Dirac 1.0.0 Released 127
dylan_- writes "According to their website, 'Dirac is an advanced royalty-free video compression format designed for a wide range of uses, from delivering low-resolution web content to broadcasting HD and beyond, to near-lossless studio editing.' Now a stable version of the dirac-research codebase, Dirac 1.0.0, has been released. The BBC have already successfully used the new codec during the Beijing Olympics and are looking to push it to more general use throughout the organisation. The latest version of VLC (the recently released 0.9.2) has support for Dirac using the Schroedinger library."
really? (Score:3, Insightful)
Remember when we all used GIF until somebody came out of the closet with a patent claim. How can we be sure about this one?
Re:really? (Score:5, Informative)
Read their site. From the FAQ [diracvideo.org]:
Do you infringe any patents?
The short answer is that we don't know for certain, but we're pretty sure we don't.
We haven't employed armies of lawyers to trawl through the tens of thousands of video compression techniques. That's not the way to invent a successful algorithm. Instead we've tried to use techniques of long standing in novel ways.
What will you do if you infringe patents?
Code round them, first and foremost. There are many alternative techniques to each of the technologies used within Dirac.
Dirac is relatively modular (which is one reason why it's a conventional hybrid codec rather than, say, 3D wavelets) so removing or adding tools was relatively easy, even though this may mean issuing a new version of the specification.
Re: (Score:3, Interesting)
The real question is, how does it fare against good H.264 encoders e.g. x264? And how are the encoding speeds?
The few comparisons I've seen put H.264 as having the edge when it comes to both, but not by a lot.
Re:really? (Score:5, Informative)
Encoding and decoding is presently fat and slow. It's very much in development.
Re:really? (Score:5, Interesting)
Since you claim this I assume that you tried the 1.0.0 already - I watched the promo vid, and it says the BBC is using the codec to handle HD content over their standard def infrastructure at very low latency (a few ms, if I remember correctly).
Nonetheless, this seems to be an interesting thing to keep an eye on, because the codec specs address good compression especially for very high bandwidths, which is going to be an important issue for movie post production/processing, HD content and the likes. The promo vid is well worth watching.
Great, but... (Score:5, Funny)
I tried using the Schrodinger library but I'm uncertain it works. Plus, I can't find my cat.
Re:Great, but... (Score:5, Funny)
Plus, I can't find my cat.
Be glad for that, as long as you dont you might still have one.
Re: (Score:3, Funny)
I had success integrating Dirac, and got a result of 1.
Re: (Score:1)
Re:Great, but... (Score:5, Funny)
I tried using the Schrodinger library but I'm uncertain it works. Plus, I can't find my cat.
I hate to tell you, but your cat is dead. And/or not.
Re:Great, but... (Score:4, Funny)
You may have loaded libheisenberg instead. They sometimes become entangled.
Re: (Score:2)
You may have loaded libheisenberg instead.
Are you uncertain of that?
Re: (Score:1)
You just your whole cat! Are you sure this works?
*ducks*
Re: (Score:1)
0xBBCD (Score:5, Interesting)
I see the first 4 bytes are 0xBBCD.
British Broadcasting Corporation Dirac.
Re: (Score:2, Interesting)
Did you mean to say the FOURCC (which is usually not the first four bytes) is 'BBCD'? 0xBBCD is usually two bytes...
Re:0xBBCD (Score:5, Informative)
the fourcc is 'drac'
Re:0xBBCD (Score:5, Insightful)
Isn't that just 2 bytes? :)
*nibbles on parent's geek card*
Re: (Score:2)
Re: (Score:2, Informative)
The size of a byte doesn't have to be 8 bits, though it usually is.
Re:0xBBCD (Score:4, Funny)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
But it's not communicating between different platforms. Or rather, it need not be. The amount of hardware on which a byte equals 8 buts, but still use '9600-7-e-1' or 9600 baud, 7-bit, even parity, one stop bit serial communications on their primary serial port is quite large, even if they're only talking to another such machine. And irrelevant of the internal computation byte size, most machines have to deal with flat ASCII text, which is typically 8-bit bytes.
My point is that byte size is also protocol de
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Open source overkill (Score:5, Interesting)
From the FAQ:
What are the license conditions?
The Schrodinger software is available under any of the GPLv2, MIT or MPL licences. Libraries may also be used under LGPL.
Sounds like someone wanted there to be no question about whether it was open source.
Re:Open source overkill (Score:5, Funny)
The Schrodinger software is available under any of the GPLv2, MIT or MPL licences. Libraries may also be used under LGPL.
Sounds like someone wanted there to be no question about whether it was open source.
Sounds to me like the license exists in multiple states at once, which may be exactly the way Schrodinger would have liked it.
Re: (Score:2, Insightful)
I am not sure, but isn't MIT one good enough to relicense it to (L)GPL or MPL?
Re:Open source overkill (Score:5, Informative)
"Each contributor grants you a non-exclusive, worldwide, royalty-free patent license under the contributor's essential patent claims, to make, use, sell, offer for sale, import and otherwise run, modify and propagate the contents of its contributor version."
So if you use it as a GPL licensed library you can't get sued by the BBC or other contributors to the code.
Re: (Score:1, Informative)
Yes - GPL says that explicitly, bit MIT (copyright holders) grant you:
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so ...
So they GRANT you to do with it more less an
Re: (Score:2)
Isn't that clause new to the GPLv3?
Re: (Score:2, Informative)
Re: (Score:2)
I seem to recall something about a Novell-MS dealbreaker being in the GPLv3 that, if adopted by the kernel itself, would mean a lot of fireworks, but I might be thinking of something else.
I'm definitely not talking about the Tivoization clause.
Performance? Benefits? (Score:4, Insightful)
How does it stack up to other codecs?
Do we need another codec?
Re:Performance? Benefits? (Score:5, Insightful)
Re: (Score:2)
Exactly, Matroska is great and all from a freedom standpoint, but technically it's far behind the encumbered ones.
At least we have ogg for audio, it seems like nothing can beat it in terms of quality/bitrate:-)
Re: (Score:1)
Re: (Score:2)
Yeah I realized that after I wrote this:-) I meant Theora. Oopsie.
Re: (Score:1, Interesting)
Also, try to encode with the newer versions of Theora. It has gotten much improvements in the last year - quality problems were never in the decoding as some will have you believe, but that the encoder pretty much sucked.
Not sure what you expect and I'm no video buff... but it sure looks a LOT better.
(You may still be right, of course. I've just found that 99% of all who state anything about anythings quality usually have formed their opinion once, maybe years ago, and then keep on repeating it).
Re: (Score:3, Interesting)
Matroska is not a codec. It is a container format, and it beats any closed-source competitions hands own on features (e.g. as far as I know it is the only format that supports embedding custom TrueType fonts for subtitles).
The best video encoding combo right now is:
- Matroska as the container
- H.264 for video
- Ogg Vorbis for audio
- ASS for subtitles
Re: (Score:3, Interesting)
How does Vorbis really compare against AAC? Besides the whole royalty/patent free issue, does Vorbis really beat out AAC? (Ignoring royalty/patent issues here because you also mentioned H264)
Re: (Score:1)
If it's really important to you, don't guess; do your own test. The Hydrogen Audio folks know something about this.
http://wiki.hydrogenaudio.org/index.php?title=Choosing_the_best_codec [hydrogenaudio.org]
Re: (Score:2)
Re: (Score:2)
According to Wikipedia [wikipedia.org] aoTuV beta 4 Vorbis encoder beats WMA, MP3 and AAC at 64-128 kbit/s.
Free as in beer or speech? (Score:2)
This came up in yesterday's discussion of the Canonical codec pack.
Standardized codecs, like VC-1 and H.264, have full open specifications and typically even reference source code implementations that can be reused in a variety of ways.
However, they also require patent fees depending on use and jurisdiction.
The issue of free software has always been asserted to be about "speech, not beer" but it seems like there's an assumption that it has to be free as in "speech AND beer." I'm sure all kind of arguments c
Re: (Score:2)
How many copies of mplayer do you think are out there? Its a lot more than 100K copies.
But here is the real rub. Even if you pay the fees, they give *no* guarantee what so ever that you are not infringing some other patent.
Also these
Re: (Score:2)
You are right and wrong. If I want to keep MPEG-LA happy, i need to pay quite a bit of money if there are more than 100K downloads IIRC. For a open source project that is *not* selling the code it is a lot of money. Most OS projects are not done with money, but with time.
But if you're just distributing the source, you wouldn't need to pay the patent. If it's a real product, or if users are compiling it into one, is doesn't seem to be an infringement on "freedom" in the classic RMS definition to wind up paying a fee. You still have full control over the code and technology used.
But here is the real rub. Even if you pay the fees, they give *no* guarantee what so ever that you are not infringing some other patent.
Theoretically true, although that hasn't happened much in practice, at least in this space. And Theora and Dirac are in the same legal position, and don't have the market effet of lots of companies
Re: (Score:3, Interesting)
But if you're just distributing the source..
So now I can't also distribute binary but my freedom is not affected? I don't think I would want to test source=ok, binarys=!ok as far as patent law is concerned with my wallet. Economic harm is all thats needed if software patents are valid.
Theoretically true, although that hasn't happened much in practice, at least in this space.
So you pay a crap load of money and only don't get sued much? Thats a raw deal. There has been at least one case I know of with mpeg4 | h.264, and thats a lot more than what both theora and dirac have had to deal with. Add the fact that theora is based on VP3 with a act
wavelet codec (Score:2)
Re:Performance? Benefits? (Score:5, Informative)
How does it stack up to other codecs?
As I say below, unfortunately the quality is lacking compared to modern codecs like H.264 and even (dare I say) VC-1. Apparently that's just the nature of using wavelets. While they give a very natural style of compression on still images (JPEG-2000, etc), they do not translate well to moving sequences because, unlike all other current codecs, the image is not broken up into blocks that can then be tracked and diff'd in time. Still, it'll be interesting to follow Dirac, if only because they're taking a radical new approach with only Michael Niedermayer's Snow as a peer.
Re:Performance? Benefits? (Score:5, Informative)
However a great codec without a good encoder is no good at all. But its early days yet considering h.264 has been around for 5+ years.
Re: (Score:2, Funny)
The codec is new, give it a few months.
Early DVDs looked like shitty 90% compressed jpegs too, you know.
Re: (Score:3, Interesting)
Re: (Score:1)
You seem to be confusing motion estimation/compensation with residual coding. Dirac does break the image into blocks, using overlapped block motion compensation [wikipedia.org]. However, the residual image is coded as a whole, thanks to wavelets. This should greatly r
Re:Performance? Benefits? (Score:5, Informative)
Dirac employs wavelet compression, instead of the discrete cosine transforms used in most older codecs (such as H.264/MPEG-4 AVC or SMPTE's VC-1). Dirac is one of several projects attempting to apply wavelets to video compression. Others include Rududu [2], Snow and Tarkin. Wavelet compression has already proven its viability in the JPEG 2000 compression standard for photographic images.
Yes it does :|
Content (Score:5, Interesting)
Underwhelming Results (Score:1)
News from OGG Theora, too! (Score:5, Informative)
Dirac isn't the only royality-free, patent-unencumbered video codec there is - Xiph's OGG Theora has been around a while already, yet failed to impress quality-wise up until recently. There's some really cool development going on however, and you may see some of the results achieved over there: http://xiphmont.livejournal.com/35363.html [livejournal.com]
It's noteworthy that the changes made only affect the ENCODER, thus no changes to the DECODER (the part of a codec all applications used to play back files have included) are necessary. This bodes very well for HTML5, which will include some support for Theora on at least Mozilla (and iirc Opera) browsers.
Re:News from OGG Theora, too! (Score:5, Interesting)
Re: (Score:2)
Except that Theora not only start being behind the then-emerging H.264 and VC-1, but its implementations are launch were quite a bit weaker than even Xvid of the era, and have essentially stagnated.
And since then, it's fallen even further behind; the implementations of standardized codecs has been improving a lot more each year than Theora, as have proprietary codecs like the later entries in On2's VPx series (Theora was forked from VP3; On2 just announced VP8).
There's been some interesting work in the last
Re: (Score:2)
A patent free codec is still good as long as its pretty close to say mpeg4. It would end up in a lot of games etc, and it keeps everyone else playing nice because there are alternatives. It not
Re: (Score:2)
AFAIK, the deblocker in VP3/Theora is NOT an in-loop filter, but just a simple post-processor, with just a little more smarts due to being codec-specific.
VP3/Theora also has some significant inherent DISADVANTAGES over MPEG-4 ASP, such as lacking B-frames (which Xvid uses to great effect). And yet, while VP3/Theora isn't competitive with H.264 o
Re: (Score:2)
My mistake. It is rather simplistic, but it is in fact an in-loop deblocking filter.
Interesting you should say that, because I just looked through the specs and I see NO indication at all that Theora can use "adaptive block sizes" as claimed in the parent post. In fact it repeatedly says Theora uses fixed 8x8 blocks, just like MPEG-1/2/4ASP.
*Ah
Re:News from OGG Theora, too! (Score:5, Insightful)
Both Theora and Dirac have plenty of space to move with regard to encoders.
However there is no easy way to measure "distortion" of the encoded image that matches the human visual system all that well. (unlike audio). But I expect most codecs to get better in the next few years because of encoders. (including h264).
Ironically h264 does so well because of the availability of a free, fast and good quality encoder done my the community. Not the license owners.
Re: (Score:3, Insightful)
However there is no easy way to measure "distortion" of the encoded image that matches the human visual system all that well. (unlike audio).
How do you objectively measure psychoacoustic distortion? Do the same techniques not apply to vision simply due to unknown constants or is there some more fundamental reason?
Re:News from OGG Theora, too! (Score:5, Informative)
But the human visual system is a *lot* more complicated. IIRC about 1/3 or our brains are used for visual perception. Currently we use PSNR (Peek signal to noise ratio) as a measure. But this has been shown many times to be a very poor indication of what we perceive. One example is blocking. Blocking cause straight lines to form in the image and our brains lock on to them far more quickly that other artifacts.
Next is the colour and the 2d nature of a image. Then add that the eyes do a bunch of preprocessing on motion perception and its getting quite difficult. Finally we have the method of comparison. Which often involves comparing still images from the video stream. Yet if thats a high motion scene the codec might be better off encoding these frames with low quality because we can't perceive the quality loss combined with fast motion.
Lets also not forget how many people think youtube is good quality or at worse, good enough!
Re: (Score:2)
However there is no easy way to measure "distortion" of the encoded image that matches the human visual system all that well. (unlike audio)..
I'm not sure I agree with that...and I think the fact that there are people who *can* tell the difference between a 256kbs MP3 and CD-audio and those who *can't* perhaps shows that there's no easy way to map quality of audio onto something that matches human perception. There are plenty of technical ways however, both for audio and visual. I'm not sure where you're getting this from.
Re:News from OGG Theora, too! (Score:5, Interesting)
I don't doubt that some people can tell the difference between flac and mp3/ogg/aac. But the true number is far less than the claimed number (do a proper blind test to really find out). Also you don't design codecs for 0.5% of the population that can hear the difference, but for the 90% that can't and the other 9.5% that don't care.
Now its a fact that PSNR is used in most encoders. Its also widely recognized that it is not a good measure. I have done my own image compression and got better PSNR than jpeg per bit, and yet it looked far worse.
So I'm not really sure where you getting the idea that is even in the same category as audio.
Re: (Score:2)
In fact you do design for the 0.5%. Testing of codecs specifically uses expert listeners, with an in-depth double-blind testing setup, hidden anchors, and the like. Of course you aren't going to please everyone, all the time, but codec developers certainly do try their best to do so.
Re: (Score:2)
In fact you're wrong.
MP3 was never designed to be indistinguishable from CD-audio. It was designed to sound GOOD at even lower bitrates (64kbps mono). MP3 has several limitations that prevent it from doing so. It's immensely ironic that people are now cranking up the bitrates
Re: (Score:2)
Yes, let me know when "Thoera" has b-frames. You know, like those things in MPEG-1.
hooray for dirac (Score:1)
Dirac? Schrodinger? wha...? (Score:2, Funny)
Who needs to compress time, when all you hafta do is compress the video.
If I took a video of my cat, and then compressed it with this new codec, would the cat be...
Umm.. never mind...!
.
Yay! More codecs to download! (Score:1)
The porn-viewing experience just gets better and better.
Re:For low values of success (Score:5, Insightful)
Could it be that the BBC's slowness to offer HD is related to the fact that most license payers receive their broadcasts via analogue or "Freeview" digital, neither of which currently support it? I guess they have better things to spend their limited budget on.
Re:For low values of success (Score:5, Insightful)
Re: (Score:3, Interesting)
I strongly disagree that "most of the improvement has come from the change to digital, not HD". TVs don't magically become a higher resolution when you add a digital decoder! The main benefit(?) of digital has been more channels.
I see a huge difference in quality between SD and HD. The most damaging thing for HD that I've seen is that many retailers used to play SD content on HDTVs, which isn't particularly suited for a TFT/LCD screen and can look terrible.
Comment removed (Score:4, Interesting)
Re: (Score:2)
Re: (Score:2)
The main benefit(?) of digital has been more channels.
The main benefit to consumers is less noise in the signal, drastically improving the quality. Resolution isn't the only measure of quality.
Re: (Score:2)
Especially since (in Europe at least) most people don't use wall sized TVs and hence don't care all that much about resolution.
Although maybe 70inch diagonal sets may be common in the US... Here I don't know anybody who would see the point of going beyond 35 or 40.
Re: (Score:2)
Exactly! And "most of the improvement" is more channels! Outside of (a subset of) Slashdot and a few A/V forums, nobody really gives a shit about HD!
Re: (Score:2)
Who on earth modded that insightful? It makes no sense! It's completely wrong. It has nothing to do with the broadcast format, but the recording format. Using 16mm film is FAR, FAR more expensive than HD video.
It also means that other countries
Re: (Score:1)
These days HD content is more likely to be recorded onto a high end digital video format like HDCAM. Many of the cameras are descended from the old DigiBeta camcorders seen on news cameramens' shoulders. There are also cameras such as the Sony F23, Thomson Grassvalley Viper, the Panavision Genesis and the Arri D20 which bear a closer resemblance to those from the world of film. These are mostly used on features, commercials and music videos.
Re: (Score:3, Insightful)
Re: (Score:3, Funny)
And I thought that was some kind of UK New Wave artsy Duane Hopkins/dogma thing they were going for.
Once again, I mistake incompetence for artistic innovation.
Re:For low values of success (Score:5, Insightful)
Re: (Score:2)
Not to mention that Top Gear looks absolutely stunning in HD - check the Polar Special, it is definitive HD demo material.
But the grandparent has a point. Digital terrestial TV is kind-of half baked in most European countries because they went with DVB-T (MPEG 2). (As did my country, Finland). That means that without new boxes the terrestial customers are left with SD until the next "big change". And since the as in the UK and in Finland there are similiar terrestial networks the national TV companies (BBC
top gear = rockin' :) (Score:1)
I've seen only one episode (vsiting a friend who watches TV more than I do, and who gets more channels), but it made me want to see many more. (The one I saw was about zooming and crunching through East Africa in beaters purchased in-country -- pretty impressive how they all held up, actually, though all suffered pretty badly.)
Thanks for your support, TV licensees of Britain!
timothy
Re: (Score:3, Funny)
Thanks for your support, TV licensees of Britain!
Uh, that's OK. Just send me some money, and I'll make sure it gets passed on to the right pub... I mean people.
Re: (Score:2)
Someone's evidentially not been watching Top Gear [topgear.com], which features some of the best camera work on TV and film.
Yes, well, you know, to some of us a show about cars is about as exciting as a show about computers would be to the general public (look ! it's got wheels ! and seats ! whoohoo ! -- yawn)
Re: (Score:3, Insightful)
Re: (Score:1, Insightful)
Windows is not 99% of the world. It is, depending on who you ask, between 88% and 95% of all desktop computer users. Of course, most of those don't give a damn about encoding Dirac files; the proportion of Windows users among highly IT literate, technically minded people is somewhat lower.
Re: (Score:2)
Hmmm... "Killer application" drives adoption of platform.
Most people on Windows only consume, not produce.
Flower chase the sunshine.
Re: (Score:3, Interesting)
I downloaded the code from sourceforge and compiled the code using Visual Studio 2008.
Looks like the encoder is distributed in source format only. I could not locate any pre-built binaries.
I am having trouble figuring out what the command-line parameters mean from the README supplied in the source tarball.
This certainly needs better documentation for non technical users.
The samples certainly look impressive. I will try to compare it against my current favorite encoder -- x264 -- over the weekend.
Re: (Score:3, Interesting)
Took me a while to figure I needed to use YUV input. :(
Unfortunately, it looks like Dirac is no match for x264.
Even VC-1 beats Dirac.
Re: (Score:2)
It's development software; new innovative software tends to be one platform, generally whatever the devs are most comfortable with. I don't know a thing about windows software engineering since early xp, but if I make something cool enough for linux someone else will probably be interested in porting it.
Video processing is commonly done on non-windows platforms because then you don't need a per-computer license for each node in the render farm, and FOSS software can follow an edl as easily as the proprietar