VP3.com: Future VP3 Releases To Be LGPL

sudog writes: "According to this vorbis-dev posting and The VP3 Homepage VP3 (QT5-type movie compression scheme) is now under the LGPL! What's not clear is whether they intend to offer it guaranteed royalty and patent free to the community. They're actively looking for help, too. Does this mean that we no longer need the OGG-Tarkin to save us from our movie-less, video-app-less emulating?" Of course, they don't say starting when, exactly.

Comments

Looks like a good thing to me.

[noser]

I have never heard of this codec, but it seems to me that this is more or less what the LGPL is intended for. Take a quick look at the LGPL and note this section:

For example, on rare occasions, there may be a special need to encourage the widest possible use of a certain library, so that it becomes a de-facto standard. To achieve this, non-free programs must be allowed to use the library. A more frequent case is that a free library does the same job as widely used non-free libraries. In this case, there is little to gain by limiting the free library to free software only, so we use the Lesser General Public License.

(Emphasis mine)

Seems to me that the people at VP3 would like as many people as possible to start working with their codec, allowing it to gain ascendancy over other codecs so that someday they will be able to make money selling their own "enhanced" version. Not a bad deal for GNU, because we get something badly needed. I hope that we start to hear more about this codec being used in some interesting projects in the future now that it has become more available.

Re:I'd never trust anyone except Ogg

[JabberWokky]

DivX ;) has had a very bumpy and very forked road - primarily because it wasn't a file format so much as a hacked DLL that was a stolen Microsoft test codec, with the audio replaced with (originally) MP3. Since there was no formal definition of the file format, it's fragmented pretty heavily with different encoders producing output that works on some of the variants, and other encoders working on different variants. At least two efforts have been made to rewrite the whole thing from scratch, and at least two alternate sound codecs exist (thus the common cry of "there's no sound in that file!"). There are questions of patents being filed even now (since it was a "format" swiped from a corporation), and, all in all, the situation has been very tangled for a single, real standard to emerge.

LGPL grants the same public use that the GPL does, except you can also combine it with commercial software (you have to release source only for the modifications done to that specific part of the code). It's not a "bastard license", but rather a compromise to allow commercial software to link to fundimental system libraries and run on a Free system. YMMV on what you think of that, but for things like file formats and reference code to file formats, IMO, the LGPL is the best license around - it keeps the whole thing open, including any changes that anybody makes to it, so the standard is open for the whole world no matter who uses it in any application.

I'm personally of the opinion that an LGPL library to read a few types of XML documents (a word processor format, a spreadsheet format, a bitmap with annotations, a vector art format and a vector engineering format) should be made, and maintained by all major office suites, probably starting off with the various open source projects. Even if a company didn't use the exact code, it serves as reference code for compatability tests and extensions.

--
Evan

This doesn't obsolete Tarkin BTW :)

[Skuto]

Tarkin is currently working on bringing new technologies such as wavelets and 3-d transforms into video coding. It's not finished yet, but it offers more possibilities for really new technology and further development.

While this is great news, it by no means means that Ogg Tarkin suddenly is obsoleted :) VP3 is available now though, and Tarkin isn't.

--
GCP

Re:Well

[Anonymous Coward]

Failing website? Kuro5hin's star continues to rise, while Slashdot continues to stagnate. K5 gives its users complete control over the submissions queue, whereas Slashdot sits on an important VP3 story for days and posts stuff about foam-rubber computer cases instead (LOL!) The Slashdot crew isn't made up of bad people, but they're yesterday's news. Kuro5hin.org is where it's at.

Re:ogg tarkin is somewhat dead.

[jonathan_ingram]

Tarkin is not dead.

Tarkin is in the same state that Vorbis was 3 years ago. No-one sensible thinks that it should be competing with MPEG-4/Sorensen/VP* at the moment. No-one connected with the project (only a couple of people, working in their spare time) has been promoting this project as competitive -- only some losers who hype every piece of open source software, no matter how far along in development the software it.

Come back in 2/3 years, and Tarkin will be looking much better.

Re:Patent protection - from Vorbis mailing list

[Anonymous Coward]

VP3 isn't their newest codec

so you get
1. the lgpl'd version or
2. the closed sourced, expensive licensed but more powerful new version

Re:Well

[Anonymous Coward]

Have you been to Kuro5hin.org lately? They have more users, bandwidth, and processing power than they ever did under OSDN. They've got several paid sponsors, as well as an innovative text ad system where users can submit their own ads which are displayed non-intrusively on the front page. People have been having a lot of fun with it, submitting joke ads, etc. And it all helps the site out.

Factor that in with an open submissions queue, a journal/diary system that actually works, a fair and equitable moderation system, and a virtually troll-free environment, and one wonders why Slashdot should even be read. For all practical purposes, Slashdot is dying.

Not really streaming, but progressive download

[benwaggoner]

A nitpicky point - VP3 as implemented in AVI and QuickTime files is designed for progressive download, not true, real-time streaming. Thus, you get the classic movie trailer wait-awhile-and-play experience, but without the ability to do random access over long files and that kind of groovy stuff.

Good support for real-time streaming would require a native packetizer to build a hint track that the (open source) Darwin Streaming Server uses to determine packetization of the stream, and which helps loss recovery and other good stuff.

Adding a native packetizer for VP3 would be an excellent open source project for the codec.

Re:What's with XviD?

[rsborg]

XviD falls prey to the problems inherent with MPEG4, as previously discussed here. Fix that, and I'm all up for it. (personally, I don't see any way around the issues)

In the interim, there's Ogg Tarkin, but it looks like they're too busy with Vorbis right now.

Re:XviD as alternative, Ogg Tarkin in the future

[Pathwalker]

I wouldn't have much faith in the tests you link to. I looked at the settings that the tester used, and it looks to me like they deliberately crippled vp3.

To quote the article:

"In VP3 I set the bitrate, keyframe interval to 9999 seconds, auto keyframe turned on as well as quick compression"

In my experience VP3 only gets noticibly blocky (the tester's major complaint) when it is prevented from creating a keyframe when it wants to. Here, they pretty much prevented vp3 from generating keyframes at all. The keyframe interval should have been left BLANK not set to a stupidly high number.

Additionaly, there is another menu of keyframing options (the one he should have used to set the adaptive keyframe rate rather than locking it) of which he writes nothing. Here, I probably would have set the minimum time to about 1/4-1/2 second, and set the maximum time to the highest supported number.

Furthermore, There is an image quality control which controls the tradeoff between image quality, and the risk of dropping the frame rate. No mention was made of the setting of this control, but the complaints about low detail make me wonder what it was set to.

Finally, turning quick compress on does lower quality. For a test which did not involve encoding speed, I have to wonder why the tester chose to turn that option on, as it trades off quality for faster encoding!


I use vp3 to encode DV streams (in Quicktime) for viewing over the web. Vp3 is a very good quality codec, superior in many cases (unless you are streaming from a QTSS, or the source was shot under unusual light conditions) to the free version of Sorenson. It is excellent under these conditions.