implement hacks to have one webpage rendered the same way by both (Mozilla-specific margins hidden behind tags unrecognised by IE, this kind of stuff).
And still, we are better off now, with everybody settling to W3C standard.
People want free and open access to THE STANDARD for web video.
Otherwise, if you are excluded from the proprietary de-facto standards, you are excluded for the majority of web video.
Fortunately, Google´s actions in no way impede you continuing to send h.264 files to your clients (unless you´re relying on them using Chrome/Firefox to view your videos AND they don´t have an h.264 decode codec on their computer). So either there is no problem playing back h.264 (on vast majority of platforms) or they will be delighted you use WebM.
I think your concerns about the current state of WebM encode/decode are misplaced, because that is very definitely it´s current state only. As already mentioned, both of those aspects are being improved on (are you speaking of the latest decoding improvements?), and if you have encoding concerns, I don´t see why most of x264´s optimizations can´t be applied to a WebM output, but read with any reference WebM implementation.
Edited 2011-01-14 00:18 UTC
Google isn’t imposing webm, they are giving it equal footing – something that cannot happen without someone with a bit of clout and money doing it… Without someone doing something webm had absolutely no hope of even being noticed. I was honestly just waiting for Flash to support it- that would have been good enough for me, but Google dropping h.264 is definitely going to stir things up now. At least it has a fighting chance; if you use h.264 it will actually be because it is better for you, not because it is the only realistic option.
You are right though, I am perfectly content with the soon to be reality: Mozilla, Opera, and Chrome supporting it natively, and having Flash as a fallback for the IE and Safari. It doesn’t have to go any further than that to make me happy.
That is fair enough. h.264 is probably the only sane format for you to use for the time being – I never said otherwise. I just don’t see why you would even try to use webm for what you do – in the professional world you would just be swimming upstream…
Read my reply to your first post – there is no need for a 2.0.
Yeah… That isn’t going to happen. Google might offer the help, but Sony and Apple wouldn’t touch it with a 10 foot pole. That is why I never expect webm to penetrate into the editing arena – Apple, Sony, etc. at best don’t care about webm and will pretend it doesn’t exist. More likely than not they will be downright irate at the mere suggestion of supporting it in their tools. Hopefully a 3rd party can fill the void, but I’m not getting my hopes up too much.
I don’t see a problem with that – I never expected otherwise. I only challenged the notion that for it to be successful it had to be good enough for someone like you to use it – I still don’t believe that to be the case.
Actually, he will, if he simply downloads & installs the Windows Media Player or quicktime codec.
http://www.webmproject.org/code/#webm-repositories
These are improving all the time, so check back periodically.
Enjoy.
Indeed, which is why h.264 doesn’t have much of a long-term future.
I really don’t know what video world you’re living in. (My reply below was about your first two posts). But there’s no such thing as a single intermediate codec and there’s no such thing as a single “delivery” codec these days. There wasn’t in the past and won’t be in the future. Not in Europe at least, were I live and work.
Again: professionals choose either uncompressed (high end, almost everyone in advertising and many in film) or any intermediate that fits their needs (a bunch to choose from), and then encode in whatever the customer wants, be it an MPEG2 file, an AVI file encoded for DivX, a Quicktime MOV file (usually H.264), a DVD, a BD, a Digital Betacam tape, a D5 tape, you name it. Most serious professionals don’t even care about what the delivery format is, since someone else in the company will do the encoding for the client.
Broadcasting has even less to do with how you edit. Absolutely NO ONE cares what’s the broadcasting format, since no one will control bitrates and final resolution. That’s the networks’ decision. You better do your homework in “FullHD” (or 2K/4K if you work for the movies) and let the broadcaster destroy your precious work compressing it to the brim.
You’re afraid the WebM encoders are slow these days. Well, unless you do single day final cuts on a feature film I can’t understand the issue, really. You either have a short film which will encode in a timely manner, or have a very long film which will take longer but still little compared to the whole process. Encoders will sure become faster in no time if the standard gets broad support, as it surely will after this move by Google.
Edited 2011-01-15 03:31 UTC
How about saving 6.5 million USD per year?
This is the “Disney Pixar” Enterprise Cap.
The annual licensing fee you pay for every cable and satellite service you own, every broadcast station, theme park, web site, theatrical release, stage show or arena production unleashed anywhere in the Cosmos.
The royalty on every Blu-Ray and DVD sale that bears your name or the name odf your subsidiaries.
“And don’t think for one moment that Hulu, FOX, ABC etc will ever switch to WebM online, because they ALREADY pay for the h.264 license! There is no reason for them to support yet one more format!”
Except perhaps the fact that they could give 40% of their users a much better experience by switching?
But, they won’t! Without hardware decoders, they won’t!
Tegra2 has hardware decode of VP8 1080p.
Old news. There is quite a lot of hardware decode support available for VP8.
http://blog.webmproject.org/2011/01/availability-of-webm-vp8-video-…
The really interesting thing is that, since WebM is significantly easier to decode than H264, existing general purpose 3D graphics hardware/shaders within GPUs can also be used to good effect (via GLSL) to decode and accelerate WebM video. You don’t actually need new hardware, existing hardware GPUs already have the grunt and the programmability required.
But they will! Hardware encoders are almost here (the first is due in Q1 2011), they are!
http://blog.webmproject.org/2011/01/availability-of-webm-vp8-video-…
Bad example. H.264 is the king in the area of ugly blocks and pixelation, while WebM resorts to a smooth, much better-looking blur.
The WebM blur happens in areas of quick motion. The human eye sees rapid motion as blur anyway.
H.264 compromises in a different area … the picture is sharp, but it contains artefacts which are not present.
H.264 scores better on computer evaluations such as PSNR, but the WebM video still looks better.
FLV ended up dying because it wasn’t a ubiquitous format to handle things in. Nobody really knew about it as a format and no one’s tools used it apart from Flash. It wasn’t that it was bad. It had a shot at becoming some kind of standard for web video and it failed.
These things always have a habit of getting driven by the majority and the path of least resistance. YouTube will support WebM natively without conversion right down to Android so users don’t need to do any conversion, or indeed, anything at all, to upload. When they get to a certain point they’ll just stop taking uploads with h.264 as a format and cut off Apple’s h.264 output. h.264 won’t die, but given the history of how these things spread it’s going to be restricted to a lot of industries that will have to have some specific reasons for using it.
WebM and VP8 are quite young formats, and are highly likely to get picked up for all kinds of uses now. The base source code is also there for everyone to refer to, which there isn’t for h.264. The quality arguments are never going to last.
Frankly, no one cares about Hulu, ABC and especially not Fox. If anyone thought these sites mattered then no one would care that Google was ditching h.264. Clearly, a lot of people do, and more than I thought.
However, this is about the future of native video in a format delivered by the HTML5 VIDEO tag, and there’s little if anything around delivering h.264 via that.
Eugenia,
I’ve read your posts and I think you must be talking about really amateur video makers. I’ve been in the video industry for 20+ years and I have yet to meet a single video maker that does any postproduction work in a heavily compressed format like H.264.
Please read point #6 on Haavards reply to Ars’ article. Offline video standards have nothing to do with web video standards. No, there won’t be an VP8 camera any time soon, but that doesn’t mean a thing regarding how people watch video online.
Edit your videos in whichever format you like (and take my advice: forget about any serious work being done in H.264), then encode your final work to whatever codec your favourite site is using. They will recode it again anyway to different bitrates, sizes, etc. (only Vimeo, as far as I know, offers sometimes the option to download the original uploaded video).
Edited 2011-01-15 03:12 UTC
You must have been reading different posts to me then ’cause I saw her talking lots about EXPORTING or SAVING into those formats, not about working in them.
Why is that so much different from saving in an uncompressed format, then converting?
“Oh, noes! One moar steps!”?
Actually, it is being worked on right now, in the soon-to-be-released “Bali” version discussed here in a blog about the previous Aylesbury release last October:
http://blog.webmproject.org/2010/10/vp8-codec-sdk-aylesbury-release…
Regardless of whether the format is DONE or not, you can still improve the encoders. Why would you think otherwise?
Even the now ancient MP3 is still being improved by better and better encoders.
Actually, history shows the opposite. History shows that good enough is good enough:
MP3 vs CD
Wii vs. PS3/360
VHS vs. BetaMax
CD vs SuperCD
The list goes on and on and on. Inferior formats and technologies have frequently won, because it isn’t the quality that matters.
Huh? WebM is getting widespread support from both software and hardware vendors. Even Flash will add the ability to decode VP8.
It’s new. It can be improved (and has been, in many areas). Comparing it to a mature, well tested encoder like x264 is silly.
That depends, and no matter what happens, an open web is better than a closed (h264) web.
>Actually, history shows the opposite. History shows that good enough is good enough:
Not in the case, because even h.264 is not good enough. When Youtube/Vimeo re-encode 720p in 2 mbps VBR, this is BARELY good enough, and it’s DIFFICULT to decode in most PCs without hardware acceleration (my *brand new* laptop, with hardware acceleration, BARELY can decode 720/30p). If WebM requires 2.5 mbps for the same quality, AND requires faster CPUs, then this is a lose-lose situation for content owners.
So yes, there are examples of better versus worse, with worse winning, but in this case, h.264 is itself not good enough.
>Huh? WebM is getting widespread support from both software and hardware vendors. Even Flash will add the ability to decode VP8.
Which part of “EDITORS” did you not get? WebM does not have ANY support from video editors so far. Adobe might add support for it in CS6 just because they add it on Flash, but Apple won’t touch it, and Sony has too few engineers working on Sony Vegas, they can’t even find time to go to the bathroom, let alone write support for a brand new codec.
>Comparing it to a mature, well tested encoder like x264 is silly.
It is not silly at all. You like to think in terms of “well, at some point it will get mature”. And I’m telling you that Google does not have that luxury. For WebM to catch on, it should have been BETTER than h.264 at all levels. Otherwise, a negative chain reaction, and chicken and egg problems arise.
It clearly is, since it’s being used. And WebM is good enough compared to h264.
Even if you don’t think h264 is good enough, WebM is good enough compared to h264.
FFmpeg can encode and decode VP8, and video editing applications will support it soon enough.
Sure they do.
I have just shown you how this is wrong. It only needs to be good enough.
>Even if you don’t think h264 is good enough, WebM is good enough compared to h264.
No, it’s not. And h.264 is used just because there’s nothing better, not because everyone likes it.
>FFmpeg can encode and decode VP8, and video editing applications will support it soon enough.
FFmpeg is not a video editor. Stop running around the issue and reply EXACTLY to the things I talk about.
IF you think that people will edit their video, THEN export it to Huffyuv losslessly, just so they can use ffmpeg, or a GUI ffmpeg, to export to WebM, you’re living in dreams. Why the heck would I do 3x the work, when I can just export in h.264 from within my editor?
Oh, and btw, the only video editors matter, in this order: FCP, Premiere, Avid, iMovie, Vegas. FFmpeg does not plug in any of these editors, so stop bringing ffmpeg as an example, because just like Gimp, it doesn’t matter in the grand scheme of things. It’s these editors that must support WebM. Not ffmpeg with its spyware Windows wrapper GUIs, because no professional content creator would use these utilities, when he can perfectly both edit and export within a single app.
Eugenia, is there a technical reason why all these video editors can never support WebM ?
I explained this in my previous message. Third party support can’t happen for webm, for example, taking the ffmpeg code and writing a codec wrapper around it to work with various editors. The reason is because most of these editors support either AVI or MOV as supported third party containers, and all other codecs that use different containers are not loaded. To add different containers the company that makes the editors itself must add support for it. And that just won’t happen, apart for Adobe.
And that WON’T happen because what? It is NOT a technical reason. It cannot be financial, because that would be a one time operation and a market advantage over the competition. So it’s actually a competitive race to be the first to support WebM which will be ubiquitous because that’s the only fit for the web. And we all know more and more video usage is moving to the web.
So if all these companies put the brakes on this potential competitive edge, they are simply stupid. Like all big companies missing a shit in technology.
Disclaimer: I’m the creator of Matroska so obviously biased. But I already know by experience that a newcomer can win. Why all the HD files on the web are in .mkv ? Because sutpid MOV could not store AC-3 or DTS alongside H.264. And the tiny HW manufacturer took this opportunity to sell more hardware. Then the big guns realised there is a market. And now at CES there was hardly any TV without native .mkv support.
If the big video editors don’t feel the gap, the user demand, someone else will. Sooner or later they will have to follow.
You are the author of matroska. And what does that tell you about webm (which is based on matroska)? Your own format is poorly supported by these same video editors mentioned earlier, and you still have faith for webm support by these editors? How many beatings do you have to take until you get it, that either you make the specific codec work yourself and offer it to these companies for free, before you realize that releasing a generic format, or generic implementation is not enough?
Your very own format and experience is a perfect example why webm won’t succeed. I’m surprised that you still believe in success after .mkv hasn’t managed to get very far itself in terms of editor support. Why would .webm go?
The problem is technical, because as I explained 3 times now, it requires the companies themselves to add .webm container support. And they won’t do it, apart from Adobe. The problem becomes technical because third parties can’t do it, the editors won’t accept their DirectX or Quicktime plugins under the .webm moniker. So that part, IS technical.
And it is financial too. Sony does not have resources to put an engineer to work on webm for example. I know this.
Edited 2011-01-13 23:29 UTC
Funny how not succeeding in video editors, means failing in all areas. Like I said Matroska is a big success, beyond our wildest initial dreams.
AFAIK no video editors it putting their raw editing files on the web. This is never going to happen. So whatever internal format they use is totally irrelevant.
Now it’s sad no serious video editors support Matroska, because it was the main goal when it was designed: replace AVI that I was using that had so many flows. But in the end we never pushed on that market (yet). In fact we hardly pushed anything, just made the technology available then wait and see. And it worked.
So I still think Matroska has a chance in video editors, even that’s totally unrelated to this WebM discussion in web browsers. With all the press and many people implementing their own WebM/Matroska parsers, at least they can see it would be nice in a video editor ecosystem as well. They can see the merit for themselves (rather than assume it’s a technology for pirates).
If I’d get a dollar every time someone told me not to do Matroska because something else almost as good existed…
And the smaller editors already have support for WebM/VP8, which will put pressure on the larger companies to add support.
Kdenlive got support in May 2010:
http://www.kdenlive.org/forum/webm
OpenShot got support in Sept 2010:
http://www.openshotvideo.com/2010/09/export-html5-webm-video-format…
GStreamer supports WebM so editors like PiTiVi support it.
VLC supports WebM in their player, so it should be supported in their editor.
Sure, maybe none of the big, fancy, corporate, commercial video editors support WebM yet. But that doesn’t mean they’ll never support it.
No, these editors won’t put pressure at all. These editors DON’T MATTER, they’re too obscure. Not even bigger editors, like Edius, and Pinnacle matter. What matter are FCP, Avid, CS, and Vegas. Not even FCE matters, since most people prefer to just pirate FCP instead. Not even iMovie matters all that much, because Apple DID try to impose the “iframe” format 2 years ago for consumer editors. It failed (thank God).
Everything starts out small.
The game can change, if you are willing to let it, or even if _you_ are not.
Excuse you me, but you seem to defy the logic of technology.
Apple was started from a freaking garage.
Google used commodity hardware.
Look at both companies now.
Now why cant one codec be more popular than the other.
In this case, Its just a popularity contest.
What do i mean by that? Well look at the current trend.
The most visited site in the interwed is facebook.
In addition to all of these, Miro has an nice converter:
http://www.mirovideoconverter.com/
If Google switches YouTube to use WebM exclusively, a sizable fraction of all videos available on the web will use WebM. If this happens, nobody will be able to ignore WebM without losing a lot of market share.
Googlehas deep pockets. They “could” send engineers to these companies to work with them to get export up and running. They could even pay the other companies engineers for the time they are working on webm support. And if they were smart thats what theyd do.
That is open to debate. Many, if not most, disagree with that assessment.
Consensus != truth.
Bush was voted into office at least once, and at one point had a very high approval rating.
You need to define which version of webM.
Are you talking about the original version released on the launch announcement of WebM, circa March 2010, or are you talking about the “Aylesbury” release late in October, 2010?
http://blog.webmproject.org/2010/10/vp8-codec-sdk-aylesbury-release…
Perhaps you might even be thinking about the still-beta “Bali” release which is due (soon) in Q1 2011.
You need to specify, because the performance of each of these releases is markedly different.
For Aylesbury, compared to the original release, the following performance improvements were achieved:
– 20-40% (average 28%) improvement in libvpx decoder speed
– Over 7% overall PSNR improvement (6.3% SSIM) in VP8 “best” quality encoding mode, and up to 60% improvement on very noisy, still or slow moving source video.
So, which release are you talking about?
And many, if not most, disagree with your view on the previous view…
Tomcat… none of us expects you of all to embrace open standards.
It’s positively baffling to see how arrogant you are. How much experience do you have in video production? How many years?
You’re arguing with someone who does have years of experience, yet are so pompous to think YOU’RE right.
Incredible. It may be that Eugenia isn’t the be all/end all of facts as they pertain to video production, but come on, man. At least acknowledge that she may know more about the subject than you.
Edited 2011-01-14 19:11 UTC
This isn’t about _production_.
It’s about _web distribution_.
VHS is far more poor in image quality than 35mm.
Production/editing in one format, mass distribution in another.
Why do you find that so hard to grasp?
The Chrome discussion obviously centers around default-included codecs vs. add-ons. Any video editor competent to think about how they want to compress their film should be able to add on a codec accessable by the default methods, right? WebM´s container format isn´t really innovative.
And I dispute your suppositions about the future: For one, OF COURSE Adobe will support it if it is included in Flash, and as for everybody else, I think they will be following quickly if YouTube is using it, along with wide-spread hardware support. WHY NOT? Google is clearly going forward with developing it´s reference implementation just so that lazy/understaffed developers don´t have to dedicate many man-hours to the task of integrating WebM support.
It’s not as simple as you think. FCP only allows AVI and MOV containers to be implemented by third parties, so .webM won’t be supported (and that’s the container that matters to Google). Avid won’t give a sh*t about it, as they haven’t even given a sh*t about other major codecs in the past. Avid users had to sacrifice a goat to get Avid to implement h.264 in the first place! Sony Vegas does not support anything apart Video for Windows, so third party non-AVI codecs don’t work. It’s either Sony adding the codec in .webm by themselves, manually, or there won’t be support for it. And they don’t have the man power anymore to do so anyway. The only editor that might add support is Premiere/AfterEffects, under CS6. But that’s just one suite, with less than 30% of the overall video editing market, and it will take a long time until all its users have moved to that new version!
Edited 2011-01-13 22:31 UTC
I guess I´m reaching beyond my expertise, but is there anything preventing the WebM CODEC from being used in another format (compatable with these other editors), at which point shifting to a different container is relatively simple/low-cost operation?
In any case, I believe Adobe will be supporting WebM fairly quickly, at which point other vendors will have pressure to follow suit (obiously, AVID is a different case). The YouTube tie-in is what gives WebM more of a ´beneficient aura´ than simply the new container Matroska offered. I guess all I can say if you´re worried about fragmentation… start bitching to Adobe and other developers to start supporting WebM! 🙂
(and BTW, thanks for sharing your perspective of ´it´s got to work good in my real-life eco-system´, without that kind of dose of reality, we get open-source stuff like the GIMP)
Google hired BBB (Ronald S. Bultje, maker of ffvp8 among other things) to start working on xvp8, a derivative of x264 for VP8. He’ll start in march.
Got a link for that, or is it through the grapevine? If true, that is even better news than the dropping of h.264! I was really hoping something along these lines would happen…
Have any Editors come out in a new version in the last 7 months, you know, when WebM was unveiled?
Google is doing everything it can wrt to WebM as we speak: software implementations, encouraging video hosts to support it with the carrot ‘look, 50% of browsers will have this in 2011’, converting all of YouTube, getting browser support, devices with s/w support, chipsets with h/w support, h/w taping so anyone can make a chip…
You’d think you tried hard to find the one angle they haven’t covered yet
Have you contacted any of those vendors to see about future support?
Sony Vegas had a new version recently. There was no webm support, and there is no plans for it either, I asked. FCP and FCE new versions are coming this April. Apple won’t have support for webm, and that’s important for especially FCE that doesn’t get new versions often. Avid won’t have support for it, not even in a million years. Adobe MIGHT be the only major editor company that will add support. CS6 won’t be ready for at least another 10 months. And very few upgrade their CS all the time. Getting both Premiere and AE costs over $2000, it’s not cheap. So the penetration will be minimal. The rest of the editors don’t matter.
Edited 2011-01-14 01:18 UTC
Maybe I’m missing something here, but are you actually wanting to *edit* in webM or just export it? I ask because I tend to see webM as an export format.
In my short stint of video editing (months, not years) I remember having to use a format that worked as close to the editor and the original as possible. Using another format would cause filters to be applied slowly and an inability to do live previewing/processing. For sharing you’d compress it eh?
Comparing this to the still images, I’d say it’s the same way isn’t it? You work with images as close to the original as possible but what people view is going to be different. No matter what you’re going to lose information when you do a transference of mediums.
Talking as a consumer, it’s rare where the video I want to consume actually requires me to see every pixel. Would it be sharper? Heck yes. Do I know the difference? Only in a theoretical sense. Besides, my consumer devices alter what I see so it, uh, “looks and sounds better”.
But going back to still images example and bias. When a designer works with his/her high res image he/she has that image memorized in all it’s fidelity. Every nuance of the image is captured perfectly. When it’s compressed into a smaller format or a different colour space and etc etc to be frank it looks like crap – to the designer. The consumer gets to enjoy it on his $50 LCD screen with speakers picked up for a dollar from the next door neighbours garage sale. Do you see where I’m coming from?
I guess I’m trying to get into your shoes a bit more and experience your views.
Edited 2011-01-14 03:08 UTC
I’m no expert on this but my
opinion would be that yes I hear what your saying from a content creation/editing point of view,
that the WebM codec is less efficient leading to a variety of bandwith& decoding problems and you make sense. BUT my gut feeling is that if YouTube
and all the other most relevant Web Video sites switch to HTML5/WebM or Flash/WebM that that will in the end be ‘ENOUGH’ of a tipping point on all fronts. Hardware decoders/encoders will start being built into recording equipment, probably initially on consumer level mini HD camcorders with ‘instant upload to Facebook/YouTube’ type features and then on professional Recording and Editing equipment once the encoders are at least approximately up to current x264 standards (and I really believe they’ll get to that level in around a year! – just from following trends in the field). I’m willing to be wrong of course.
MP3 was better in many important ways. It is much more portable, easier to distribute, easier to catalog, and has a marginal production cost of essentially zero. CD was only better in audio quality, which while important, is just one aspect of total value.
For quite a while, the Wii was both cheaper and provided unique (and to some consumers, more desirable) input modes.
No idea, honestly. DVD killed both by being demonstrably better than either, and downloaded and streaming video is set to eclipse DVD eventually for the same reasons that MP3 is eclipsing CD–portability, distribution, and production costs.
SACD failed to catch on because while it improved on CD’s sound quality, it had all the same production, distribution, cataloging, and distribution drawbacks of CD. As importantly, this comparison demonstrates the difficulty of attacking an entrenched format with another that is fundamentally similar.
Raw technical performance matters, but that is only one aspect of a product’s total quality. MP3 is inferior to CD in only one way (sound quality) while it is superior in many other ways. By focusing on a single aspect of technical performance to judge superiority/inferiority, you miss the larger picture.
It makes as little sense as asserting that laptops (inferior) have triumphed over supercomputers (superior) despite their obviously limited processing capabilities… Sort of, but it turns out that being smaller than a minivan had a quality all its own.
Edited 2011-01-13 22:51 UTC
>which is more like 720p or lower video at maybe 2-4Mbit or so. If that is the metric WebM is pretty damn close to on par with h.264.
No, it’s not close to h.264. It’s visibly worse. Also, you’re forgetting how much slower it is to decode webm video. Which is as important as good quality is. If quality won’t kill Webm, decoding times for users will. h.264 is already slow to decode and people complain all the time on Vimeo and youtube for HD videos!!! Webm is many, many times slower!
>The argument could be made that for a content producer webm is mostly irrelevant. It is a last leg delivery format. While I’m sure the tools will come they certainly are not here yet.
Content producers are not irrelevant at all to the whole deal. They’re in fact a big part of the road to success for webm. WMV is a good format for example when trying to play it back on Windows Media Player, but since it’s so slow to decode inside non-MS editors, and it doesn’t stream as well as h.264, it never took off.
Edited 2011-01-13 23:05 UTC
It IS close. I have worked with it. I do not at all claim it is better or even equal, but close is a fair statement. It does have some differences (more blurring during fast motion) which are due to some of the approaches it takes. But if I take a 720p video at 2Mbit properly encoded with h.264 Baseline, webm, MPEG-4, and MPEG-2 and show them to someone, they will almost certainly rank in that order as far as quality goes – with the first two being significantly better than the last two and occasionally trading places.
I don’t know when the last time you looked at it is… It can decode 720p video at well over 200 FPS on a modern single core processor (its about 25% faster than it was at first release). If however, you meant encode instead of decode, then yes, it has a long way to go.
I did not say that. I said webm is irrelevant to you, at least for the time being…
Edited 2011-01-14 01:44 UTC
I think you meant encode. WebM is faster to decode than H264. webM is a simpler format, and it requires less computational complexity to decode.
As for WebM encoding, that hasn’t been optimised yet. Wait for the next update, called the Bali release, which is focussed primarily on encoding speed (and to a lesser extent of encoding quality), and which is due out in Q1 2011.
Ever actually tried 720p WebM on Youtube with a recent browser ? Contrary to Vimeo’s HD videos (h.264), it works in a perfectly smooth fashion on my laptop and my not-so-old 3000+/1GB/7800GT desktop.
mp3 had exactly two things going for it:
1. small file size at a time when disk space and especially bandwidth were not cheap
2. the specs were available so anyone could make an attempt at writing codecs
Keep in mind that when Winamp added MP3 support the alternative wasn’t Theora or FLAC, it was MP2. How many lossy audio formats can you name that existed at that time? Now how many were not controlled by a single company?
The success of mp3 or something like it was inevitable. Of course Justin Frankel helped (-:
While those two things are important, they are not most important. You are assuming away the fact that MP3 was the avenue to a new way or storing and using music for consumers. For that it was, as the original poster said, more or less “good enough” (it didn’t have to be perfect) though as you suggest, it was probably the best overall option at the time.
Implicitly though, you are comparing MP3 to other digital music options available in the 1990’s, rather than as the original poster did, file-based music (as represented by MP3) to CD-based music.
File-based music was just radically better and different in a lot of ways. It could be taken anywhere, delivered nearly instantly, copied infinitely at no cost, and thousands of albums could be stored and cataloged on modestly sized hardware. The only downside MP3 had was sound quality, and people were willing to overlook that to get all those other advantages.
But those were its advantages compared to CD, not to other file-based options.
Ogg, FLAC, and even more popular formats like AAC have not succeeded in killing off MP3 despite (in many cases) technical superiority because they are much the same thing, only a little bit better quality, or a little bit smaller, or a little more open or a little more something, but not enough of anything to get people to change wholesale.
MP3 was radically different than CD. It changed the way people transported, archived, and delivered music.
Ogg is only slightly different. From an end-user perspective it is just like MP3, but more open. Is Ogg taking the world by storm as a result? After all, it is, “good enough”… MP3 encoding requires licensing, rather (but not perfectly) analogous to this situation. Why isn’t everyone jumping to Ogg?
FLAC is making more headway by being better in at least one important way (sound quality) but it has a long way to go just to get a glimpse of being equal. Most people still haven’t heard of it, and WMP, iTunes, and almost all portable devices don’t support it natively, so most users don’t even know how to make it play. (FWIW, I like FLAC. I rip all my music FLAC, and if buying online, get FLAC when I can.)
On the other end of the spectrum, even a heavily-backed format like AAC hasn’t managed to unseat MP3, though it made more in-roads than Ogg or FLAC.
So, neither openness on one hand or commercial backing on the other have managed to unseat the incumbent.
(Of all of them, I think FLAC has the best long-term chance. I think the winds are blowing against DRM, and toward better archiving and convertibility, and FLAC might be the “good enough” option for that. We’ll see though, it will be a while yet…)
In a sense, I am both endorsing and rejecting the original poster’s view. I agree that “good enough” is sometimes good enough — at the start.
What I disagree with is “good enough” and coming in late in the game is still “good enough.”
A product entering an established market either needs to be radically different (MP3 vs CD) and so provide lots of new value to people, or it needs to be radically better (DVD to VHS).
Being “good enough” and the second (or fifteenth) to enter an already running game isn’t going to do it.
Even being free and open (like Ogg vs MP3) isn’t a radical enough change unless the licensing terms were very onerous to begin with, which for most commercial producers they don’t seem to be.
Edited 2011-01-14 17:09 UTC
While you are not wrong, that was all sort of my point. MP3 itself was irrelevant; “something like it,” as I said, was inevitable. The advantages of file-based music over CDs is enormous. This ties directly in to my cited advantages of MP3 over competitors for file music format: Free or cheap players, small file size. Once it won there’s little incentive to change.
Interestingly, FLAC is quietly taking over the digital audio device world. It has yet to reach many consumer-grade audio hardware but I keep seeing high-end audio systems designed for the connoisseur coming out with FLAC support, even if it’s not much advertised. This is something that will trickle down, I feel. FLAC is here to stay; I don’t forsee any near-term incentive to switch to anything else once it’s entrenched.
IMO the next big trend in music format will not be a format thing at all, but delivery. It will be what mp3.com wanted to do with its music locker, but slicker, more automatic and with mobile device support.
Edited 2011-01-18 13:31 UTC
Nonsense. MP3 won because it’s not tied to a particular type of hardware and traverses storage boundaries easily. Not because it’s ‘good enough’. Portable media formats always win over physical hardware formats in the end.
No. It’s not a quality issue. They were targeted at different gaming markets altogether.
Here, we agree.
DOS/Windows vs AmigaOS/MacOS/Unix/Novell…
x86 vs Alpha/Sparc/MIPS/HPPA/IA64…
Thanks for your perspective on this, but I think your mention of x264 is important.
x264 is an independent open source implementation of the patent-encumbered h.264 spec.
it happens to be basically the fastest implementation.
Google has very recently released it´s spec for WebM, along with it´s current reference implementation. Check out their project page, they´re very clear that encoding optimization is their current goal for the next release of the reference implementation.
Just like x264 (well, not JUST like, but equivalently in this case), alternative/independent implementations of the WebM spec are completely do-able, and should be expected to proliferate as the format is further adopted. If somebody wants to merge patent-encumbered optimization code from x264 into a WebM encoder, I don´t see why they couldn´t do so, the end product would just be subject to IP-licencing in software-patent-encumbered countries, though users already with h264 licences (tv production, etc) should be just as covered as they are currently with x264.
WebM´s advantage is that being an open format, ANY platform can include decoding, and ANY platform can include encoding, even if using a patent-encumbered x264-influenced encoder is ´optimal´. WebM is clearly the best option at this point for a HTML5 video tag standard. Saying that the obvious standard can´t be adopted amounts to saying that we can´t have an effective open video standard. I disagree with that.
Lets make sure we are talking about the same things… Google is dropping h.264 support in chrome – they have made no announcements as yet for any other products, i.e. Youtube, Android, Google TV, etc. Fair enough?
My point is for chrome this is relevent to use of the video tag and only the video tag.
“Better”, in this respect, has a whole lot more to do with quality at (relatively) low bitrates than overall quality. Pristine 1080p 50Mbit+ video over http is not practical at this point and probably won’t be for at least a decade… WebM does not need to be good at what h.264 is really good at, which is namely high bitrate HD video.
It needs to be good at what is practical to stream currently, which is more like 720p or lower video at maybe 2-4Mbit or so. If that is the metric WebM is pretty damn close to on par with h.264. I’m sure not saying it is better, but it IS really close to equal quality in most cases.
The argument could be made that for a content producer webm is mostly irrelevant. It is a last leg delivery format. While I’m sure the tools will come they certainly are not here yet.
That said, I do expect tools will start becoming available real soon that allow for exporting to webm, but I think you are putting the cart before the horse with that argument. There needs to be some demand created for those tools before anyone will write them.
THAT is the big one. The encoder has to get faster, and I mean a LOT faster before people start to jump on board with it. The codec, however, its not really the problem (although it does have a few quirks that don’t help matters) – the implementation just sucks. It will get better. But I agree, it needs to get better fast.
That is not true at all. The x264 encoder has gotten at least 2 or 3 times better in quality since it was released AND faster at the same time – all without having to come out with a 2.0 version of h.264.
There is no need for a 2.0, there just needs to be work put into the tool set. You would be amazed at the tricks you can do with a video encoding without having to change the binary format…
I would rather a fragmented web that allows me to do what I wish with my videos than a cohesive one where I have to pay for the privilege.
That top statement seems to have been lost in the (above) increasingly shrill arguments. So, to address that, I have a couple of questions for you:
1) Why are Google right to push a format you believe is doomed to failure?
2) By failure, do you mean a failure to completely replace h.264, or that is will be a flash in the pan format, soon abandoned?
3) Do you regard a fragmented web (with regards to video) to be worse than a web unified under h.264?
Expanding on the last question, it is my opinion that even if WebM cannot replace h.264 and that the best it can do is muddy the waters, it is better this than the wrong kind of unity.
Gratuitously pompous quote coming up …
“Free men pull in all directions.”
>1) Why are Google right to push a format you believe is doomed to failure?
The act itself is right, it’s just the wrong timing, and the wrong codec. But I agree that we need something that’s not h264.
>2) By failure, do you mean a failure to completely replace h.264, or that is will be a flash in the pan format, soon abandoned?
I believe it to be a format that won’t take over the web, or anything else, because h.264 has better support for ALL sides of the coin. You want TV? check. you want cameras? check. You want online video? check.
>3) Do you regard a fragmented web (with regards to video) to be worse than a web unified under h.264?
Yes. As bad the situation around h.264 patents is, I still prefer to have a unified format that works perfectly well, rather than have fragmentation.
That’s an interesting point of principle, but one which I (and apparently Galvanash; see his mod-up-able thrupence worth here: http://www.osnews.com/thread?457703) would have to disagree with.
Except h.264 doesn’t work perfectly well, for the very fundamental reason that you yourself state. The licensing and patent situation for h.264 is diametrically opposed to the W3C’s own policy and to the entire concept of an Open web.
Au contraire, history is littered with examples of not just competitive, but actually inferior formats, which one. VHS vs Beta is one example. MP3 vs Vorbis is another. For both of these the inferior format won.
If you run a Mac, to get your existing tools to support output to WebM, I believe that all that is necessary is to install a quicktime component for WebM. Download site is here:
http://www.webmproject.org/code/#webm-repositories
The encoder for WebM is indeed slow, and there is no hardware support yet for acceleration of encoding, it is a yet available only for decoding. I believe the first hardware accelerated encoder is running now in an FPGA environment, and rigorous testing is underway. This hardware encoder will become available in the first quarter of 2011.
http://blog.webmproject.org/2011/01/availability-of-webm-vp8-video-…
The first efforts at WebM optimisation have concentrated on decoding performance:
http://blog.webmproject.org/2010/10/vp8-codec-sdk-aylesbury-release…
The next release after Aylesbury due in Q1 2011 will be named Bali, and in that release the aim will be to increase the encoding speed.
I have no idea where you got this from … the WebM format is indeed frozen (i.e. “done”), but performance improvements within the existing format are going on all the time. I hope you will enjoy the Bali release dure in a month or so, it should address all your “gripes” expressed here. No need to wait for a “WebM 2.0” at all.
I accept that this is your opinion, but since it is based on utterly false preconceptions, perhaps you might wish to re-assess in the light of actual facts?
Youtube has the second largest number of searches after Google, it is not insignificant.
It would have been awesome for everybody but a few companies to have H264 as _the_ standard. The best format with the best support. Sadly licensing terms and software patents have destroyed that dream.
You can’t use H264 unless you are big enough to pay the fee. Small companies will probably start by using webm until they make enough money to pay for H264. Podcasters will have to use webm to escape the fees of H264. Games that are using ogg probably take the next step to webm.
I don’t get why it would be such a big problem having 2 codecs for video to choose from. Even intel cpus are getting fast encoders that make re-encoding as fast as copying the file to your harddrive. Your camera saves to H264. Your editor saves to H264 and when you are ready to publish you either save it to H264 for blu-ray or webm for the internet.
John Bright did have a point about there being another side of the “Open” coin, in that H.264 was developed openly by the video industry, while VP8 was not.
I think the value of this is worth as much as being royalty free for use.
H264 is not open. If you look at the patent policy from the W3C, an open standard is required to be free of charge.
H264 is an industry standard. It is not an open standard.
It’s not open in the sense that anybody can implement it, but it is open in the sense that it was developed by means of a collaborative effort by members of the video industry.
Like I said, there are two sides the openness coin: Defining a standard and implementing a standard.
And, it doesn’t really matter what the W3C defines as an open standard, as in the grand scheme of things they are only a niche standards group.
Let’s not play with words. It’s not an open standard, period. At least not as far as web standards are concerned.
The W3C is a niche standards group? Yes, because HTML, CSS and all that is so niche!
H.264 may not be open in the same sense as web standards, but neither is VP8.
Web standards are not only royalty-free standards (Like VP8, but nto h.264), but are also developed openly, with input from members of the relevant industries (like h.264, but not VP8)
So, h.264 is open in areas that VP8 is not, and vice-versa.
And, yes, W3C is a niche standards group (which isn’t the same as unimportant). They govern technologies related to the transmitting, formatting, and interacting with web pages, and that’s it. This is only a subset of Internet technologies.
Try to sell a codec with ever evolving specs! That’s simply not how it works.
Ideas that can’t be done in VP8 are kept on a side project for the next iteration. That open process is already happening.
On the contrary, VP8 is free and open. VP8 is not a standard (yet), but it’s free and open.
It is not. WebM is owned by an open-source project, which is sponsored by Google and others.
This is about video on the web. The web is probably the most important part of the internet.
Claiming that the W3C is a “niche standards group” is simply insane.
You didn’t address my other point about the collaborative development.
H.264 was developed collaboratively by industry members, while VP8 was not. This is the other side of the “open” coin.
As for W3C being a “niche standards group”, from Random House:
–noun
3. a distinct segment of a market.
Compared to, say, ISO, or IEEE, W3C is a niche standards group, with a much, much limited scope compared to larger organizations.
Au contraire, here is the WebM license page, with links to Contributers license for individuals and corporations.
http://www.webmproject.org/license/
And here is a list of the major contributing corporations:
http://www.webmproject.org/about/supporters/
I have no idea why you are making up your outrageous allegations against the open nature of WebM, but at least they are entirely trivial to rebut.
None of contributers are names big brands electricals.
Nevertheless, HTML5 is a W3C standard.
Oh, and VP8 is indeed open, read the license for yourself:
Streaming license:
http://www.webmproject.org/license/bitstream
Software license:
http://www.webmproject.org/license/software/
Additional IP Rights Grant
http://www.webmproject.org/license/additional/
Individual Contributor License
http://code.google.com/legal/individual-cla-v1.0.html
Corporate Contributor License
http://code.google.com/legal/corporate-cla-v1.0.html
Au contraire, here are the industry participants working right now on polishing the WebM codec:
http://www.webmproject.org/about/supporters/
Here is a blog so you can keep track of some of the main improvements and polish they are contributing to WebM:
http://blog.webmproject.org/
Not pushing for h264 at all, but the same arguments you make could almost be same about Flash, which Google proudly ship with its Android product, a so-called open-source project.. shipping with a proprietary product, not even a standard specified… Still not a single word on that ?
HTML5 adoption should be your objective, not just how <video> tag is supported by browsers, which again is still good to debate. But so much noise about h264, so little about Flash.. I don’t get it..
Search the OSNews archive, kiddo.
Did you read the article?
Flash is just a plugin. It’s everywhere on the web, so it’s a necessary evil. But with html5 video the dominant format is yet to be determined.
html5 video is useless if it’s based on a proprietary video format.
Plus, Flash is used for more than just video.
Oh, well.. I guess we’re not visiting the same websites to make such a statement. Starting with OSNews, Google, FB, TW… You’ve never been there i guess.
What a convenient statement. Glad not everyone are giving up so easily like Linus vs “Windows everywhere”. You just don’t want to criticize Google, and not give credit to Apple : It has made the evil not as necessary as you say.
Getting all browsers (starting with Chrome) to support full HTML 5 spec deserve more lobbying/noise than the video codec story Sorry, that’s just my opinion.
Now I wish I would have saved my earlier post for this article:
http://www.osnews.com/thread?457625
Edited 2011-01-13 22:07 UTC
Probably one of the best comments I will read this year.
Nou, ja mijn moeder is een vis, maar je zult niet eten mijn hoed.
Geen belediging bedoeld, ik moest echt Google Translator gebruiken om te controleren of uw onzin was niet de Nederlandse.
While I’m sure the patent card trolls will continue to harp on H.264 knowing the whole browser is patented and that the Internet is swimming in patents. And that patent is a government caused problem.
I’d like like Steve Jobs take on this one. And does Apple feel about WebM versus H.264. Will Apple change it’s course? Or just ignore WebM.
Edited 2011-01-13 22:19 UTC
You’ve got Gruber. That should be enough.
Gruber’s function is to validate the RDF. Not take active part in it.
H.264 is deeply entrenched outside the web.
Here is a small sampling:
http://en.wikipedia.org/wiki/List_of_video_services_using_H.264/MPE…
So what does your employer say when he learns that his corporation’s internally generated H.264 video can’t be viewed from within a Chrome or FF browser –
but can be viewed from within IE9 or Safari?
With GPU support in the OS and hardware as a given.
Think Security.
Video confrerencing.
Streaming video from certified mobile medical and industrial devices, like an ultrasound scanner.
In-house video production — and go on from there.
The same thing he would say if you were to tell him to redo the whole website because a certain somebody bars its clients from installing flash.
http://scobleizer.com/2010/10/22/starbucks-cio-shows-why-next-versi…
But this is about video on the web.
That’s a cop-out. Web video does not exist in a vacuum. It is supported by a whole host of hardware and software that exists outside of the browser.
That’s a cop-out. Web video does not exist in a vacuum. It is supported by a whole host of hardware and software that exists outside of the browser. [/q]
WebM is supported by a whole host of hardware and software that exists outside of the browser, and it is about to become many times more hardware and software.
But this is about video on the web.
The web doesn’t produce anything. The web is a distribution network.
Content providers produce content – and content providers are touchy about video quality and distribution costs – and even more touchy about content protection.
20% of peak hour download traffic is a Netflix stream.
Paid subscription content that can be routed directly through your HDTV, video game console, DVD or Blu-Ray player.
The web is not the internal network of the corporation. Where Microsoft and IE remain very strong – and where the iOS and OSX are gaining traction.
You boss isn’t going to replace dozens – perhaps hundreds – of remotely mounted H.264 cameras to keep Chrome on your desktop.
He isn’t going to replace – any – web based H.264 video app that “just works.”
Those terrible cameras existed before HTML5 video work, and come with terrible Flash apps to view them with.
If the sites to view them were coded correctly, if the browser couldn’t open a h264 stream it would see the Flash applet anyway. No biggie.
because Flash has already won.
HTML5 is just a debate for second place.
Google doesn’t have the same level of influence that they used to. Even if Google requires WebM for YouTube everyone will install Flash for the other media sites.
When did Google have more influence than they do now?
I don’t think Google are picking a fight with Flash, but rather trying to control the “second place” option, HTML5.
Google had more influence with YouTube before Hulu and Netflix showed up.
US-centric view. Hulu and Netflix aren’t even a blip on the radar of the ‘net. Nobody gives a shit about Hulu or Netflix. Their user base is <0.5% of the web. They might be all the shit in Starbucks-going, Apple-toting America, but in the real world, NOBODY CARES.
Edited 2011-01-14 15:32 UTC
If Hulu and Netflix required a WebM compatible browser that would be enough of a market to give it a foothold. The US has a lot of tech influence, it’s nothing to be bitter or proud over, just a fact of life.
And last I read Starbucks can be found everywhere but Italy. BTW eggnog lattes are only available for a limited time.
Why should those adopt anything but a combination of Flash and other proprietary apps for mobile devices ? They’re looking for DRMs, so obviously HTML5 video is not for them. No matter which codec is being used.
Lets say Google plays the nasty cards.
webm only from youtube. Any browser that don’t support it will have to install webm codec.
Then google plays this out on android as well ie only default video format webm.
Hello Google profit. Either other video sites take up webm or they have market of android out box to themselves. Money talks get use to it.
H.264 when it was first released had poor hardware support as well. New format lack of hardware support. Same with the start of all formats. But webm already has signs it will take.
Google is in the location in history todo what ever they like to webstandards. Were MS use to be. Lets hope the power does not go to Google head.
http://twitter.com/gruber/status/25699562303787009
Hilarious, HTML5 <video> cheerâ€leader failed to notice Web‣M support on YouTube months ago.
Oh, right, that’s because Gruber is just a H.264 fanboy, and doesn’t really care about the Web as long as it feeds his Apple addiction.
Depressing to see what persons tech blogs have to look up to and quote nowadays, isn’t it?
You should tweet back to him about how he railed on people for not adapting to the Flash-less i-devices. I.e., he lambasted those who stated: nothing will work on i-devices!
He has really been exposed these past few days. Notice, too, how he ignores the good pro-WebM articles. Fcuking fraud.
Oh, I’ve been tweeting him nonâ€stop, he pretty much selectively replies to my comments.
His most recent offensively wrong comment was iDevices can never do WebM without discrete hardware decoding. Which is just fibbing, webm is lower complexity, they’d throw llvm at it, and you get a whole 1Ghz to mess about with for playâ€back.
I remember encoding some (SouthPark DVD etc) early DivXs on a 1Ghz Thunderbird, and getting double digit framerates, I was ecstatic. You can’t tell me these things can’t decode much quicker. Yes, battery life would be affected. But everyone understands that for ‘s/w’ decode.
You sound like you know how to waste a lot of your day getting into useless barb tossing contests.
http://www.youtube.com/watch?v=eNxygsLGHSQ
There is no useless barb tossing contest here, it’s a case of informing ignorant people. Gruber has a large fan base, what he says, they repeat. They don’t fact check. And that damages the work sites like OSNews do.
For the record, with CoreAVC optimized for ARMv7 we never managed to play even 720p flawlessly. So don’t expect much video codec support without hardware decoding. On the other hand my Nexus S can play WebM files and it doesn’t have hardware decoding.
I have funny humour.
I read from your article, “If this is indeed somewhat reflective of Microsoft’s stance in this matter, than the company is even more out of touch with the web than we already knew.”
I then followed the link and read “The MSDN Blogs are currently undergoing scheduled maintenance.
We will be back online in a few hours. Please visit us again soon.
Thanks for your patience!
– The MSDN Blogs team
”
I think the Microsoft PR team is performing maintenance
While I agree with you, that’s a rhetorical tautology.
Humour is inherently ‘funny’.
I’m just being a jackass because I find it funny.
I actually think that “H.264 is English and WebM is Esperanto” is a much better comparison than “To me, demanding they pay up millions of dollars per year just to throw away the very ideals that made the web what it is today feels a lot like breaking the arms of the paramedic who saved your life only moments ago.”
I also think that you are using the “a known patent troll” way too much. It is really becoming your version of “let’s think about the children / terrorism” and shouldn’t be used in EVERY article you write about this topic.
Calling Firefox the “single-biggest thing to have happened to the web” is nonsense. It is just one step in a long series of developments with IE6 actually being on that list as well. When it came out it WAS by far the best browser. There were other steps before it and other steps after that (webkit, improved javascript, html5).
I like that WebM is going to be huge on the web and it will be great that hardware support and browser support is going to be improving this year. But not supporting H.264 in a browser is like not supporting jpg or mp3 because something new and exciting (but not widely supported) has shown up. It MIGHT be better in the long run, but it is unacceptible right now
What about it is unacceptable? That is what I don’t get with people arguing on this point. The facts are very simple and very clear:
1. Probably upwards of like 99% of all video served using the <video> tag currently have a flash fallback built into the code. If they don’t they are already ignoring probably 50%+ of all users…
2. Chrome has Flash built in.
3. Flash has h.264 support built in.
Therefore, Google removing h.264 from the browser itself has roughly no practical effect on anything at all. If you are serving h.264 content right now, that update that removes h.264 is going to have what effect on your users? None. Notta. Zip.
Whats the big deal really?
The big deal is that Chrome used to play this content without Flash. Than they added Flash in their installation but it wasn’t needed to play this content. And the next step is that Flash will be needed to play this content.
The whole idea behind the video tag is that browsers can play this content without needing Flash. Chrome is taking steps back, not forward.
Also, think about this: Chrome has added support for WebM and will remove H.264. They have asked Flash to add support for WebM as well, what if Flash decides to do the same as Chrome, remove H.264? If it is a good idea for Chrome to do it, why not for Flash?
Removing something that works and is widely used is just a bad idea. Everybody thinks re-encoding is so easy? You need to keep all the original files for it (LOTS of storage), re-encode all the files (lots of processing/energy) and store both files (even more storage needed). And maybe add a Flash version as well?
Would you like to keep all psd/bmp/tiff and wav files and convert them again because a brower that previously supported the jpg/gif/mp3 files will no longer do that and require you to use png/ogg?
That’s a skewed point of view: very little videos content is watchable directly in html5 without flash. Most website who started an html5 player program ask you to visit another page than the standard homepage or to change your user settings. And all the Hulu & co won’t recant on using Flash for DRM and advanced country detection reasons.
No. The whole idea behind the video tag was to foster an open standard for video on the internet. The developers for Opera are the ones who initially proposed the video tag – and it was expressly intended at the time to work with Theora – because Theora was open source and royalty free. It got corrupted in committee because of Apple, Microsoft, and others – but the original intent was for fostering an open format, escaping from the plugin dependency was only part of it.
Then h.264 won’t work anymore in Chrome by default. I never said it was a good idea for Google to remove it from chrome – I simply said it had no practical effect. The decision to remove it was theirs. I think the web is better served by and open standard for video, even if it is somewhat inferior – but at the same time I think browser makers should be free to implement whatever video codecs they would like – open or not. Google has simply chosen not to implement h.264. If you are a big fan of h.264 you won’t like that, there is certainly nothing I am going to say that will change that.
Like I said, that was their decision. I don’t know why they did it, I suspect it is in part because of the desire to escape from being at the mercy of MPEG-LA for future royalty payments. It is also in part an effort to accelerate adoption of webm at the expecse of h.264. That said, none of this changes my original point – it has no practical effect on anything at all, this whole stink is about intentions and future consequences. The act of removing h.264 from chrome really doesn’t affect anyone.
I was wrong in saying the “whole” idea. But allowing browsers to play AND control video-content natively was certainly a reason. “It got corrupted in committee” is your opinion. I think not specifying a codec (and actually allowing multiple codecs within 1 video tag) was a good idea, just like allowing multiple formats (gif/jpg/png) for images is a good idea. This committee that corrupted it according to you is the W3C I assume?
I think the web is best served with choice. Allow people (users and content producers) to use H.264/WebM/Theora/Flash. Let codecs compete on quality/bitrate first and openness/price second.
You say “Google has simply chosen not to implement h.264.” but that is not true. They did implement it, but are removing it. There is a big difference.
Finally you say “The act of removing h.264 from chrome really doesn’t affect anyone.” It affects me as a Chrome user and it affects content producers a lot. If it wouldn’t affect anyone there wouldn’t be so many people up in arms about this. It took Google half a year to re-encode most of their videos on Youtube. That isn’t a small task!
Yes, W3C. The spec was always going to allow for multiple codecs… But Opera, Mozilla, and some other members of the committee envisioned a default codec that was royalty free and all browsers would have to support it – a laudable goal and completely in the spirit of the web. Theora was the obvious choice at the time.
Frankly, in hindsight things would have been better for Apple, Microsoft and the others who fought this to just let Theora be the default and implement it – because it would have never posed much if any threat to h.264. It would have made the open source and non-commerical world happy enough and everyone would have gotten back to watching videos in h.264 more than likely. Firefox would have never supported h.264 and would have lost marketshare over it more than likely, but the other major browsers would have likely went ahead and implemented it (I expect even Opera). h.264 would have won, but Theora would have been a builtin insurance policy to keep them from getting too greedy…
Alas, that isn’t how it went down and now we have WebM – which actually IS good enough to at least make some people nervous…
So do I. But at the same time I don’t see Apple and Microsoft giving anyone any choices here. Put it this way… If Apple and Microsoft make WebM available by default in their respective browsers, Ill be the first one on board petitioning Google to put h.264 back into Chrome. Hows that?
THAT is the issue right there. In no way shape or form should the order in which you prioritized those things be applied to the web. EVER. Open for any and all to implement outweighs all other considerations – go read the W3C charter. It has to, the web cannot continue to exist otherwise – it will simply devolve into a corporate marketplace. The web is supposed to be a global public commons, do you not realize the importance of this?
It has no practical effect – you will still be able to watch the same videos you watched before. They may load in Flash instead of the video tag, but nothing becomes inaccessible.
I realize I’m splitting hairs here, so let me get right to the point. If Apple and Microsoft had simply built support for Theora into their browsers (or into their OS) we would not be having this discussion… They didn’t. Something needed to be done to level the playing field. Google buying VP8, releasing WEbM, and now dropping h.264 is the result. Its certainly not what I wanted to happen, but the gloves had to come off at some point.
Edited 2011-01-15 07:04 UTC
Chrome having H.264 first, then removing it and then adding it again if Apple and Microsoft do something? No, no, no. That is starting to sound more and more like blackmail.
The web is full of non-open and non-free technologies and there are many ways around it and the web is doing just fine. I know that royalty-free is what the W3C requires and it is a great idea, but the web isn’t just what W3C wants. For that, the W3C is FAR too slow, indecisive and fickle. A few years ago the direction was to go XHTML, not HTML5. The web itself and browsermakers are the ones that really dictate the future.
From a purely technical point of view I would like browsers to NOT support any image/audio/video technology at all. I would like them to rely on technology that is already present in the OS. That way H.264/WebM should only be included in the OS and can then be played by all media players, including browsers. Improvements will only need to be applied once and licenses wouldn’t be a problem for most users. As far as I understand Chrome is actually capable of this, as will IE and Firefox be. If it does work that way, I would actually welcome all browsers to drop support for H.264, WebM, Theora, etc and would hope that plugins/codecs for OS’s would be made available until these technologies get included in the OS by default.
I would like (web)video to “just work”, without the need for Flash (clientside) or re-encoding (serverside)
This is not the way it works. HTML5 is a W3C standard. It is a standard written by W3C. The HTML5 standard therefore IS just what W3C wants.
What the W3C wants is driven by many, many more participants across the industry other than just Microsoft and Apple. What the W3C wants is that technologies used within W3C standards must be royalty-free. W3C have an industry-wide consensus on this policy. All other W3C standards conform to this policy.
HTML5 is a W3C standard. Therefore, HTML5 must not include technologies that are not royalty-free.
That is the way that it works.
Edited 2011-01-15 13:19 UTC
There is NO current standard for video on the web! The only standard is what people did because there was (and is) no standard. THAT is why Flash-Video, H.264 and other things browser makers and content providers came up with (realplayer, activex mediaplayer, quicktime) had to be used and are now in use. This has been a PITA for the last 10 years.
http://www.w3.org/standards/techs/html#w3c_all
HTML 5 is listed as a working draft on January 13. It isn’t a recommendation or standard at all.
Dropping support for something that is used right now, to support something that will be part of a future standard is a powerplay.
(I am really hoping that MPEG-LA will shock the world by saying “royalty-free for web-use”. If that happens, would there still be a need for WebM/Theora?)
If h.264 was allowed to be incorporated royalty free into non-commercial software, both as an encoder and a decoder, and if playback of non-commercial content was royalty free as well, and if those terms were granted irrevocably and permanently then I would be content to use it.
I’m not holding my breath…
That is me though. Even if these terms were offered – it would still not meet the criteria required for W3C recommendation. It has to be royalty free for anyone, even commercial endeavors. And it would definitely not be good enough for someone like Mozilla.
So to answer you question, yes – there would still be a need for WebM/Theora.
Just so it is clear, to be acceptible for the W3C doesn’t mean all encoders and decoders have to be royalty free. Here is the summary: http://www.w3.org/2004/02/05-patentsummary.html
The really important thing to note is “xxx” that everyone seems to focus on. But “The license may be limited to implementations of the Recommendation” means that if the recommendation is about “playing back video in a browser” than the royalty-free license only needs to be royalty-free for “playing back video in a browser”. Encoders wouldn’t have to be royalty-free, tv’s with decodeders wouldn’t have to be free, etc.
Of course, the current license for H.264 isn’t royalty free and WebM/Theora is, but my point was: What if Mpeg-LA decides to drop a bomb and DOES make H.264 acceptible for the W3C.
It must be “royalty-free for any party to implement codecs, and royalty-free for all kinds of web use including video providers and commercial content”.
I would also like to see “perpetual and irrevocable” wording in the license grant, just as there is in Google’s license grant for WebM.
Then, maybe. H.264 would still have the problem that it is computationally more expensive than WebM, often requiring dedicated hardware acceleration where WebM would not.
PLEASE provide a link that says that H.264 is much more processing intensive (encoding and/or decoding. I think it is exactly the opposite and Eugenia seems to think the same. http://www.osnews.com/permalink?457711. There were also stories about a big video site NOT re-encoding because the re-encoding was just WAY too slow.
H.264 DOES have a lot of hardware support, but how is that a bad thing? It speeds up the decoding and saves battery life. Hardware support for WebM is just starting and it will take years to reach the same penetration level as H.264.
WebM is much slower to encode than H.264, but much quicker to decode.
Encoding WebM is not yet optimised, and the encoding takes many times as long as an equivalent H.264 video.
But decoding is the reverse. On most platforms hardware acceleration assistance is required to decode H.264 at acceptable speeds for anything other than the smallest of picture resolutions. Flash with H.264 at 1080p, 720p or sometimes even just 480p brings down most machines unless hardware acceleration of H.264 decoding is used.
WebM, on the other hand, can be played by most CPUs without hardware acceleration at quite high resolutions.
It is on page 5 of the PDF
http://www.streamingmedia.com/conferences/west2010/presentations/SM…
Simple design, low underlying computational complexity.
Also here:
http://blog.webmproject.org/2010/10/vp8-codec-sdk-aylesbury-release…
Improvements in decoder speed were made over the version of the original release announcement, but encoder speed has not yet been worked on.
The theme for the next release, called Bali, due in Q1 2011, will be a faster WebM encoder. WebM certainly needs that, because right now encoding is glacially slow.
Edited 2011-01-17 14:14 UTC
Really – who cares. h264 is fully supported either through flash (which is part of the Chrome and fully supported on Firefox) or through codec plugins, so h264 can trivially be supported even through HTML 5 in browsers which aren’t IE or Safari.
What’s relevant is always – content. Google for now provides both WebM and h264 on youtube. As long as this is true, everyone has a choice, the real difference is what you can play ouf of the box on certain web browsers.
At least h.264 has now lost the “market” battle to become a de-facto web video standard, but they could still try to start a patent lawsuit war. On2 was a good and logical investment by Google as there was a risk of catastrophically loosing control over video formats (in ~5 years) which were a foundation of one of the most important internet services.
Can I please remind you all that an attack on Googles decision in this case is not necessarily an attack on WebM.
It would be awesome if all browsers had WebM. What is not awesome is Google removing choice from users and content producers for political reasons. Also, neither Thom nor Kroc or anyone speaks for all the open source community, there are plenty of people out there perfectly happy to have the different codecs compete on their merits.
It is certainly Googles choice what codecs to include in Chrome, but it is perfectly legitimate to complain when Google takes away choice for political reasons that a lot of the user-base doesn’t care about.
They can choose another browser.
Simple analogy to help Microsoft understand.
h.264 = CP/M
WebM = DOS