With YouTube and other video sites serving up over a billion streams a day, it’s beyond contention that web-based video is not only mainstream, but has become fundamental to the web experience. Why, then, is a huge majority of web video in a wrapped in a proprietary Flash candy coating — essentially making Adobe the gatekeeper of video content? It’s worked okay so far, but it’s hardly a fertile ground for innovation, not to mention the fact that Flash is a real dog on OS X and any kind mobile browser (if it’s even supported).
The next iteration of HTML standards is poised to introduce a <video> standard, putting moving images in the same natively-viewed category as images and text. Flash video has become so ubiquitous that you hardly think about it, but we all get a reminder every few months or so when we have to upgrade or re-install the plug-in, and the continuing difficulties with .flv support offline show that Flash is far from the ideal delivery method for such a (now) basic resource.
Unfortunate, it seems that the powers that be (heavyweights Apple, Mozilla, Microsoft, Google, &c.) can’t agree on what format the <video> tag will indicate. The battle is between the reigning champ, H.264, and the open-source alternative, Ogg Theora.
Whatever, let the format geeks work it out, right? I’m afraid not: there’s more than image quality and codec efficiency on the table here. H.264 is a property of the MPEG standards organization, which places it somewhere east of proprietary but west of public. Whatever its status is (I don’t pretend to understand exactly), it’s not free, and although it’s well-maintained and extremely common, many think that implementing a patented technology for a fundamental standard is a bad idea when there is an alternative.
And that alternative is Ogg Theora. While the Ogg formats (maintained by Xiph.org) haven’t taken off in popularity when compared to their MPEG cousins, they’re competitive and have the very attractive quality of being free and open source. Recent statements by Google’s __ suggesting that Theora is simply not efficient enough have been challenged, although it seems to me that there would certainly have to be some work done if Ogg were to roll out its format as a standard on this scale. Dailymotion has a corner of its site (you’ll need a compatible browser like FF 3.5) dedicated to using HTML5 and the <video> tag, but they admit that neither the audio nor image is up to snuff quite yet. It’s worth mentioning that The Video Bay is dual-wielding HTML5 and Theora as well, but to say its future is uncertain would be somewhat of an understatement.
Not easy, is it? Damned if you do, damned if you don’t. The devil you know or the devil you don’t know… and you have to pay for the devil you know. Plus, of course, I’m simplifying everything to my own level — and I’m decidedly not a developer. Personally I’m rooting for an open format (I suppose x264 is out because it relies too much on H.264), and I’m sure a little elbow grease would shine Theora up but good. I’m also unsure as to the possibility of supporting multiple formats, as the <img> tag and others obviously do (I may be missing something here).
Last, who’s to say that competition would be bad? You’ve got your open standard, free to all, and you’ve got your (perhaps slightly better) closed standard, easier to use and with better support. Fight! As long as it’s transparent to the user and it doesn’t stifle innovation, that sounds like the kind of rumble I can get behind.
This isn’t really a new battle, nor is it likely to be resolved any time soon, but discussion is ongoing elsewhere on the internet (such as at Ars’ excellent examination of the legal issues, and of course Reddit) and we may as well bring on over to the Crunch (again). Any codec nerds or patent-mongers care to chime in?
To hell with the video tag. Not worth this foolishness.
To hell with proprietary flash players that all have different interfaces, and have a tendency to occasionally hang my browser.
To hell with both of you, flash sucks because it relies on an external proprietary library and Adobe is stupid, but how would the video tag be rendered? By the browser? More bloat in FF because they need to maintain a video player as well? I think not, but depending on external libraries sucks as well esp on Linux. VLC used to be teh awesomeness, but it seems kind of stupid lately. I guess I’ll be stuck with gstreamer. I hate pebkac technology! I was happier with HTML 3.2 index pages and downloading mpegs from ftp servers
To Hell!
That’s a good question – I’d hope that there’d be a basic hardware solution, all video information being rendered through the video card. That’s a bit much to hope, though.
I disagree… Even with support of 4 out of 5 major browsers, it’s light years beyond what we have to do today to get cross-browser video working. Have you ever tried to get a video from your computer on the web without using YouTube?
Heck, even if we have to encode twice (mp4/ogg), it’s *still* easier than today’s tag soup of video objects, embeds, etc, hoping that you cover all your bases of mobile/desktop browsers.
Basically, all we need to do is this (hopefully TC won’t eat my tags, so bear with me):
This assumes that IE won’t be supporting it for a few years ago. Once IE is in the mix, it’ll probably boil down to:
If Apple stopped digging its heels in on the HTML5 list and just supported OGG, you could do this:
Oh great… TC silently ate my markup!
Let’s try again. Here’s the markup for supporting IE + video-capable browsers. Once IE is in the mix, we can probably nix the OBJECT part of things:
<video>
<source src=”video.ogg”>
<source src=”video.mp4″>
<object type=”application/x-shockwave-flash”>
<param name=”movie” value=”movieplayer.flv”>
<param name=”video” value=”video.mp4″>
</object>
</video>
seems like a good solution… except that most people in video industry are on VERY old browsers and i doubt there is will be cross-platform cross-browser support for any time soon. the reason everyone uses flash to display videos is simply the amount of flash players installed. it solves an issue of making crap work for every browser with native plugins… ever tried playing mov in windows xp without quicktime installed? have you tried playing wmv on mac or *nix without proper codecs installed? non-technical people dont want to install crap they dont understand on their computers and thus developers use flash players. i can only dream about support in 90% of browsers so i can finally stop using flash players! hell… if i have to encode videos in both ogg and h264, im fine with that as long as it WILL failback to proper codec without prompting user to install anything.
bleh… TC stripped my tags and my post barely makes sense without them
I’d say forget the object tag and just add the fallback with JS. Methinks the best way to vote is with your markup
BTW, if anyone has trouble with OGG in FF3.5, I recently realized the dimensions have to be in multiples of 16.
Nonsense. The problem with Ogg is that is is far from ready and will likely not be any time soon. Google has stated that to convert to Ogg Theora, it would require the total bandwidth of the Internet to stream the billion streams a day they do now with that format.
Ogg is old, with questionable patent liability and the open source true believers are being disingenuous to claim otherwise. Apple and Google both know that they open themselves to massive liability if they were to adopt it. And it does them no good, since they already pay for h.264. There are plenty of free h.264 players that rely on Apple’s paying its h.264 license already. If you have Quicktime, and you do if you have iTunes, then you’re already covered.
This is pure FUD. When Thomson claimed that Ogg Vorbis was infringing their patents (when it wasn’t), they were made to eat their words. There have been zero claims or even specific rumors that Theora infringes any particular intellectual property rights. The source is open, and all rights holder have the ability to compare it to their patents. Not one has even hinted that they believe there is any overlap with their patents. So you can spread misinformation, but others should realize that it is only misinformation.
That statement by google is widely questioned…
The video tag allows for multiple sources so you can say, for example, <video src=foo.ogg src=bar.mp4>. Browsers will use the first source if they know that codec and if not, will try the second source.
But it’s not quite the same as with images and multiple formats because you can export an image into two or three formats in a matter of seconds with little processing power, while encoding video is a much longer and more processor intensive task.
True, though it doesn’t stop the big guys like YouTube from re-encoding in several formats.
Devin, absolutely.
I’d love for the big guys, an it’s completely within their capabilities, to offer a Theora+Vorbis in Ogg version for Firefox, Opera, and Chrome, and then offer H.264 in .mp4 for the other <video> supporting browsers, and finally H.263 or H.264 in Flash for the no-<video> browsers.
H.264 is much more better for optimized streaming for higher quality video.
Define “much more better”?
http://people.x...comparison.html
http://people.x...merten/youtube/
Much better, as in over 60% better.
http://www-cs-f.../theora-soccer/
Did you read those comparisons? Using YouTube as the H.264 encoder? Giver me a break. Everyone knows YouTube’s encoding quality is crap, most likely because they trade-off encoding speed for quality. Interesting too that the videos chosen don’t feature traits that Theora is crap at encoding like fast motion. Pointless comparison.
Exactly, you provide a truthful test of H.264 encoded with slow care against ogg.
Nothing against ogg but H.264 performs.
Stream service providers don’t want to use all the h.264 power because it takes time to encode and decode. One aspect none of the comparisons makes is the time it takes to encode / decode the videos.
Very interesting topic. I’ve been following this a bit lately and I’d love to see H.264 win out but I can definitely understand the concerns surrounding it. Also, article could use a little proofreading.
You’d love for more vendor lock-in and for open source browsers like Firefox to not be able to implement support for Web video?
Firefox has been able to display video since the beginning. People choose how they display video and Flash has been the most popular choice.
The notion of tie-in applies just as well to Ogg, if not more so since there is no other use case except in HTML5.
I’d also love to see H.264 “win,” which would clearly require reasonable and non-discriminatory licensing of the underlying patents per the insanely comprehensive documentation of the matter Ian Hickson put together last week (http://lists.wh...une/020620.html).
I’d also like a pony and world peace, but I know I won’t be getting either of those any time soon, either. Oh well.
The biggest problem isn’t with Ogg Theora competing against H.264 – Its the damn name!
Ogg?? OGG? Really.
Now, if it was named something like OT.29993 or equally boring and byzantine, we’d have a hit.
amen.
Branding isn’t open source’s strong suit. I’d love to help them with that.
Also, I think your use of terms like proprietary, free, and open source miss the big issue here and that’s patent licensing fees.
You can have a free or open source implementation (x264), but in countries that enforce software patents, you can’t use or distribute it without paying the patent holder.
Right now, the W3C has a policy of avoiding, where ever possible, codifying Web standards that would require licensing fees.
So, H.264 cannot and should not be written into the standard because to implement it requires paying MPEG-LA licensing fees.
Theora+Vorbis in Ogg require no such licensing fee AND there are open source implementations AND it’s plenty good for the overwhelming majority of Web videos currently in circulation. Oh, and because it’s open source and has the support of a dedicated developer community, it’s getting better all the time.
Agreed about the mixing of terms. Just because Ogg is open source and no one is trying to collect licensing fees right now does NOT mean it is not infringing on anyone’s patent. With H.264, the terms for licensing are known. With Ogg, the hammer may fall at some later date.
Neither format (or indeed any software) is immune to submarine patents.
Ogg Theora could be viable, but at this point in time that’s not the issue. The key issues are:
1) it’s only been a few days that a non-beta browser can support the video tag, so unless you’re a JS/HTML5 fanboy it doesn’t mean anything.
2) converting existing video to the new codec would be a monumental (if not prohibitive) task.
3) the biz of video codecs is not kid’s stuff – the video codec battle has been going on for over a decade. Apple and Microsoft are still very deeply invested in proprietary codecs (and related hardware). Flash video was lucky to break through at the right time.
4) Flash only recently (2007) started supporting H.264 because that codec has a large support-base from premium content producers.
5) Ogg doesn’t offer 10x value over H.264 as H.264 had over Sorensen, etc. It’s only value is open source for the sake of.
Having done enough work with embedding media in web browsers I can’t stress how excited I am for the video tags. Plugins are fickle creatures and are compounded by browser differences and wacky standards (cough cough ie).
As far as the formats go, from a developer standpoint open source is so much easier to deal with. I’m still sitting here waiting for the whole mp3 license shenanigans to blow up.
i think that the video tag is just fantastic, as a video content producer, i’m tired with flash, no flash, and as another person said earlier, flash is just incredibly slow on mac, i just hope all the major devices are going to support html5 asap
There was an article about this on ars today. He one point they made that was missed here is that there is a possibility that further development of Ogg could be hindered by unforeseen patents on other technologies. As was mentioned above, it’s a decision between going with the known and havin to pay for it, or going with the unknown and hoping for the best.
What will likely happen is that he video standard will go through as is, and the market will decide which is the better standard. It worked for XHTML vs. HTML.
Why does there need to be a video standard? There isn’t an image standard?
HTML isn’t supposed to endorse one image format over another. It shouldn’t endorse a video format either…
There is a major difference between the image issue and the video issue. While there is no standard for images, all browsers support the use of at least the three major image types (jpg, gif, png).
With this video standard, there are some browsers willing to support one choice and not another. This would create a major headache for developers and content producers, as they would always have to serve up two video formats for each video.
Flash would be better i thought .!
Flash isn’t even a video format.
there’s a good background article on this with excellent tech info as well: http://people.x...comparison.html
there’s a good background article on this subject with some excellent technical insight as well, if anyone’s interested: http://people.x...comparison.html
Anything is better than FLV on the Mac. FLV on the Mac is like going to the movies in the 1920s. End the Adobe rein of terror!
I have constant problems with flash on linux and would love to see video move to open source tech, but it’s not like this will _change_ anything, flash is the de facto standard. What might be lost is the opportunity for the W3C to get behind a push for more open tech.
“It’s worked okay so far, but it’s hardly a fertile ground for innovation, not to mention the fact that Flash is a real dog on OS X and any kind mobile browser (if it’s even supported).”
I think that’s kind of a ridiculous statement. Do you remember video on the web before Flash? Real Player and Windows Media? Video on the web became mainstream because of better bandwidth and because Flash made it easy for everyone to use. We continue to spend a lot of time and money to innovate around web video.
And how many mobile browsers support the video tag? We’ve got H.264 support in every phone that supports Flash Lite 3.1 which is a number of devices including the HTC Hero with Android.
Advancing web standards is always a good thing but Adobe has pretty good web citizen when it comes to video. We’re paying the licensing fees so that anyone can consume high quality H.264 video.
=Ryan
ryan@adobe.com
Agreed. Also, I’m puzzled why people are spreading lies about Flash being slow on the Mac. I’ve done numerous Frame Per Second tests on both platforms and don’t see a significant difference (on 9 and 10). Flash 7 years ago was a different story however. Anyway Hulu is Flash done right. It works great on my both of my Macs.
Looking forward to Flash 10 for mobile early next year.
I have to agree with Ryan.
We’ve built a business around encoding into Flash, and there is no debate about the ubiquity of Flash video.
Granted the open source standardization impetus is necessary to keep Ryan and his team on their toes, but I have to say they’re doing a fantastic job already, WRT platform independence, video quality, and accessibility.
I have seen Internet video in the days of Real, Windows Media and even player-less Java based streaming, and those memories are best left buried.
Sure, Flash is better than what there was before, but that’s like saying being punched in the face is better than being stabbed in the face. Besides, flash would likely continue to be supported for rich video players (chapters, annotations etc) but for basic video display there really should be a standard tag and at least guidelines for a format.
I agree with the author that flash has become a hindrance instead of an enabler. The fact that the servers spends countless hours converting video from one format to another is also an issue. Hopefully the new HTML 5 will offer a better solution to upload/download than ftp too.
If the H.264 folk were smart they’d encourage browser vendors to support H.264 and license it to them for free. Why? Because that would ensure that H.264 becomes the dominate standard, and open the floodgates to users creating and uploading and playing H.264 video.
In the meantime, the H.264 group makes its money off the hardware guys, as now every computer, notebook, phone, and media device will need low-power dedicated H.264 hardware decoders.
Further, many H.264 hardware decoders exist NOW in graphics cards, notebooks, and even the iPhone. Ogg? None.
Keep in mind the production side – most video cameras are using H.264 (AVCHD is a H.264 variant) and the great things with H.264 is the profiles allow support from low power devices through to studio equipment. So really Ogg has a lot of ground to cover to compare.
As far a Flash player goes I have always found it works great and upgrading is easy.
The kicker here is Microsoft. I hate IE as much as the next guy but you can’t really move forward with web video when half or more of the users are on a browser that can’t support the video tag. Hopefully IE’s market share will continue to erode but until it is reduced a sufficient amount Flash will remain the de-facto standard for web video for good or for bad.
H.264 is already supported under Quicktime, and Windows 7 will natively support H.264.
http://www.zatz...y-support-h264/
The HTML guys can access the codec APIs and get H.264 for free.
Much ado about nothing.
Even while having to worry about the two formats, I was rather mad to find out Firefox 3.5 didn’t full integrate the video tag with their release. Both the loop and poster attributes currently do not work on Firefox 3.5… and those were two I wanted to try and use in an experiment!
How are features and capabilities going to be added to the HTML spec for the video tag? (such as adaptive bit rate switching, multiple audio tracks, etc…).
When these features are rolled into the spec and adopted at different times across different browsers and platforms, won’t we end up with a horrible mess of inconsistency?
Adobe (love them or hate them) has done an amazing job of providing consistency across browsers and will continue riding this success into the mobile market with OSP and FP 10. I dream of the day that I can build a flash app and have it run consistently across desktop, tv, set top box, mobile device, etc..
I do find it interesting that there is so much convo about either the ‘overwhelming’ enthusiasm for the flash streaming model, or whether h.264 is better than OGG, yet practically no Publisher is profitably delivering video on the web. Can you even consider Internet Video an ‘industry’ when it loses money hand over fist?
A profitable solution needs to address two issues simultaneously to have any chance at profit. 1) Consumer Access 2) improving the delivery cost vs. advertising value ration. The ubiquity of flash makes access easy for consumers and there are lots of tools, making it easy to create the content (and of course the tools is where adobe make their money), but the cost to value ratio of flash streaming is unsustainable. If you provide advertisers no better value than display advertising, they will not pay the premium required to cover the BW and content creation costs. H.264, flash, Ogg/Theora, it don/t matter.
My company is trying to find another way through audience insight and testing…..and our solution supports OGG/Theora/Vorbis because you don’t have to go through Adobe or the MPEG foundation or pay their tolls. Hopefully at the end to the day we can provide Publishers and Advertisers additional value. Here is a link to a download our Ogg/Theora player with DRM that supports our ad testing solution. http://www.noss.../downloadPlayer
Well, despite the lack of returns on internet video so far, it’s going to be an industry whether people like it or not, so I think it’s important to establish some standards like this first, and Adobe and the MPEG group aren’t going to make it easier in the long run.
Couple of points:
1) The tag in html5 is a really big deal for developers. The tag too. They become part of the DOM so they can be addressed in various interesting and useful ways.
2) H.264, .flv, OGG, etc. in their current state of development are but experiments on the way to ubiquitous and functional video on the internet. The public web is only about 15 years old, and multimedia on the web only about 10. The question is, 10 to 15 years from now where do we want to be with online media? The tag will be universally supported by browsers and mobile devices, which will also support standard formats. Standards matter a great deal, and as a developer I would rather spend time with open standards.
Shoulda realized my html tags would get stripped! What I meant to say above is:
1) The VIDEO tag in html5 is a really big deal for developers. The AUDIO tag too. They become part of the DOM so they can be addressed in various interesting and useful ways.
The statement of Google to use the entire bandwidth is nonsesne. Flash video exist before html5 and it will exist after it. So, what´s the problem?
Apple and Google are at h.264 side because of Apple iPod and iPhone. The statement that “ogg lacks hardware” it’s nonsense too. If a z80 can play ogg audio, than a smaller adjus for new hardwares can play Theora easily, and Theora it’s lighter than h.264 in consumig procassor resources.