Xiph.org Releases Theora Alpha One

Pajama Crisis writes "Xiph.org, the crazy guys behind Ogg Vorbis, have released the first alpha version of Ogg Theora, an open video codec. Downloading, hacking and smashing into little pieces is cheerfully encouraged. Theora has been mentioned on Slashdot before. Also, Xiph has been working with a couple different companies to bring Vorbis to a portable near you; stay tuned."

Comments

Re:What I want...

[Bogatyr]

I'd love ogg vorbis as optional for use on my Apple iPod.

Killer App

[wls]

The killer application is to insert a DVD and have it simplistically rip and encode it for personal viewing. Make it simplistic enough for the masses to use, and let the codec take off as a standard induced by practice not dictates or technology.

Re:Killer App

[ryanvm]

The killer application is to insert a DVD and have it simplistically rip and encode it for personal viewing.

Yeah - I'm sure the MPAA is going to let that happen.

They sue you into oblivion just for making an unlicensed DVD player. Just imagine what they'd do to a company trying to distribute a packaged DVD ripping solution. If there were a "scrotal electrocution" clause in the DMCA, it'd be reserved for this.

Re:Killer App

[utexaspunk]

likewise, OGG could stand a better chance as an audio codec if an insert-and-be-done program were made. Windows Media Player (and iTunes) has this feature for CD's. You stick it in, it gets the track names and rips it. Match that with a P2P client, and it could get adopted just for downloading ease. If everyone were using the same program, files that were the same would have the same names, which would make it much easier to download a whole album, or switch from one peer to another mid-download, thus reducing download time and bandwidth usage.

Re:Killer App

[rseuhs]

likewise, OGG could stand a better chance as an audio codec if an insert-and-be-done program were made. Windows Media Player (and iTunes) has this feature for CD's. You stick it in, it gets the track names and rips it.

KDE's Konqueror does exactly that. Just stick an audio cd in and type audiocd:/ in Konqu's location and see what happens...

Re:Killer App

[cryptochrome]

No, what would be ideal is for it to duplicate not only the video but the entire DVD structure, menus, subtitles, extras and all. A single, fully functional video file, except of course using more modern codecs so it's smaller. (And since this is a still a very big file, it should also contain data that allows the software to detect and repair the inevitable errors.)

Of course the only file format/architecture up to the challenge for that is Quicktime and maybe MPEG4.

Re:Killer App

[dasunt]

GordianKnot (available here) helps to automate the button. Someone should ask the author for an 'idiot' button. Select the soundtrack you want, resolution, size of the finished product, and click 'rip'.

Of course, there are enough open source and free apps that could be automated, its not stopping anyone from making their own home-brewed gordianknot. To piss off the big guys, remember, make it a win32 binary with very easy setup. :)

Re:Killer App

[defile]

I wrote a script to automate this process. You need cddb.pl (comes with perl CDDB module), oggenc, and cdparanoia (last two by the wonderful Xiphophorus people).

The script is called cddbrename. It automatically looks up the CD against CDDB and makes the appropriate artist/album subdirectories, and copies the tracks there with appropriate names.

There are a couple of steps:
cd ~/ogg
(insert audio CD)
cdparanoia -B
ls *.wav | cddbrename
cd (**artist name/album name**)
oggenc *.wav
rm *.wav

Surely this can be automated. But it's not hard to do. Oh, sometimes a pregap will be saved as a zero track. Just delete it.

---
#!/usr/bin/perl

# Er, uh, GPL distributions terms.

open(D,"cddb.pl|");
my @cddb = ;
close(D);

my $album = '';
my $artist = '';
my @tracks;

foreach (@cddb) {
chomp($_);
$_ =~ s/\r//g;
$_ =~ s/ $//g;
$_ =~ s/\//-/g;
last if ($_ eq '');
if ($_ =~ /^title: (.*)$/) {
$album = $1;
} elsif ($_ =~ /^artist: (.*)$/) {
$artist = $1;
} elsif ($_ =~ /^track [0-9]*: (.*)$/) {
push @tracks, $1;
}
}

open(D,"ls *.wav|");
my @files = ;
close(D);

foreach (@files) {
chomp;
$_ =~ s/\r//g;
last if ($_ eq '');
if ($_ =~ /^track([0-9]*)\.cdda\.wav$/) {
my $srcfile = $_;
mkdir("$artist");
mkdir("$artist/$album");
#my $dstfile = "$artist/$album/$1 $tracks[$1].wav";
my $dstfile = "$artist/$album/$1 $tracks[$1-1].wav";
print "$srcfile -> $dstfile\n";
rename($srcfile,$dstfile) or die("Cannot rename(): $!\n");
}
}
---

Re:Killer App

[defile]

Oh, the ls *.wav | cddbrename is no longer necessary. Who knew? Just cddbrename.

And that weirdo looking line should be:

my @files = <D>;

Slashdot interpreted it as:

my @files = ;

Re:Killer App

[Alan]

The only killer app I want is an affordable handheld mp3 player. I'd love to get an iPod, but I'm not about to shell out $600CND for one, or shell out $300CND to get a knock off with 128mb ram. I want something I can go to the gym with, or take on a bus ride that's going to give me more than 10 or twenty songs, and that's small enough that it can clip to my waistband or arm. Currently I'm using a cheap ass $100 discman that can play mp3s for the gym. It works, but I can't edit what is on there without burning a new disk, and it's still prone to skipping while on the treadmill.

Please please please give us something small, cheap, and with lots of hard drive space! If you make it play .ogg as well then even better, I'd love to re-rip all my cds to a "better" format than mp3!

Re:Killer App

[EllisDees]

You mean like Grip? Well, at least the ripping part.

Re:Killer App

[wls]

Actually, the proposed idea was not to break copyright nor distribute movies. I've done neither, and the size of my personal commercial movie collection assists in demonstrating that.

Where I was going was that adoption of standards don't really happen by dictate of use or by people going "oh that's a really cool technology." Nor, does it even reside with quality. It resides with ease of use.

No matter how good or how bad this codec is, if someone makes a utility that is just so pathetically easy to install and use, then it will gain adoption because people won't abandon it for the wrong reason.

Now, yes, I'd have to concede that it's likely that people will engage in illegal activities. However, let's face it, they'll do that with whatever is at hand -- AVI and MPG being a good example, and those have dozens of codec choices. Ironically, if the "copyright breaking bad guys" flood the net with stuff in this codec, the technology and applications will mature faster, much the same way you're benefiting from the VCRs, recorable CD-ROMs, dual layer DVDs, high bandwidth internet, and streaming video technologies the porn industry funneled lots of money into.

I say a technology is neutral, and it's user is where the ethical standpoint resides. But it's demonstrated through history "great technology" is not the factor that leads to a win, it's adoption by the masses. And that means making something so simple to use that the masses can use it -- the good part, however, is that when the open source community does that, they don't skimp on quality to make a buck.

Ogg Theora?

[spectecjr]

And I expect that this has nothing to do with Max Headroom whatsoever, right?

Re:Ogg Theora?

[Spazholio]

Actually, if you check the FAQ, that's exactly where it came from.

Re:Ogg Theora?

[Ziviyr]

I'm waiting for the Ogg Rebus video codec. :-)

Re:Ogg Theora?

[Megane]

For maximum compression, the next generation will be Ogg Blipvert. Once they get the bugs out -- one in every hundred thousand people or so explode when watching Ogg Blipvert videos.

So, what's so special about Theora?

[Crapflooder Supreme]

Its benefits sound, well, theoretical to me.

vp3

[sootman]

...has been rolled into Theora, as is said at both sites.

from vp3.com:
NOTE TO ALL VP3 DEVELOPERS:
Monday, September 9, 2002 -- Starting today, all source code development and maintenance for the VP3 open source codec has moved to a new home: www.theora.org. Piloted by the open-source wunderkids at xiph.org who brought you Vorbis audio, Theora heralds a new era of open and license-free multimedia.

from theora.org:
What is Theora? Theora will be a video codec that builds upon On2's VP3 codec.

So, in case anyone was wondering (like I just was), there you go.

Re:vp3

[mcelrath]

So it's great that they released this codec into the wild, but On2 Technologies is now up to a VP5 codec.

Will the features in VP4/VP5 ever be folded in to the open source version? (a la Aladdin Ghostscript) Are the differences between VP5 and VP3 enough for us to care? Are any On2 developers actively working on Ogg Theora?

Basically, is On2 playing the version game to force customers into upgrading (and getting more money), or are there real improvements in VP4/5 that Ogg users will want?

Anyway, kudos to On2 in any case. I will definitely be using it.

-- Bob

Re:vp3

[sootman]

1^2=1; (-1)^2=1; 1^2=(-1)^2; 1=-1; 1=0

Shouldn't the last bit be '2=0'? :-)

Re:so?

[josh crawley]

Too true. OSS ideas are to re-invent the wheel, even though many mathematical algorythims are patented (like Ogg - psychoacoutic model of elimination). Still, it may be needed, as Sigma pissed off the guys who were creating Xvid. Cause a company was selling open source software without providing source, the Xvid team is now rioting and is quitting making xvid.

And a second point: what would be more "together", 2 totally diffrent codeds slapped together, or a dual thought out codec. I figure they use similar measuers to make both. If the do, you might only need 2 chips on a Ogg: AV displayer (handheld device). A Ogg decoder, and a video displayer. Just an idea.

Re:What I want...

[silvaran]

You may get it. I know my SonicBlue Riovolt has had numerous updates. I'm just hoping the amount of RAM on the device is enough to accommodate upgrades. That would absolutely rock (pun intended).

Where's the link?!?

[ryanvm]

Also, Xiph has been working with a couple different companies (iRiver, Frontier Labs) to bring Vorbis to a portable near you; stay tuned.

Oh come on! News like that and no link or credible source?!?

For those who don't know - iRiver is the company that makes Rio's Volt lineup of portable CD MP3 players. Top notch players, but up until now they only play MP3 and WMA.

The Rio Volt's lack of Vorbis support is the only reason I haven't switched to encoding all my music in Vorbis. I've got to believe that the lack of hardware supporting Vorbis is the number one reason for its modest adoption rate.

Re:Where's the link?!?

[jonathan_ingram]

Oh come on! News like that and no link or credible source?!?

"Pajama Crisis" is Emmett Plant, Xiph's CEO.

Re:Where's the link?!?

[BrookHarty]

It takes time to design hardware, OGG Vorbis 1.0 hasnt even been out for full year. I was surprised that video games like Serrious Sam 2 used it. Even game companies are starting to use DIVX for video files for games(WarCraft3). The word is out, people are looking at it, just give it some time.

Soon as the video codec is released, and then production tools, we should see video hardware follow. I hope they dont make the same mistake as DIVX and have 20 different versions.

Re:Where's the link?!?

[big.ears]

The Rio Volt's lack of Vorbis support is the only reason I haven't switched to encoding all my music in Vorbis.

Coincidentally, it is also the only reason why I haven't bought a Rio Volt.

Re:Where's the link?!?

[m_chan]

I saw this post at the iRiver user forum:

Emmett
2002-09-24 15:51:43 (Xiph.org - Ogg Vorbis) Hey, folks, please read. Ladies and Gentlemen and other assorted interested parties:

This morning (2:28am, EDT), I received an E-mail from Y.H. Lee, the Chief Engineer of iRiver's products. He has informed me that they are currently porting the fixed-point Ogg Vorbis decoder (Tremor) to their product. We will endeavour to assist them in any way we can.

In addition to letting iRiver know that we stand behind them ready to assist with technical issues, I believe a huge thanks is in order to the people who have contacted iRiver asking them to support our format. We're indebted to all of you who have posted looking for Vorbis support on this thread, as well as all of you who have sent iRiver E-mails and called them on the telephone.

With any luck, it won't be long before iRiver starts releasing products with Ogg Vorbis support. When the day comes, I hope you'll all take a moment to congratulate yourselves, and celebrate with a portable full of music encoded in an open and free format. :)

If anyone has any questions, comments, or anything they want to talk about, please feel free to drop me an E-mail to emmett@xiph.org, or give me a call at 215-820-9583. We're an open source project, and we love hearing from people, talking about the codec, and discussing cool stuff. Please don't hesitate to contact us if you want to gripe, complain or compliment.

Thanks again for all your help, folks. We couldn't do it without you.

Emmett Plant
CEO, Xiph.org Foundation

We demand benchmarks!

[YellowSubRoutine]

How well does this thing compare to stuff like mpeg4? How well does it stream?
If we need a killer on video formats, it better be a damn high compression ratio...

Re:Theora...

[Spazholio]

If you check the FAQ, they do say that this name came from Max Headroom. The others, I'm not sure about, but I think so.

Which?

[Grendel Drago]

Really? Do you mean DivX 3.11a with SBC? Or DivX 4 or 5? Maybe the new XviD codec, which is replacing (various versions of) DivX in the pirate scene.

Seriously. "DivX 0wnz". Put a little thought into what you write---unlike MP3, "DivX" comprises a wide variety of codecs and licensing schemes.

--grendel drago

Top 10 things still missing from all Ogg software:

[Eric_Cartman_South_P]

1) Add DRM so my privacy and security will be protected.

2) Make it impossible to uninstall.

3) Necesitate a live web connection to play files.

4) Upload information to a "security" server every time I play a music file or insert a CD.

5) Utilize DirectSound.

6) Get rid of *nix versions.

6) Release only binaries so terrorists don't get the source code.

7) Use attorneys to bring down all mirror site distributions.

8) Pick sellout a$$hole to pump out CD that installs trojan uninstallabe player onto people's system. Is Peter Gabriel available? http://www.theregister.co.uk/content/4/27272.html

9) Make sure user interface is horrbile, avoid simplicity and good design (so see QuickTime and WinAmp for what NOT to do).

10) Add interoperation with future .NET (Passport) purchasing system.

Re:Theora...

[Jugalator]

Being a stupid Star Wars fan, I always associate Tarkin with Grand Moff Tarkin. :-) oh well...

About the name

[shut_up_man]

Q: Why the name 'Theora?'

A: Like other Xiph.org Foundation codec projects such as Vorbis or Tarkin, Theora is named after a fictional character. Theora Jones was the name of Edison Carter's 'controller' on the television series Max Headroom. She was played by Amanda Pays.

DAMN they have cool names for their stuff. I wish other people would follow their lead (bloody "Opteron").

Re:About the name

[foobar104]

Okay, I'm a geek. I didn't realize this would be a FAQ, because it never occurred to me that this would be a question at all, asked frequently or infrequently. I heard the name and thought, "Oh, yeah, Theora makes perfect since for a video codec since she was a television producer. Neat." And that was the end of it.

I'm such a nerd.

Compression.

[Grendel Drago]

Not exactly. It had better have a configurable compression ratio. Ideally, you'd be able to take a clip, tell it "Make it 200MB" and come back in an hour (or three) to find a high-quality encode waiting for you.

The problem is really the quality that comes out of the encoding process. The real comparison done nowadays is to encode the same movie at the same bitrate with two different codecs and compare the quality. Of course, this is a subjective and nontrivial process...

--grendel drago

They never actually say...

[Spazholio]

...whether or not it's better than DivX/MPEG4, only that that would be its main competition. Does that mean that they're banking on the fact that they'll never start charging licensing fees, but they're not *quite* as good as DivX? Or that they're just as good, if not better, but are going to let the people make up their own minds?

Either way, I'm stoked. Ever since I discovered Ogg Vorbis, I haven't regretted it and haven't looked back since. Hopefully I'll be able to say the same about Ogg Theora.

Re:They never actually say...

[jonathan_ingram]

Theora is basically a cleaned up VP3, and is likely to remain so. Although some visual improvements may sneak in along the way, this is not the main focus of the project, which is to convert VP3 from being a Windows only mess into a portable cross-platform codec which plays well with the Ogg container format.

Previous tests by places such as Doom 9 have shown that VP3 is beaten by the various pseudo-MPEG4 encoders, although not embarassingly. I imagine that this situation will continue: DivX/XviD/etc. will continue to have marginally better quality.

Re:They never actually say...

[Dr.Dubious DDQ]

It's come up in one or two places that I've lurked that a lot of the niceties of the video codec quality actually come from pre- and post-processing, so improvements may very well come for it. As you say, though, they may not come from the Ogg Theora project directly.

I WILL, quite frankly, be happy to see an encoder and decoder that I can use for vp3 on my Linux box, though. The CVS repository seems to have been dead for months...

I get the IMPRESSION (purely from various readings, not from experience) that VP3 could, in the end, be as good or marginally better for a lot of 'live' footage than mpeg4, while mpeg 4, tending to be 'sharper', may give better quality for 'sharp' images like digital "movies" (i.e. 'cutscenes' from games) and some types of animation.

Just a purely subjective, speculative observation though. I'm most interested in the 'free as in free-from-patent-lawyers-suing-you-for-distributin g-stuff-in-this-format' aspect of it :-)

Re:Meritocracy OpenSource Progress

[NineNine]

Doesn't commerical software work even better as a meritocracy? You make something only if you think it'll be useful, and people vote with their wallets. To me, people willing to pay for a product says *much* more about the quality of a product than if it's given away for free. It's a difference between "Yeah, I'll pay money for that" versus "Eh. Well, it's free."

Re:They don't have the best track record.

[jonathan_ingram]

What the fuck are you on?

Seriously. My job's getting me down a bit and I could do with some cheering up.

Re:Meritocracy OpenSource Progress

[rseuhs]

Not really. For example the CSS model completely eliminates small packages. - Nobody would pay for a command line utility like "grep" or "wc".

This is why every commercial package gets bloated and reinvents the wheel numerous times.

How does he know?

[Frag-A-Muffin]

From the post:

"Also, Xiph has been working with a couple different companies to bring Vorbis to a portable near you; stay tuned."

How does he know this? Can anyone confirm this? I'm a NEXII owner and have been emailing Frontier Labs for a bit now. (Every friend that I recommend this to, that has subsequently purchased a NEXII(e) have been harassed by me to email them as well). But I've never heard any official word from them that they're even working on it. Is it true!? If it is! Woohoo! Anyone have links to at least one other site that has these rumours?

Re:How does he know?

[Frag-A-Muffin]

I've been talking to the Xiph guys, it's true, I believe Monty himself will be starting work on a flash for the NEXII/IIe within a couple weeks, if it turns out to be technically possible.


Sweet! I don't see why it wouldn't be technically possible though? What's in one of those NEXII(e) anyways? Or is it possibly the firmware size won't fit? Anyone know how big the flash ROM for the firmware is?

Re:How does he know?

[Sesse]

"Pajama Crisis" is an alias for Emmett Plant, CEO of Xiph.org. He would know. :-) (It isn't all new either -- browse through some of the previous Vorbis stories and I'm sure you can find a few comments regarding it.)

/* Steinar */

Re:How does he know?

[brain159]

Any chance he could fix up some of the other creeping horrors in the NexII's firmware (read over the nexiimp3 yahoogroup to get a feel for the outstanding issues) while he's at it? :-)

BSPlayer

[Grendel Drago]

Yeah, it'd be nice if BSPlayer were the only media-playing app I needed on the Windows box. It's lightweight, fast, one-key shortcuts for the important stuff...

And most importantly, it played DivX just fine on my friend's 450MHz K6-2. With subtitles. WMP choked and died. I think that's reason enough not to use their bloatware...

But really, there's no reason for the format to be tied to the player. BSPlayer is enough for me...

--grendel drago

IRivers Response

[emolitor]

"Thanks for your suggestion. Actually iRiver have listened to our users' need and start working with Ogg Vorbis format, ManPower have been allocated to develop the Ogg Vorbis, let's give our engineers more time to develop this format. Some users suggested to give a schedule on this issue. However, it is really difficult to tell at this moment, let's just hope it to be released as soon as possible."

http://www.iriver.com/user/user_view.asp?page=1&id x=1268&mode=Àüü&strque=&field=1

Thats clipped from their forums.

Here's the link!

[XJoshX]

Here's the link!

We'll see if they're able to get it into an upgrade of their current firmware.. Even if they don't its still quite exciting news for the future.

Here it is

[Malc]

It's funny, I was just reading this two hours ago: http://www.xiph.org/ogg/vorbis/hardware.html

Linux?

[The Pi-Guy]

I can't seem to find how to compile it for Linux.

Will there be support, I wonder....

Re:Linux?

[Dr.Dubious DDQ]

There is supposed to be support for it - you are supposed to be able to run the ./autogen.sh script that comes with it to generate the necessary makefiles.

Unfortunately, I'm getting a bunch of errors from automake when I do this, so I can't test it...

Re:Linux?

[xiphmont]

Read the README file... then if you still have trouble, let us know.

(This Alpha release was developed and tested on Linux)

Monty

obscure

[Lovejoy]

Yet another name to insure that Ogg is completely unmarketable!

Re:obscure

[kraf]

The solution:
OGGX THEORAX

Anything can be turned into a trendy brand name by adding an X.

Re:They don't have the best track record.

[Microlith]

I don't know about anyone else, but something tells me you're full of shit...

I'd like to see more evidence of this "increased rate of speaker decay" you claim that Vorbis causes. I honestly don't see that happening, unless the amplifier supplying the signal is fucked, or your DAC is shot and is feeding bad signal (but you'd hear that). Once audio is decompressed, it's just PCM.

IIRC, wasn't Vorbis's big feature that it supported up to 255 independent audio tracks? No joint-stereo crap at all?

And I've encoded karaoke tracks into Vorbis. I imagine the only thing keeping me from singing along is that my Japanese skills aren't that high. Interestingly enough, the tracks I have encoded (so far: Jin-Roh and Escaflowne: The Movie OSTs) are some of the hardests tracks I've thrown at any encoder, and Vorbis pulls off a damn fine job.

Sorry prof., but I'd like more info to back up your claims before my skepticism is anything but high.

Wow, this is like watching mainland Chinese news.

[xiphmont]

Forgive the tone of this reply, but this guy is smoking so much crack that it's simply impossible to take seriously. There's nothing I hate more than self styled 'experts' who just make it all up.

Sir (and I use the word loosely), if you're a plant for the competition, try a bit more subtlety and you'll blend in better. I've also found that borrowing buzzwords liberally from 'Star Trek' will help your gobbledygook sound more convincing to both low level nerds and the common man.

Off we go...

I've been studying psychoacoustics in my spare time,

Oh boy, a *real* expert... Give me a second to contain my excitement.

It sacrifices a lot to "sound better" than MP3, and while some of their tradeoffs do manage to improve sound quality

A terrible, terrible thing it is to sound better... there must be something wrong

First off, Vorbis concentrates its encoding in the more audible midrange

...as does every psychoacoustic compression, because your ears have the greatest perceivable signal depth and resolution in the midrange. This is page one of most psychoacoustics textbooks (and it also happens to be true). Go study some more.

completely cutting out higher overtones. While MP3 works similarly, it manages to keep enough of the high range to maintain the "feel" of the original music.

Bzzt. False statement number one. Go study in your spare time a little harder, do some ABX testing then come back and tell us what you learned. However this one is almost forgivable compared to the nonsense below.

Vorbis claims to support more than two channel audio, but this is misleading.

Bzzt, no it's not. 255 channel support, all of which may be totally independent and un-coupled. You need not use 'joint stereo' (our method is more general and we call it 'channel coupling') at all.

MP3 encodes stereo using a "joint-stereo" method, which couples both tracks together into a mono track, giving each frame a different balance to simulate stereo on a mono track. This is equivalent to playing a mono tape and turning the balance knob!

No, idiot, that's 'intensity stereo', not 'joint stereo'. Vorbis does not use intensity stereo.

Obviously, this is less than optimal

It certainly would be, unfortunately--- *GASP* ---it's not true!

While Vorbis supports true stereo encoding, it fakes 5.1-channel audio using a "joint-joint-stereo" method, where the left-back/left-front and right-back/right-front channels and joined together into the two stereo tracks in a similar fashion. Not very good at all.

Bzzzt. Go read the spec again Bucky. You could do what's described above, yes, but that's not 'the way Vorbis does it'.

The way that Vorbis compresses its audio accelerates speaker degradation

Actually using the speakers accellerates degredation too. They last alot longer when you leave them in the box they came in and don't plug them in.

It breaks sound up into an evenly-spaced array of harmonics which approximate the original waveform

Those are not 'harmonics', and Vorbis's compression pays no particualr attention to sinusoidal harmonics. Perhaps you'd like to wait until college and get some signal processing lectures under your belt before coming back.

"Big deal", you say, "that's how all lossy encoding schemes work!"

[sigh] No, no it's not.

But if we assume, for a second, that you said, 'Vorbis is a transform-domain codec', which is what you meant, no, not all lossy audio compression formats are transform based.

But the way that Vorbis does it causes a noticeable amount of harmonic resonance in speaker systems, stressing their driver system and accelerating the rate at which they decay.

The problem Sir is that you have a surplus of Zackthorp particles coming from your warp core, a well known source of wear and tear on cheap speakers. Make sure your speakers are rated for greater than warp 4 before trying to use them so close to a Gammagorp Modulator and your worries are over!

I listened to the result, and believe me, it's true! Because I said so.

If you know the story of the first Tacoma Narrows bridge [carleton.ca], this is the same principle, working at a smaller and more gradual pace.

...which has *nothing whatsoever* to do with what we're talking about. But hey, I loved that TNG episode where Data gets his dick caught in the food synthesizer, so that goes to show Vorbis is Bad. Vorbis made him do it. Really. I heard the voices whispering all through that episode.

Given Xiph's poor track record with Vorbis

OK, let's stop here. Everyone gather around, point and laugh!

Monty
[sheesh]

Re:They don't have the best track record.

[JabberWokky]

I cannot say anything about the your first point, but as your latter two points are blatantly false, I'd imagine that the first is BS as well.

While Vorbis supports true stereo encoding, it fakes 5.1-channel audio using a "joint-joint-stereo" method, where the left-back/left-front and right-back/right-front channels and joined together into the two stereo tracks in a similar fashion.

While it might be capable of 'faking 5.1' in that way (and almost all physical surround sound systems do the same when plugged into a sterio source, like the common surround sound on a sterio cable or VCR system), Vorbis *does* support ~255 synced and totally seperate audio channels. This makes it ideal for interactive sound, say kareoke, game soundtracks that can turn up the anxiety track, the awe track, etc, as you encounter certain things. It also makes it great for eventually supporting all the alternate audio tracks (languages, commentary) of DVD style video when Theora is finished.

--
Evan (no references)

Re:Top 10 things still missing from all Ogg softwa

[Dirtside]

5) Utilize DirectSound.

6) Get rid of *nix versions.

6) Release only binaries so terrorists don't get the source code.

7) Use attorneys to bring down all mirror site distributions.

No! You fool! There is no rule six, not two rule sixes! You've reversed the polarity on the reactor and doomed us all!

Re:so?

[nanojath]

I think you're wrong, I think you're looking at the process from too short-term and narrow of a point of view. The beauty of open source is that it creates a basis for the way certain things should be: to wit, that the fundamental architectures of how we digitize media should be a free, open standard so that we can concentrate on what's worth selling (services that organize and deliver content and devices that play it). Having any degree of proprietary encumbrance on a media codec is the equivalent of patenting the alphabet or the left-right up-down orientation of conventional english text. It's just stupid, it serves no purpose.

At worste, it doesn't do any harm. At best, it may serve as the basis for a resonable, sensible multi-media platform at last. So what's your problem? It's not like anyone is making you use it.

Re:They don't have the best track record.

[Yarn]

When something looks like bullshit, but you're not sure, check the poster's record.

This guy could troll for his country.

Re:File Extensions

[Hitokage_Nishino]

Xiph intends to continue using .ogg

Other people already using OGG as a movie format use .ogm

Re:Top 10 things still missing from all Ogg softwa

[ellem]

6) Get rid of *nix versions.

6) Release only binaries so terrorists don't get the source code.

6) Get zelots to start a religious war over formats

mplayer/mencoder

[peter_gzowski]

Evidently oblivion exists at www.mplayerhq.hu. They have an encoder that lets you rip your dvd to DivX4 using 1, 2 or 3 pass encoding. Instuctions are here. Is three lines at a command prompt simple enough?

Re:mplayer/mencoder

[Alan]

root@gentoobox # USE="+dvd" emerge mplayer :)

Monty's putting the smackdown on this guy!

[Jason Earl]

Ouch, that was hilarious. Not only was it entertaining, but it was educational as well. Five stars!

Oh, and thanks for Vorbis. Ogg makes me happy.

Re:File Extensions

[ncc74656]

If Ogg Vorbis are .ogg files, what are they going to call Ogg Theora files?

If they're smart, .avi. Make it a codec that plugs in alongside XviD, Huffyuv, etc., and you'll have a sizable amount of capture/editing/playback software that'll be able to use it right off the bat.

Re:Wow, this is like watching mainland Chinese new

[tempest303]

Wow... leave it to Monty to drop the hardcore De-FUD bomb and donkey punch a troll into oblivion. :-)

well done, man... I'm still laughing over this one...

Ogg petition for Ipod

[huper]

Attn: All ./'ers

There is a petition in the applecare forums to bring ogg support to the Ipod. It would make a good product even better.
Take a look and sign it if you can.

VP3 background

[imsabbel]

The vp3 video codec was released to open source by on2 technology mainly because its image quality was noticable worse then (divx 5.0, Xvid ,wmv8).
They tried to get some publicity while selling the succesor.,vp4 (or 5), which isnt open sourced.

Any improvement to bring it on the same level as xvid or divx would just retrace on2's steps of progress.
How can OGG be sure that they wont violate patents in this process?

Re:They don't have the best track record.

[0x0d0a]

by Professor Collins

Yeah. People tend to give people with a "Professor" in front of their name more credence. Good nick choice.

I've been studying psychoacoustics in my spare time

Heh ...in my experimentations.

Heh. ...concentrates its encoding in the more audible midrange...

Yup. Be a pretty pathetic audio codec if it didn't. ...because the brain unconsciously uses inaudible overtones as a guide to determinine the tone of the music...

Given 128kbps, I should expend more data on the inaudible portion of the spectrum to hint people as to what the audible tone is that I'm already emphasizing in my encoding? Uh, huh.

The way that Vorbis compresses its audio accelerates speaker degredation

Tech can now destroy speakers! Muahahaha!

Re:Wow, this is like watching mainland Chinese new

[0x0d0a]

He didn't have any *true* points, much less salient ones.

Re:Wow, this is like watching mainland Chinese new

[0x0d0a]

Monty, love your work. Thanks, and congrats on the 1.0 release of Vorbis!

But hey, I loved that TNG episode where Data gets his dick caught in the food synthesizer, so that goes to show Vorbis is Bad.

This may be making a serious run for being my new signature.

Re:File Extensions

[kevin lyda]

they should use .the. that way people talk about ogg theora files will say things like, "did you get the dot the file?"

Real Killer App

[SuperKendall]

I think what you propose would be nice, but not killer app...

My idea of a killer app is close, one that would let you take any music you like and grab video snippets from multiple DVD's or other video sources to create your own amazing music video, then spit out the result on your DVD writer or email it off to some friends. I think people would be drawn in and it would absorb teenagers to no end.

Allowing consumers free and easy mixing of various copyrighted media is the ultimate nightmare of RIAA and MPAA though!

Re:File Extensions

[lunatik17]

Ogg is merely the container format that both Vorbis and Theora use. Just like Mov is a quicktime container that can be any number of different codecs, as well as avi, asf and wmv. Calling a Theora file .ogg is perfectly valid and calling a vorbis file .ogg is just as valid.

Not worldwide, and it was 15 years ago

[Goonie]

Max Headroom wasn't such a big deal here in Australia (we saw the Coke ads, but I'm not sure that we ever saw the series), and it was 15 years ago (and so people under about 22 won't remember). Hence the need to explain it to some people.

Re:Wow, this is like watching mainland Chinese new

[Compuser]

I think the parent is a troll, but in case he is
serious:
What does a square wave do to your speakers?
We have some decent peakers down in our lab,
because we do vibration testing before real
experiments run and we run all kinds of sharp
looking waveforms thru them and we don't see
any problems. This is in fixed setup, inside an
acoustic room with acoustic level meters and
accelerometers. Doesn't get any more precise than
that. Our speakers are fine after 5 years.
What's supposed to be the problem?

Re:File Extensions

[ncc74656]

AVI can only multiplex one video and one audio stream

It's usually not supported by most programs (maybe not even by the Video for Windows and DirectShow APIs), but you can have multiple audio and video streams in an AVI. At work, we have some lossless video-compression code that deals with multiple video streams in an AVI. We had to write our own AVI-handling code to create some of the test files used since we didn't have anything on hand that would do that for us, but one of the test files used has three or four video streams in it. Play it with Windows Media Player and you only get the first video stream, but that's a DirectShow limitation. I'd post a link to the relevant specs, but they didn't turn up after a few minutes of googling (it's not my project, or I'd more than likely have the link in Mozilla).

Then what's a "media type"?

[yerricde]

Just like Mov is a quicktime container that can be any number of different codecs, as well as avi, asf and wmv.

True. However, containers typically have their own associated most common codecs. For instance, a .avi file is generally either some sort of uncompressed video or MPEG-4 video, along with either PCM audio, MP3 audio, WMA audio, or Vorbis audio. Likewise, .mov typically implies Sorenson video with QDesign audio.

Calling a Theora file .ogg is perfectly valid

Not for file systems that determine a file's media type (e.g. audio/ogg vs. video/ogg) and disposition (e.g. open in background in an audio-player style interface vs. open in foreground in a video-player style interface) from its extension. The common uses of the common Windows and UNIX file systems have this limitation. In this case, a fellow could use .ogg for audio but .ogm for a multiplexed audio/video/caption stream; this appears to be the convention that has arisen among users.

Re:Then what's a "media type"?

[Junta]

Mostly a me too post, but I think there is some more meat here..

I'll leave the avi/mov descriptions alone, they are a bit incorrect, but your point that remains true is that the extensions of .mov, .mpg, and .avi all imply video streams by convention, even though not technically required by the applicable standards in any one of them. Thanks to this convention, file managers hold the reasonable expectation that a program like xmms would likely not be able to handle those formats.

But with regard to mime-typing, the audio/ogg vs. video/ogg is handled by some systems as just saying application/ogg, which is supposed to be neutral but in actuality seems more wrong than either one. A media/ogg or something might be more fitting. The intent from the beginning was to have a truely flexible media format and some programs have recognized this. However, the extreme delay in production of 'endorsed' codecs (Theora and Tarkin) has created the convention that .ogg=audio only. And this has created a problem. Even under OSX and *nix in addition to windows now, Mime-typing based on extensions governs file manager behavior. Now with people who have winamp or xmms set up to open up extensions they have come to expect to be audio (.mp3, .ogg, etc...), Theora presents a problem. Even if both xmms and winamp are capable of playing back video, the interface is clearly not designed for video playback. Additionally, xmms video plugins have to cheat and manage their own video output, so video in xmms is even more of a hack, saving the developer only the trouble of UI design and audio output. Because of this, I suspect that the encroaching .ogm will indeed become the video extension by convention. It is sad because at least to me, .ogg as a unified media format sounds nice, and even though entirely superficial, somehow different naming seems to hurt that image.

But this just goes to show that sometimes a beautiful vision just can't work with practical usage. It would take one hell of a UI for a player to truly deal appropriately with a format that could be a very simple audio file to something with DVD-level complexity. No single UI can be fit to all of them. WMP tries and fails miserably.... I guess the answer is a single app that radically modifies its UI based on content.... Until that is here, you are right, .ogm is a necessary divergence from the vision..

Vorbis

[moogla]

The audio codec is called vorbis.

VORBIS

That's all you need to say. OGG is a format container for audio, among other things. No one goes around talking about their RIFFs or ASFs do they? You don't call a DivX movie an AVI, do you?

FUCKING VORBIS!

Re:Vorbis

[Dahan]

No one goes around talking about their RIFFs or ASFs do they? You don't call a DivX movie an AVI, do you?

No one talks about RIFFs because nobody sees "RIFF." However, people certainly do talk about ASFs and AVIs. And likewise, since the popular file extension is .ogg, people are going to call the files OGGs.

Quality of Theora can be improved

[tangent3]

Although VP3 didn't quite perform that well in Doom9's codec comparison test, I think there is still a lot of room for improvement. What's so special about the codecs which perform well? Why, 2-pass encoding of course, something which VP3 didn't have. Now that VP3 is opened, I'm sure we'll see 2-pass encoding being hacked in soon and odds are the quality of Theora encodes will be able to match those of DivX3-SBC, DivX5 and XviD.

Re:so?

[Salsaman]

many mathematical algorythims are patented (like Ogg - psychoacoutic model of elimination)

Can you provide evidence of this ? I thought the ogg team were very careful not to infringe any patents.

Now I'm confused

[Salsaman]

I thought ogg's video format was going to be called Tarkin ?

HLS

[Dahan]

Fish on!

Tarkin

[dmaxwell]

That is still planned but the release of the vp3 codec and the existance of vorbis means they can get something out NOW. I understand they still intend to develop their own video codec. THAT is what will be Tarkin and I wouldn't hold my breath waiting for it. Making a video codec from scratch with out patent entanglement is going to be super hard.

Opening every file in a folder is slow.

[yerricde]

Wouldn't it be better to associate *.ogg with a configurable wrapper which then spawns the appropriate media player?

And then have the operating system have to search inside the file when you Start > Search > For Audio... The shell wants to know very quickly (i.e. from the directory, not from opening the file, which incurs an additional rotating-media seek per file) whether to display an "eighth note" icon or a "filmstrip" icon.

Re:File Extensions

[ncc74656]

...in other words, you broke the AVI standard to make it do what you want.

We didn't break anything:

http://msdn.microsoft.com/library/default.asp?url= /library/en-us/dx8_c/directx_cpp/htm/avirifffilere ference.asp

In fact, it turns out that Windows Media Player (v6.4, anyway, which is the newest version that I'll use) is able to play multiple video streams in an AVI simultaneously, which surprised me. One stream plays in the WMP window, while the other streams open additional windows and play in those.