Slashdot Mirror
VP3, Open Source Video at 200kbs
Honest Man
noted that intel is hyping VP3 as the first low bitrate open source video codec. 200kbs for VHS quality video sounds good to me, especially when I can apt-get it. But is DivX already to entrenched in this niche?
So quick, apt-get it before it gets banned!
A low-bandwidth codec might have more success than DiVX (which, while lighter than mpeg-2, is still 800 meg for 90-100 minutes at decent quality).
warning: not a troll, just an observation: Is this just a cheap ploy to sell P4's? This seems like Intel is just tooting their own horns about this technology, and claiming it's only for P4's.
I bet this is not "First Post."
Is DivX ;) entrenched in the market? Well, how many non-technical people have heard of it? How many PCs is it bundled with? It has a reputation for being primarily used for pirated video (regardless of the truth). So, the answer is a resounding "no, it isn't entrenched".
501 Not Implemented
More info here at http://www.on2.com
Yah, its open source for development but it costs $395 to license? You do the work, but we'll take the profits.
The page linked in the article is notably lacking in any resembling information on this codec. For more info try: On2's website
You can't get a blue screen on a black and white monitor.
The page says datarate as low as 200kb, but what's the average datarate and what's the worst case scenario datarate?
There are patents on the technology, which means it is of no more use to the open source community than True Type font hinting and MP3.
:)
I hope that they address the patent issues, and not just brush them aside like the DivX guys have done.
There's a reason the Xiph.org project is trying to develop a video codec too
The real open source VP3 site
The VP3 open source license
The VP3 license claims to be MPL derived. Would be interesting to see if it still fits the open source criteria.
Having to fill out any registration forms will push people away, and not being able to find the codec online will obviously throw people off. The easiest way to get around this would be to encode a bunch of movies, or tv shows.. in the name of the file put the URL to download the codec, and message everyone who is trying to download it, telling them where to download the codec.
With a bit of support from its users, this could easily take over as the common standard. With mp3's, people were only used to hearing about one specific codec (mp3). With video, people already know there are multiple kinds, each with different qualities. They know of real video, mpeg, divx, quicktime, and a few other formats. Throwing another one in the mix won't be surprising to them.
Also, after someone has downloaded a 600Mb file, they are more likely to go out on a limb and install a codec, than if they just downloaded a 3Mb mp3.
it may be open source, but it sure isn't cheap...
Intel's link takes you to on2.com's website where they have this to say:
With the VP3 for Windows codec, you can encode VP3 video and play it back through the Windows Media Player! The VP3 for Windows codec allows you to encode VP3 video using any Video for Windows compatible encoding application (such as Adobe Premiere and Virtual Dub) and play it back through the Windows Media Player. This version comes with limited email support. $395 USD"
The free open source versions can be found at www.vp3.com, but it looks like Intel is promoting them the big bucks version.
Consider this pice:
(e) Notwithstanding Sections 2.1 (a), (b), and (c) above, no license is granted to You, under any intellectual property rights including patent
rights, to modify the code in such a way as to create or accept data that is incompatible with data produced or accepted by the Original Code.
Yeah, that's real fucking useful -- we can view the code, but we can't improve it (incompatibly).
This is the problem with the "Open Source" movement -- it's become such a buzzword that morons like VP3 think they can make up licenses like these.
From the vp3.com website:
"With VP3, there are no platform limitations and developers are not required to pay the restrictive license fees that other open source codecs make mandatory."
restrictive license fees for open source...? They seem to have a very strange concept of the term "open source".
"We won't use guns, we won't use bombs, we'll use the one thing we've got more of and that's our minds" - Pulp
Where, pray tell, is the link to download the source?
With DivX, if you want to copy your video collection, your hard drive is likely too small, plus you need special equipment to record. The alternatives are open as the linked article demonstrates. Via KazaA (or Gnutella, or whatever), Video files (even a 22 minute Simpsons episode) take a long time to download via cable modem, and is not realistic for the majority of people who use dial-up connections.
Don't get me wrong, DivX ;) is great, just as MP3 is. All I'm saying is that the differences are big enough to prevent DivX from being entrenched at this stage of the game.
501 Not Implemented
> Am I missing something here????
It uses a modified Mozilla license, and I just downloaded source for free (reg. required) from here:
http://www.vp3.com/
On2 could have been there first, but they squandered their chances by charging money for a codec while everyone else and his mother were giving it away for free.
.bomb
Their technology was slightly better than the latest mpeg at the time, but marketting ruined another
Disclaimer: I used to have money invested in this company.
For informational purposes relating to the on2 codec, check out http://www.duck.com
200kbit/sec for video? so what. double that if you want VHS quality sound along with it!
Is that video *and* audio? Or is that video only? Either way, it seems too good to be true. Typically, 128kbps is the considered the bottom end for near CD quality for MP3 audio... at 200kbps for this VP3, if they have decent stereo sound encapsulated, that doesn't leave a lot of room for the video!
Even if that figure is for video data only, that seems way too good... 200kbps is barely enough to describe audio, let alone a decent representation of video! Don't forget, DivX takes about 10Mbyte/min or 1365kbps for audio and video at decent quality...
I wonder what the quality and resolution are truly like...
MP3 finally has a video counterpart - a file-compression algorithm that makes it possible to send large multimedia files over the Internet on demand. VP3 is the first open-source video codec to truly support VHS-quality video at bandwidths as low as 200 kbps.
isn't MP3 a patented, non-free algorithm? isn't that why Ogg Vorbis exists? so the only reason Intel is comparing VP3 to MP3 is marketing crap, right?
either that, or they are hoping people will compress millions of DVDs into VP3 and set up giant file-swapping services, that would be a video counterpart to MP3.
in other news, are there any side-by-side comparisons of VP3 and DivX? and how does Ogg Tarkin fit into all of this, now that there is an 'open source' codec?
-sam
burn the computers. go back to the abacus.
Ok, folks - since so many people seem to be having trouble with the concept, we'll go over it again. I'm no expert (feel free to flame me if I'm wrong - oh, you would have anyway...), but Open Source is not the same as free software. Open source merely means that your customer can obtain the source after purchasing the product. Free Software is (as its name implies) free. There are many great products that are open source, but not free. Similarly there are many free software packages that are not open source. It just so happens that a lot of software for Linux/BSD/whatever happens to be both.
Ok, back to my lurking.
Monday is a horrible way to spend 1/7 of your life.
If all this should have a reason, we would be the last to know.
For the most part, isn't DiVX illegal, in that it is based on Microsoft Intellectual Property... Namely, their proposal to MPEG-4, which was not accepted? Isn't the DiVX 3.11 codec, just a hack of the MS Codec? I believe 3.11 is just the older version of the Codec that allowed encoding to non .ASF formats or something like that.
;) ? In addition, its hard to defend DiVX in a court battle. I mean, how many teleconferencing apps do you know run DiVX? I'm sure Intel will be able to show that the main purpose of this codec has nothing to do with pirating movies, even though it could be a good use of it ;)
This new Codec developed by intel, is open, so there is no immediate legal issues pertaining to its use, unlike DiVX. Also, this may open the doors to commercialization. I mean... How many vendors do you think would want to release something called DiVX
I think it would have better market value knowing it was a codec developed by a real company, not a hack of someone else's work.
Besides, isn't the bitrate of DiVX like 910 kb/sec in most applications? I think 200kb/sec for the same quality is awesome.
Key difference being: WMA supposedly offers better/equal quality to MP3 at a lower bitrate, but nobody wants to be sucked into a proprietary format. Likewise Windows Media8 supposedly offers DVD quality video at like 500 kb/sec, but again, who wants to be sucked into a proprietary format? This new codec from Intel on the other hand is open.
Just my two bits...
It's available from their cvs server. Look at: this page for more info on browsing the CVS tree.
You can't get a blue screen on a black and white monitor.
With all the talk lately about various Codecs (divx, sorenson, and now VP3, as well as the "mpeg-4" stuff we've heard about (that may or may not be a codec :) )), I've been wondering...
...has anyone put together a good test suite to compare the various codecs at various bit rates? I'm thinking something that'd have some fixed-images (test patterns), some high- and medium-intensity moving images, lots of colors, simple and complex sounds, etc. Then put that file through all the various systems, at various rates, and compare the quality somehow...
Not that it'll really make much difference to me, as an end-user, since I'll just watch whatever someone has already encoded, but I'd be curious to see something a little more substantial and quantitative than just "sorenson's cool" sort of postings...
http://www.vp3.com has the real goods- Had you looked at the link on the bottom of On2's website, you'd have seen the link for the Open Source release of VP3. Open Source doesn't mean that they can't still be selling the versions of the codec that are "certified" (as in supported- they're offering limited support for the open source release...).
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
Uhh, yeah, it's pretty open source, in spite of what you may believe. Whoever modded this up obviously did about as much research as the poster.
// file: mice.h
#include "frickin_lasers.h"
When they can beat DiVX, I'll be impressed.
Two months ago I was compressing near-VHS quality at just over 250 kbps with DiVX. Could have gone down to 200 kbps if I had the time to tweak it a little more.
Now, if VP3 can do the tweaking for me, and is faster, then I'll be impressed.
Lots of missed info for a project that whose source was released in early September. Good to see it finally got noticed by /.
-first source is available on vp3.com. You must register to download (hrm).
-Its license is MPL derived, with some restrictions on IP for their patents. Also derivatives must always be able to play VP3.
-Its streamable with QT hinting.
-only currently available for Win and Mac. Port to *nix should be easy since there is code for OS X.
-Apple and Real will be supporting it in their players
http://cypherpunk:cypherpunk@www.vp3.com/vp3/tar/v p32.tar.gz
Standard VHS quality audio is encodable in something like 40-60kbps. It's nowhere near hi-fidelity. Now, stereo hi-fi stuff would probably require something like 300-400kbps for the video and audio as it'd need an MP3-like audio stream present.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
There was some discussion on the developer lists about getting them to allow us to release an OGG video stream using their codec. Right now, they're supposedly looking at re-working their license to make it where something like that might happen.
Otherwise, the best you can do with the current license is make a VP3 player/stream codec for Linux (Which wouldn't be a bad thing- I've seen the technology in action with RealPlayer 8 on Linux, playing some unbelievable streams from news.com.).
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
On the bottom, there is this link to the official site at www.vp3.com. The code's under a slightly modified MPL license- if you want much of anything in the way of support, you'll be paying them $395 for the "certified" version. Otherwise, it's as free and open as Mozilla is right now without the GPL license on some parts of it.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
On2 owns the patents for VP3. They're granting rights to anyone that wants to use the source and produce a codec for their platform. They're granting the rights to the patents for people that make improvements to the system so long as they don't break stream compatibility with the open source code in their CVS repository.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
Though of course Intel can easily use the Sony case as a precident.
___
If you think big enough, you'll never have to do it.
Bandwidth is a relative concept. For what you're talking about, it's the correct usage of the term.
For video, "low bandwidth" takes on a slightly differing meaning. 200kbps for a 640x480 plus audio video stream is something like 1/10 the bandwidth of what it'd take pushing it raw. And it can do 56kbps QCIF resolution feeds fairly well.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
I found out that the 'downloading' part is register locked. Big deal. guest:guest worked for me :-)
I love default passwords...(ahem, root:root)
This new video codec probably utilizes a lot of Intel specific SSE2 instructions. Maybe Intel is making this open source to encourage developers to use more Intel specific instructions. I hope the VP3 codec though doesn't require an Intel processor to work. It should at least have a back up algorithm that utilizes MMX so that those with AMD CPU's and others can use it.
Often companies say that their product produces "VHS" quality, but that is a bit subjective. For example some say that one codec sounds just as good as another (WMA at 64kbps vs. MP3 at 128kbps) but I can notice the difference immediately. And since this is a product produced by a company, not a standard, it probably won't be very popular.
Here is what I have so far : http://tv.cheema.com/vcr/ Its in early stages of development and you may find some problems here and there. I plan to release the source under GPL once I get my employer's approval.
Warning : The system above is on a slow uplink so some pages may load slowly. At some point I will start using mod_gzip.
OK, so the source code is available. That's a start. But it's not truly free. It's encumbered by patents and other restrictions.
So, given a code base for reference (ala reverse engineering), all we need is for somebody outside of the US, where software patents don't apply, to develop a GPL replacement written from the ground up, but which is unofficially 100% compatible with the VP3 format. Ideally, it may even be possible to work around their patents somehow, which would free content producers from having to pay royalty fees (as with MP3).
Of course, that's assuming that VP3 is really a format worth emulating compared to the patent-free video codec the Ogg Vorbis people are working on. But hey, even they may be able to gain some insight from looking at the VP3 code.
http://www.on2.com/quicktime/trailers/
Rather than watching the movies in a tiny window embedded in a web page, visit this page, download them, and watch them in their own player as large as you want. Personally, I'm very impressed.
Dear Slashdot: next time you want to mess with the site, add a rich-text editor for comments.
And the results - for the same file, at 910 kbps, indistinguishable quality, both had minor artifacting, etc, but looked pretty good full screen, and looked great at default res. The big difference was time to encode - divx took 6 1/2 minutes to encode the clip I selected, VP3 took 11, and size - divx was 20.7 mb, vp3 was 29 mb. All other things were equal, I used Virtual Dub for both, same video clip, and the default encoding parameters for both (Medium for speed/quality in DivX 4.0, Fast Encode for VP3). My computer's a Celeron 566, 256mb RAM, running Windows 2000 SP2.
Check out Treesandthings.com for offbeat news
VHS has several quality modes. VHS in its purest form is mono-EP. Low sound quality, low video quality, lots of stuff fit on a tape. Now there's Hi-Fi VHS, stereo-SP. High(ish) sound quality, high(again,ish) video quality, not so much fits on a tape. So which is VP3? Probably the former.
[insert witty comment here]