Slashdot Mirror
Xiph Releases Ogg Theora Alpha-3
ArcRiley writes "For more than a year Xiph hackers have been working on Ogg Theora, an improved version of On2's VP3 video codec. Alpha-3 includes several bitstream changes, VP3 to Theora "upgrade" utilities, and is now supported by Xine, MPlayer, and Real's Helix Player. We're nearing Beta-1 where the format will be frozen, fully documented, and it'll be ready for everyday use."
I'm happy to announce at long last the release of theora alpha 3. This incorporates all the bitstream changes we wanted to make both for future encoder improvements and to permit lossless transcode of VP3 content. This is an important milestone for us on the road to a stable release.
As this is an alpha release we are again providing sources only. See the files section of downloads. This version requires libogg 1.1 or later and libvorbis 1.0.1 or later.
Also new in this release are a set of experimental tools in the win32 directory contributed by Mauricio Piacentini. This includes a transcoding tool for avi-encapsulated vp3 video which also works on linux.
We hope to not make any further incompatible bitstream changes, but this is still alpha code. Don't use this for content you're not ready to re-encode!
Thanks to everyone who contributed!
Well in eg a portable music player, each codec takes more silicon and silicon development costs lots of money. Either that or do it in software and then the batteries won't last 5 mins.
Unless u could do it in firmware - perhaps FPGA?
Posters recognized by their sig,
I notice that the CBC's Quirks and Quarks radio show supports Ogg. Figures that a science program would be clued-in on latest developments. Hopefully the rest of the CBC will catch up--they still only offer the Big Ugly Three (and sometimes only Real).
One line blog. I hear that they're called Twitters now.
VP3 was meant to be a Real-killer, rather than a movie format - streaming. How does DivX work at very low bitrates? Is it acceptable at modem speed? Does it stream well?
Posters recognized by their sig,
Great, sad, point. But remember that as these devices get more complex, the time when some smart fella or lady throws linux on it get more near. Eventually, it won't matter that The Man doesn't support OV. As technology improves, the open source community has more places to innovate and use the best compression music among other things. It's only because of open sourcers creating such things as OV that arent the most useful now that we will ever have hope of such things being useful ever. I'm just glad they are ironing out wrinkles before I need their stuff to work.
Xiph is a group (I believe European or something to get the odd-to-my-USian-ear names) that puts out codecs.
Ogg is a wrapper format that they put out. It serves much the same purpose as QuickTime, AVI, or ASF does. One wraps it around an encoded stream of audio or data. Currently, Ogg is mostly commonly used to contain audio data encoded with the Vorbis codec, which is notable for tending to sound better than MP3, being patent-free and having a completely free implementation for anyone to use.
Theora is a video codec also put out by Xiph. It is based on an older, originally proprietary video codec that was donated to the Xiph project. I'm not sure how it measures up to existing video codecs.
Alpha-3 is, I think, pretty self-explanatory to a developer. It's an alpha release, so the developers are leaving open the possibility that they will make large changes (unlike beta software, where the software should be considered ready, and only lacks feedback from a broad base of people). It is the third alpha release.
Oh, yes. I love the Far Side strip about Ginger. Remember, though, that saying "awk", "sed", or "grep", which sound quite reasonable to people on Slashdot, sounds absolutely bizarre to most folks.
May we never see th
My Rio Karma works just fine with Ogg. In fact, all I have on the thing is Ogg.
It works incredibly well, and with 20gigs for $250 shipped, and a Java-based interface program (which runs on FreeBSD and Linux), I'm very happy with it.
You are right that the 'brand names' don't support these formats very well. This is why you should probably look past the brand names and check out the little guys...
Although the code of Xvid of xvid is GPL, XviD is still an ISO MPEG-4 compliant video codec. That means that the patent holders of MPEG-4 can still demand a fee for its use.
Theora doesn't have such limitations.
That's Creative Labs for you. I'm still waiting for them to make release quality (i.e. not beta ones a year after the fact) Windows 2000 drivers for my DXR3. It works great under NT4, but I haven't used that for years. This issue and their crappy drivers for the Soundblaster Live on SMP systems has convinced me never to buy their products again. They don't fix known problems with their existing products - your only hope is that splashing out more money for the next generation will resolve the problem. They're a very poor company.
Xvid is Free Software, but it's an implementation of MPEG4, which is patent-encumbered. It's the same reason that Ogg Vorbis is seen as the champion of free music, although there are GPL MP3 encoders.
I have a positive feeling about Ogg Theora. Three days ago I installed binary only DivX for linux (closed source), and I am not very happy about it. I want video codec at top performance, optimized for CPU and maybe even 64bit platform later this year, not a 32bit binary pentium-only crap with possible vulnerabilities.
There you are, staring at me again.
These things are not implemented yet, and will probobally not be useable on earlier Beta releases either, but as of Beta-1 the bitstream will not change in future-compatable ways. That is, while some optimisation fields won't be supported yet, no new fields will be added. Future players will always be able to play media encoded by the Beta releases. The same is not true for movies encoded with the Alpha libraries, so Beta-1 is really the first point where it should be used for distributed movies.
The 1.0 release will include support for atleast decoding these optional fields, it'll likely use them all too for encoding, and should be considerably higher quality than the VP3.2 codec from which it started with. It'll always, however, be able to upgrade VP3.2 media to Theora and, again, always be able to play media encoded with the Beta releases.
The Soundbridge does not support Apple's DRM, only unencumbered AAC files.
That thing looks pretty huge, so no thanks. And besides, if I wanted something with geek-appeal, I would buy Rio Karma
- Supports Ogg Vorbis
- Supports FLAC
- Has _Ethernet_ plug
- Has 20GB HDD
Neuros might have a Linux-version of it's software, but if the player appears as a regural HD to the OS, why would you need dedicated software?
Lesbian Nazi Hookers Abducted by UFOs and Forced Into Weight Loss Programs - -all next week on Town Talk.
At beta-1, the format of the Theora bitstream will be "frozen and fully documented". The reference software that implements the format will continue to be bugfixed/improved/tweaked, but they will not make any incompatible changes to the data format. Ogg files encoded using the beta software will be compatible with v1.x decoders - this isn't necessarily true of the alpha releases.
('twas the same with Vorbis audio, as it was developed, if I recall correctly).
You might try these options to the sample encoder: --audio-rate-target --video-rate-target --audio-quality --video-quality This is not two-pass but it may be close enough. I always seem to need to some fudge with XVID using mencoder so this might end-up being as close as you need in practice for now.
I was about to post the same thing, except I was going to add that this limitation can be easily overcome by a quick Google search for m4p2mp4.exe. Apart from that, the only formats I'd want supported that this doesn't handle are .ape (could just convert all my .apes to .flacs though) and .ac3 (for rips from music DVDs, I don't like to transcode them if I don't have to). I think I may have just found the digital-music-in-the-den solution for me, thank you (to the grandparent post)!
Hey kids, there's only 5 days left 'til Yak Shaving Day!
Virgin Radio has been broadcasting in 96k ogg vorbis for quite a while nowe n/ogg .html
;))
http://www.virginradio.co.uk/thestation/list
The advantages for them are quite clear: no patent costs and more listeners who just want to support ogg (ok, maybe not many, but still
Rob Lanphier
Developer Support Manager
RealNetworks
Neuros might have a Linux-version of it's software, but if the player appears as a regural HD to the OS, why would you need dedicated software?
You can put the files on the harddrive or ramdisk, but if you want them to show up in the UI, you have to build the appropriate databases. The DB schemas are documented and there are a few different implementations:
- The windows-only one included with the Neuros
- Positron, a command line utility in Python.
- A Multi-platform GUI in Java, whose name escapes me.
As for size, it's larger than an I-Pod, and probably too big for most people's shirt pockets, but it's not all that big, and is fine in a beltcase, purse, or in the car, etc. Depends on how you want use it, I suppose. I chose the Neuros over the Karma mostly because of apparent company support: Digital Innovations actually seems to care if it works with Linux or the Mac, while Rio's support is pretty haphazard.If you want a free wavelet based codec the BBC have just released one under the GPL. Its called Dirac and its on sourceforge. Its still alpha code but...
http://sourceforge.net/projects/dirac/
Samsung's patent
Each stakeholder has patented a method or slight spin on the basic technique, and so certain claims in each patent could possibly be applied to your arithmetic coder if they wanted to go after you.
You may need to wait 5-10 years before bundling an implementation in a package with high-visibility that isn't designed for educational or experimental use.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
What you describe is Trademark law, which is probably the sanest branch of north american IP at this point. You cannot trademark common words or phrases (modulo certain exceptions) so an effort is generally made to prevent brand names from becoming same by prosecuting uses of the trademark that don't refer specifically to the actual product.
Copyrights belong to the author (or sponsor) automatically and can only be given up voluntarily or lost when the rights period expires, which is now some significant time after the author dies.
Patents are granted on a first come, first served basis to whoever applies for one and provides a monopoly on the implementation of a particular method for a fixed term.
Neither copyright nor patent rights are contingent on enforcement the way trademarks are. Holders of these two rights can and do choose which infringements to pursue.
This is the problem with MPEG-4. We can avoid the copyright issue by writing an open source version from scratch, since the standard is at least published. We don't have to call it MPEG-4 so there are no trademark issues, although while the MPEG logo is trademarked in the US, one can refer to the specification because that itself is not a trademark and because there is no attempt at confusion.
But there is no way around patents because they grant a monopoly on implementation rights. Just because you wrote your own doesn't mean you don't have to buy a license, or that you won't be forced to buy one sometime in the next 20 years. If you live in a jurisdiction that doesn't enforce patents, you're fine for now. If you just want to trade movies underground, you're probably fine because there's safety in numbers. But if you're like me, and want digital media to be as easy and ubiquitous as webpages; something anyone can do, something you don't need permission for, you need a something that's Free as in Freedom and Free as in Beer. Something like Theora.
Nope, there is a royalty due on every encoder and decoder as well as every minute of content. Get the facts here.
DivX has the largest market share for pirated content distributed via P2P networks. But it isn't a meaningful player in commercial markets.
Among other things, since DivX is based on the AVI file format, it doesn't have native support for streaming. It's really a CD-ROM format at its heart.
Also, DivX uses the MPEG-4 part 2 video codec, which is being superseded.
For commercial codecs, MPEG-2 is probably 95% of the market today. And the battle is now raging between Microsoft's VC-9 and MPEG-4 AVC (aka Part 10 aka H.264) as to which will replace MPEG-2 over this decade.
My video compression blog
Well, once they've locked down the bitstream, folks can start innovating on building better encoders compatible with the bitstream format. 2-pass encoding would be an obvious initial step.
Codecs can get a LOT better due to encoder innovation. MPEG-2 has roughly tripled its compression efficiency since standardization.
My video compression blog
Once again, I'll restate that *it can* do proper mixing. You are dealing with is a limitation of your hardware, not Linux. It just goes with the territory. It's almost like the people that say: "Linux doesn't do 3D because my obscure video chip isn't supported. X sucks!"
DMIX is a mess. Unfortunately, that's the price you pay. It's not that Linux can't do the mixing that you ask. Maybe, given time, DMIX will improve. It's a relatively new component for ALSA though. You may want to consider paying for 4-Front's drivers that do it if DMIX doesn't float your boat.
http://opensound.com/
These drivers support softmixing reasonably well, but aren't free. They are not a proper alternative to hardware mixing, but they will work. They apparently have an ALSA emulation system, but it's just a "play project." I don't know if I would want an ALSA wrapper sitting on top of a sofware mixer though (their Virtual Mixer).
Your sound chip is essentially like a Winmodem, built specifically for Windows. Even some chips that do hardware mixing on Windows don't mix on Linux because of lack of proper programming documentation. What can the programmers do about that? Again, Linux's multimedia capabilities are not to blame for a driver issue. A good amount of other users don't have the same problem that you do because they have well-supported hardware.
It's not something that will change unless hardware developers take a different design approach, and a different approach to releasing their programming docs.
I must tell you that I understand how you feel though. Really - I am sympathetic to the problems you are having. It's a very frustrating situation to have to deal with this sound issue. I've been through it before, as well. This was before the CS46xx drivers were as mature as they are today. I used the 4-Front drivers to hold me over until my hardware got better ALSA support, and it paid off. If the sound issue is giving you that much of a headache, the $20 may be money well spent. That isn't to say that the 4-Front drivers are without problems though, but it is worth testing them before you fork out $20 for a version that doesn't expire after a period of time.