Slashdot Mirror

That's OggKate's job. It also works with any other Ogg embedded video codec.

Yes, and when he was called out on his BS and FUD ... he promptly disappeared.

... the video functionality that, while officially removed from the HTML5 standard, will be implemented by everyone anyway...

I really don't know where this urban legend started and why people believe it, since it's trivial to verify that <video> has never been removed from HTML5.

What has been (hopefully temporary) removed is the mention of Ogg Theora as baseline format since Apple and Microsoft haven't yet accepted to implement it (Safari supports it anyway with the XiphQT component installed). OTOH, Mozilla, Google and Opera all support Ogg Theora (and Vorbis for audio) in their browsers (current of future versions), so apparently Theora is still the strongest candidate, altough Google may change this if they buy On2 and free the VP8 codec.

P.S.: sorry fanboys, H.264 is not an option: starting from 2011 websites with H.264 videos will have to pay an unspecified amount of money to the MPEG LA.

Chrome 3 will include a Theora decoder ... a known broken and crappy one from an old FFmpeg build that can't cope with Thusnelda-encoded files, i.e. the close-to-H.264-quality encoder that Xiph and Mozilla have been working on.

They know about the bug ... but can't be bothered fixing it.

So sites with lots of Theora video will have to browser-sniff and suggest Firefox 3.5 to those with Chrome.

How to snatch defeat from the jaws of cluefulness ...

(note also that Chris DiBona mysteriously vanished from the WHATWG list after his FUD was refuted. It would be interesting to hear why.)

Actually, Apple might well implement Theora, but they are worried about the patent issues. Amongst other things, that the patent agreement from On2 for VP3 might not be watertight. See this post by Apple employee on the Xiph mailing list - it really shows Apple's stance and worries on this issue.

Google can now use On2 codecs such as VP8 in YouTube, for free. No more royalties. But the royalties are not that expensive so this isn't likely a big deal for them. (Google could save more money by using smarter settings on their H.264 encoder.)

Do you think Google will seriously try to make money by selling codecs? I don't. $100 million is small change to Google, and if that's all it cost to buy On2, then the On2 revenue stream must be trivial by Google's standards.

So, Google won't save much money and won't make much money by buying On2. I think they are up to something else.

What I think is more interesting is the possibility that Google will give On2's latest technology to the Theora guys. Just as Sun started giving away OpenOffice.org after buying StarOffice, it's likely that Google will give away some or all of the On2 technology.

Despite being based on technology that is nearly a decade old, Theora is already fairly competitive for web video. (Theora is better than H.263, which has actually been used for years, so it's difficult to argue that Theora is not usable for web video.) Now imagine that Theora gets the best technology bits from a modern On2 codec, and integrates those, such that Theora really is as good as H.264, or even better.

Now imagine that this improved Theora is bundled with Google Chrome and Firefox, bundled with Android, and bundled with Google Chrome OS. Within a few years, Theora could become firmly established everywhere as a baseline standard that anyone can use.

Google likes things that make it easier for Google's customers to use Google's services. They like their customers not being locked into proprietary technologies not owned by Google. It will be impossible for Google to take the market away from H.264, but it is very possible that they could make sure their customers can always easily access their services.

Note that this scenario utterly depends on the new Theora being free software. Google could try to sell a proprietary On2 codec and gain a significant market share; well, if they try it, all I can say is "good luck with that." It's hard to push out an established standard; to do it, you need to be significantly better, not just a little bit better. Better technology, with Google behind it, completely free (and with no need to even keep track of how many codecs you ship out) might succeed.

steveha

I thought the whole point of HTML5 video was to have it work in every standards-compliant browser...

It was. It was specified that browsers playing HTML 5 video should support Ogg Theora, the codec that Firefox 3.5 among others uses. But it was removed after pressure from Apple and Nokia.

On the bright side Ogg Theora is supported by Firefox 3.5, Opera say they will support it, Google Chrome is on board, Safari can be made to support it by installing Theora codecs for QiuctTime and there are ways to make other browsers support it as well. So the problem is solved on the publisher side: publish in Ogg Theora. Hopefully this will put enough content coded in Ogg Theora out there to make it a de facto standard that solves the problem on the viewer side as well by pressuring webpages like Youtube to offer it and Safari and IE to implement it.

I don't understand the inferiority justification, side by side they look indistinguishable to me.

http://xiph.org/quicktime/

Adds support for Ogg Vorbis and Ogg Theora to QuickTime (which is used for nearly all media playback on OSX). Easy to install (but could be made easier easily - such as making into a .pkg), and makes Safari 4 work with <video> and Theora.

Also, can we please stop whining about this in relation to the HTML5 spec? HTML has never specified file formats for media/objects (<img>, <object>) and it should *not* start now.

and the fact that it's far worse quality than h.264.

FUD. How about this for clarification http://people.xiph.org/~greg/video/ytcompare/comparison.html

You're right as for there being no hardware support for decoding Ogg Theora. I don't know enough about that to make a comment (I wonder if it is possible to make such a thing but whether or not it just hasn't been implemented). As for the rest though, the quality argument is simply not true... it looks as if in some circumstances, in fact, theora comes out on top. But even if that isn't true, we can see that it's close enough that it isn't a significant difference.

As for the submarine patent stuff, that's FUD... every codec technically has that threat. But Theora is the only one not known to have any current patent issues. h.264 has several known patent issues, but of course Apple is not worried because they are in control of that. But what about everyone else? In fact, unlike Theora, where steps have been taken to avoid patent issues, the dangers of patents are already known when it comes to h.264.

Please don't spread this obvious bullshit. One codec may have patent issues but nobody can find them. One codec has obvious and known patent problems and may have even more that nobody has found. If you're going to make an attack on the former for patent issues, you'd better not be supporting the latter.

The plugin you're referring to: http://www.xiph.org/quicktime/

(They finally released an update last month after two years of no activity. I think this HTML5 argument finally spurred someone into thinking "Hey, maybe we should provide (regularly updated) support on a fairly widely-used platform.".)

In the case of Safari, it does. Safari uses QuickTime to play videos, so you can use the <video> tag with any format that QuickTime supports. If you install XiphQT, then that includes Ogg Theora (I just checked, the video here plays just fine in Safari 4 with XiphQT installed).

I have looked at some clips posted with both VLC and Media Player Classic using CCCP. Theora sucks versus H.264. The amount of noise in the captions and images, it is like the difference between looking at an extremely compressed and noisy JPEG, and a slightly blurred PNG. So what gives? Does my player suck, or is this really the best its supposed to do? Sure it looks better than the blocky H.263, but that isn't saying much.

Vorbis did make some inroads into the game market, though. It's not all that rare that I see a copyright notice for Xiph in the opening of a game. The most recent example is Ghostbusters.

A list:

http://wiki.xiph.org/index.php/Games_that_use_Vorbis

Everything from Halo (Mac/PC) to Guitar Hero (II) and Rock Band to GTA to Quake4 / Doom 3 to Devil May Cry and all sorts of games in between.

There are three things that this test doesn't consider:

1. for the same bitrate (1000 kbit/s) the Mpeg-4 file is 5.2% bigger than the Ogg one;
2. nobody uses video alone like in this test, there's always audio and the audio codec associated with Theora (Vorbis) rocks: same quality as MP3 for half the bitrate. Bits saved on the sound can be used to improve the video; and, yes, it is apples-to-apples comparing the overall bitrate of Ogg/Theora+Vorbis against an all-Mpeg-4 solution.
3. but the most important detail is that they used a constant average bitrate encoding with Theora, which is known to give inferior results for the same bitrate to simply setting the quality to match the desired bitrate.

For real life examples, that also include sound see "YouTube / Ogg/Theora comparison" and "Another online-video comparison".

There are three things that this test doesn't consider:

1. for the same bitrate (1000 kbit/s) the Mpeg-4 file is 5.2% bigger than the Ogg one;
2. nobody uses video alone like in this test, there's always audio and the audio codec associated with Theora (Vorbis) rocks: same quality as MP3 for half the bitrate. Bits saved on the sound can be used to improve the video; and, yes, it is apples-to-apples comparing the overall bitrate of Ogg/Theora+Vorbis against an all-Mpeg-4 solution.
3. but the most important detail is that they used a constant average bitrate encoding with Theora, which is known to give inferior results for the same bitrate to simply setting the quality to match the desired bitrate.

For real life examples, that also include sound see "YouTube / Ogg/Theora comparison" and "Another online-video comparison".

See this comparison and this comparison to see how Theora compares to the most popular real-world implementations of H264 on the Web.

WP:NPOV violation

From http://xiph.org/:
Our purpose is to support and develop free, open protocols and software to serve the public, developer and business markets.

See this comparison and this comparison to see how Theora compares to the most popular real-world implementations of H264 on the Web.

WP:NPOV violation

From http://xiph.org/:
Our purpose is to support and develop free, open protocols and software to serve the public, developer and business markets.

See this comparison and this comparison to see how Theora compares to the most popular real-world implementations of H264 on the Web.

WP:NPOV violation

From http://xiph.org/:
Our purpose is to support and develop free, open protocols and software to serve the public, developer and business markets.

Why is anyone worried about the quality of videos your going to watch in your browser? The vast majority of those videos are not going to be interesting enough to want to see them in full HD glory.

That's a bullshit assumption. I watch streaming video on my TV from sites like Hulu and YouTube, and the higher quality video, the better.

Yes. You are part of a small but growing minority but even so I expect that Theora+Vorbis high-def content at comparable to H264+AAC bitrates would be satisfactory. Check for yourself here if you've got a Firefox 3.5 pre-release.

For the overwhelming majority of Web video, and for the overwhelming majority of Web users, though, Theora+Vorbis won't just be satisfactory, but will be virtually indistinguishable from what users are already quite comfortable with.

Why is anyone worried about the quality of videos your going to watch in your browser? The vast majority of those videos are not going to be interesting enough to want to see them in full HD glory.

That's a bullshit assumption. I watch streaming video on my TV from sites like Hulu and YouTube, and the higher quality video, the better.

Yes. You are part of a small but growing minority but even so I expect that Theora+Vorbis high-def content at comparable to H264+AAC bitrates would be satisfactory. Check for yourself here if you've got a Firefox 3.5 pre-release.

For the overwhelming majority of Web video, and for the overwhelming majority of Web users, though, Theora+Vorbis won't just be satisfactory, but will be virtually indistinguishable from what users are already quite comfortable with.

Unfortunately, Theora still needs twice the bitrate as H.264 to deliver the same quality, even with the "Thusnelda" rewrite of the encoder.

That's not my experience. See here and here for some real-world comparisons. H264 can be better than Theora -- though not majorly so, but in the real world at sites like YouTube, H264 doesn't stand out from Theora for most cases.

Here are the only two real-world comparisons I know of. Both have Theora making a pretty good showing. Beating H264 in all cases? Not at all. Good enough to replace H263 and H264 at YouTube, absolutely.

http://people.xiph.org/~greg/video/ytcompare/comparison.html

http://people.xiph.org/~maikmerten/youtube/

Here are the only two real-world comparisons I know of. Both have Theora making a pretty good showing. Beating H264 in all cases? Not at all. Good enough to replace H263 and H264 at YouTube, absolutely.

http://people.xiph.org/~greg/video/ytcompare/comparison.html

http://people.xiph.org/~maikmerten/youtube/

"Unfortunately, Theora still needs twice the bitrate as H.264 to deliver the same quality, even with the "Thusnelda" rewrite of the encoder."

Except that statement is provably false if by no other facts than that neither Theora nor H264 quality scales linearly with bitrate.

Beyond the obvious fail in your claim, you're also just wrong.

See this comparison and this comparison to see how Theora compares to the most popular real-world implementations of H264 on the Web.

"Unfortunately, Theora still needs twice the bitrate as H.264 to deliver the same quality, even with the "Thusnelda" rewrite of the encoder."

Except that statement is provably false if by no other facts than that neither Theora nor H264 quality scales linearly with bitrate.

Beyond the obvious fail in your claim, you're also just wrong.

See this comparison and this comparison to see how Theora compares to the most popular real-world implementations of H264 on the Web.

If Theora wants to be taken seriously they need a GPU based hardware decoder than works on the big three, Intel, ATI, Nvidia, and they need it yesterday, and they need to start offering it to the GPU manufacturers so they can bundle support like they do for WMV9, Divx, and H.264.

There *ARE* hardware chips developed.

But for GPUs, instead of having yet another bunch of transistors dedicated to yet another hardware accelerator eating up valuable on-die real estate (or an additional chip on the graphic card), having an implementation written in OpenCL could actually be the way to go.
There are already parallel implementations of DCT running on Cuda. And with the increased computing power of the GPU's shader units, this could be the way to go for the future (ATI seems indeed being interested in implementing video acceleration using GPGPU instead of dedicated silicon).

Theora/openCL could be a real killer application, specially since PowerVR people are also in the openCL design board (which probably means OpenCL capable embedable cores - makes a nice CortexA9 + PowerVR/openCL combination).

Actually, a considerable amount of commercial games use Vorbis as well: http://wiki.xiph.org/index.php/Games_that_use_Vorbis

This?

http://people.xiph.org/~greg/video/ytcompare/comparison.html

I read it, and have commented on it at length throughout this thread.

Basically the article briefly says that YouTube's H.264 is better than Theora, and then goes on at length showin how Theroa is better than H.263.

Xiph's own data shows H.264 has a big bitrate advantage at the same quality level even in a test that should favor Theora.

http://web.mit.edu/xiphmont/Public/theora/demo7.html

in the Rate-Distorion graph, note they start the plot at 50 Kbps, so look at the actual numbers.

For example, at 40 dB, x264 needs 70 Kbps and Theora needs 120 Kbps.

The gap would be bigger with higher motion, more detail, longer content, and particular when there are buffer constraints. Also, x264 is (properly) tuned for perceptual quality more than strict PSNR accuracy.

Theora suffers from not being a very mature implementation (which Xiph is making great progress on addressing) and being a 90's era codec design (about which Xiph can't do anything without breaking compatibility). And other codecs are getting better as well; if Theroa is refined enough to be reasonably optimal in a year, it'll be competing against the improved H.264 and VC-1 codecs of a year from now, and H.265 not that far away.

There are all kinds of interesting things about Theora, but competitive low bitrate compression efficiency isn't going to be one of them.

The point, if you'd bother reading THE FK!N ARTICLE is that there basically is no bandwidth cost increase for Theora vs what Youtube uses today.

Now, 5 years later I have a large collection of ogg files that are essentially useless. No one in the mainstream uses ogg, despite the superiority and price.

Weird. I started out the same, but I'm still ripping to Vorbis ogg. When I first started, I easily found the Cowon D2, which supported ogg. When I bought my Android G1, hey! Guess what? The native media player supported ogg, too. A quick Google search turns up this page, which lists no fewer than 59 flash based portable media players that will play oggs, and 38 hard-drive based portable media players that do, too. There are 5 smartphone platforms that support it (some of those through third-party apps for the phones). The last two DVD players I've bought have come with support for ... what? Playing oggs off data CDs.

There are many mainstream companies that support ogg. Some don't. "No one," however, is simply incorrect.

--- SER

What amuses me is the bias. The submitter wrote "hideously encumbered H.264 format." Hideously encumbered? Give me a break. It's as "encumbered" as MP3 is, and everybody uses MP3s.

Even Theora's developers say full H.264 edges out Theora. We're just supposed to adopt Theora simply because it's not "encumbered." Well, outside the echo chamber, not a lot of people care about that. Not to mention that H.264 has hardware acceleration support.

Hypothetically (I haven't confirmed this) you can also just install XiphQT and subsequently, Ogg Vorbis audio and Ogg Theora video should work in Safari just as it does in the Firefox 3.5 and Opera browsers.

Whether Apple® wants that or not...

To be perfectly clear, it's h264 that's a patent lawsuit magnet. Theora makes the patent lawsuit problem go away because there are no patents to worry about..

Really?

The Theora FAQ says that it is based on patented technology but that that a licensing deal permits free use. That's all well and good for the parts that are in that/those patent(s). What concerns me is the features section of the Theora wiki lists things that are also in other standards or are variants thereof.

A patent is usually meant to cover a single "invention" but I suspect that a video coding scheme would have scope for multiple different inventions. I'd be worried some of those may already be patented

I always supposed that Ogg came from Nanny Ogg.

Word of God* says no.

*I have lost hours of my time to that website.

FLAC fine. But have you anything to play Theora video that isn't a PC?

straight from the theora website

The COWON devices rock. I bought my SO a Cowon D2 for her birthday because it was tiny and had an SDHC slot, unlike any of its competitors that I can remember. She mostly uses it to watch dual-audio subtitled anime tracks in OGM format. Guess what, it also plays Theora. And FLAC. And Vorbis. Etc etc. Sweet little gadget, I need to get myself one now.

First of all, Matroska is an open spec, and most implementations (including the reference implementation, libmatroska) are Open Source (lgpl for libmatroska).

Mkv supports B-frames, Variable bit rate audio, Variable frame rate, Chapters, and Subtitles. Not all containers support all of these, and AVI only supports any of those with workarounds, modifications or just nasty hacks.

The mpeg container can't do chapters or subtitles, and obviously only holds media in the mpeg (1 or 2) format.

MP4 has limited chapter and subtitle support and only deals with mpeg media (basically 1, 2, and 4 ASP/AVC).

Ogg/ogm is designed for simplicity, streaming and specifically for Vorbis and Theora (although most/all other codecs can be used), while Mkv is meant as a completely general-purpose distribution container, and wants to replace avi, asf, mp4, mov, etc.

http://en.wikipedia.org/wiki/Matroska
http://en.wikipedia.org/wiki/Comparison_of_container_formats
http://www.matroska.org/technical/guides/faq/index.html
http://xiph.org/container/
http://www.hydrogenaudio.org/forums/lofiversion/index.php/t10426.html

Don't you think a reason for this is the lack of support for open formats in popular players?

Yes. I just identified one of the problems why it is not supported due to lack of popularity.

This is the typical chicken-and-egg problem. No one will support FLAC until there are devices that can play it. If there's no content in FLAC, player manufacturers won't see the need to include it.

I recall my Creative MP3 playing FLAC with ease. I know iPod can play FLAC after some effort yet the problem is the EFFORT part. FLAC isn't cool, FLAC isn't hip, it's not jazzy man and all the squares keep buying MP3 Players to play Mp3s.

That and I really dont want "normal people" trying to find codecs on google--most of the hits for "$AWESOME_CODEC" are usually just spyware installers.

Firefox 3.5 won't have support for other codecs than those that are built in (various Xiph codecs (Vorbis, Theora) and Wav). Since it won't be possible to install extra codecs for use in Firefox Firefox won't contribute to "normal people" installing random codecs from the net. If/when support for system codecs land (probably after 3.5) you may get the problem you describe.

I've been wondering - can anyone confirm whether or not Safari 4 will correctly handle the same Ogg Vorbis/Theora <audio>/<video> tags that Firefox 3.1 hypothetically will (assuming it is ever released[1]...) if one has installed XiphQT?

[1] Yes, yes, I know that eventually, someday, Mozilla Corporation will reach a release for Firefox 3.1. I think they just bit off way more than they could metaphorically chew and the unceasing stream of delays is making me all antsy here...

(Moral: Never agree to a feature plan bigger than your head?)

The best I can give you is hear-say -- I know someone who worked at our community radio station, and they did listening tests. MP3 is definitely the worst of current-gen codecs -- mostly because it's not a current-gen codec, it's a last-gen codec that people can't seem to let go.

So, Vorbis definitely sounds better, at the same bitrate.

I can't speak to CPU. At the time it was developed, it used a lot of CPU -- the name "ogg" came from the brute-force techniques that were used, which pushed that 486 hardware to the limit.

Did I mention, 486 hardware?

Oh, by the way: You're thinking Vorbis, not Ogg. Ogg is a container format. You could just as easily put an mp3 stream in there, or put a vorbis stream in an mkv.

And the donation was not just Ogg, but also Vorbis, Theora (video), and others.

I would say, if ogg became as successful as png, I'd be happy. Right now, png is successful enough that I can use it in a website, and every browser will support it. Even if plenty of people still use gif, or even jpeg inappropriately (for images that compress well). And then you have the brand-new applications -- HD-DVD seemed to prefer PNG images for the menus.

For something arcane like Vorbis (or the video codecs they'd like pursued) you can spend money on hundreds or thousands of programmer man-years and not get anything better.

Arcane? As in

arcane: known or knowable only to the initiate : secret <arcane rites> ; broadly : mysterious , obscure <arcane explanations>

Are you sure you are thinking of vorbis?

Uhm, its a direct show filter, which means it will work in any windows application which supports the installed windows codecs, so by going to http://www.xiph.org/downloads/ and clicking the Ogg Codec's for Windows link you will have exactly what you are talking about.

And finally, Comparing Silverlight to Ogg is like comparing Flash to MP3, its a stupid comparison as they do almost entirely different things. I stress almost.

So Ogg is free. Even if the manufacturers got $5 for each machine they shipped Ogg on, most would not do it because it would not increase sales by any measurable amount, and it would force them to pay more for hardware. MP3 decoders are mass produced and very very cheap.

Except a lot of manufacturers DO support Ogg/Vorbis (hereafter ogg because we are all sick of the container vs codec posts).
http://wiki.xiph.org/VorbisHardware

That list is in fact out of date already because the Archos 5 I have very definitely supports ogg.

The elephant in the corner that refuses to join the party is the ipod. It can already decode ogg with Rockbox and there isn't any terribly good reason they couldn't support ogg natively.

Right now I can rip music from CDs to ogg and not have any issues except if someone else wants that music (which is illegal anyway right...) and has an ipod. Then one of us has to transcode it to something they can use.

If Apple did add native ogg support I imagine the format's adoption would increase substantially. You'd just be able to find a lot more (illegal or otherwise) ogg files floating around that you could toss on your ipod and have it Just Work.

I really thought Ogg went the way of the dinosaur. Let's hope Mozilla can help it to succeed in the real world. It will be hard to beat mp3.

You thought wrong.

I-river was great in that it supported ogg/vorbis encoded music. Mine also works with AA batteries for 30 hours of non-stop music. A pity it's not called the iIriver, then it would have been more successful. Here's a list of ogg capable mp3 players. http://wiki.xiph.org/index.php/PortablePlayers

Are you sure that that OGG comparison is fair?

This http://www.xiph.org/dshow/
enables all directshow apps on a Windows machine to play all Xiph.org codecs.

This http://www.xiph.org/quicktime/
enables all quicktime apps to play all Xiph.org codecs.

Granted, this is probably just as much fucking around as asking the package manager to install Pidgin, but it's still possible!

Are you sure that that OGG comparison is fair?

This http://www.xiph.org/dshow/
enables all directshow apps on a Windows machine to play all Xiph.org codecs.

This http://www.xiph.org/quicktime/
enables all quicktime apps to play all Xiph.org codecs.

Granted, this is probably just as much fucking around as asking the package manager to install Pidgin, but it's still possible!

I bought mine at Canada Computers, a 15 store chain in southern Ontario. It wasn't hard at all. I knew what I wanted to buy. The Cowon site listed authorized retailers. http://flac.sourceforge.net/links.html#hardware lists a large number of devices that support FLAC. http://wiki.xiph.org/VorbisHardware lists devices that support Ogg Vorbis.