No, you can’t do that with H.264

A lot of commercial software comes with H.264 encoders and decoders, and some computers arrive with this software preinstalled. This leads a lot of people to believe that they can legally view and create H.264 videos for whatever purpose they like. Unfortunately for them, it ain’t so.

Maybe the best example comes from the Final Cut Pro license:

To the extent that the Apple Software contains AVC encoding and/or decoding functionality, commercial use of H.264/AVC requires additional licensing and the following provision applies: THE AVC FUNCTIONALITY IN THIS PRODUCT IS LICENSED HEREIN ONLY FOR THE PERSONAL AND NON-COMMERCIAL USE OF A CONSUMER TO (i) ENCODE VIDEO IN COMPLIANCE WITH THE AVC STANDARD (“AVC VIDEO”) AND/OR (ii) DECODE AVC VIDEO THAT WAS ENCODED BY A CONSUMER ENGAGED IN A PERSONAL AND NON-COMMERCIAL ACTIVITY AND/OR AVC VIDEO THAT WAS OBTAINED FROM A VIDEO PROVIDER LICENSED TO PROVIDE AVC VIDEO. INFORMATION REGARDING OTHER USES AND LICENSES MAY BE OBTAINED FROM MPEG LA L.L.C. SEE HTTP://WWW.MPEGLA.COM.

The text could hardly be clearer: you do not have a license for commercial use of H.264. Call it “Final Cut Pro Hobbyist”. Do you post videos on your website that has Google Adwords? Do you edit video on a consulting basis? Do you want to include a video in a package sent to your customers? Do your clients send you video clips as part of your business? Then you’re using the encoder or decoder for commercial purposes, in violation of the license.

Now, you might think “but I’m sticking with MPEG-4, or MPEG-2, so it’s not a problem for me”. No. It’s just as bad. Here’s the relevant section of the license:

13. MPEG-2 Notice. To the extent that the Apple Software contains MPEG-2 functionality, the following provision applies: ANY USE OF THIS PRODUCT OTHER THAN CONSUMER PERSONAL USE IN ANY MANNER THAT COMPLIES WITH THE MPEG-2 STANDARD FOR ENCODING VIDEO INFORMATION FOR PACKAGED MEDIA IS EXPRESSLY PROHIBITED WITHOUT A LICENSE UNDER APPLICABLE PATENTS IN THE MPEG-2 PATENT PORTFOLIO, WHICH LICENSE IS AVAILABLE FROM MPEG LA, L.L.C., 250 STEELE STREET, SUITE 300, DENVER, COLORADO 80206.
14. MPEG-4 Notice. This product is licensed under the MPEG-4 Systems Patent Portfolio License for encoding in compliance with the MPEG-4 Systems Standard, except that an additional license and payment of royalties are necessary for encoding in connection with (i) data stored or replicated in physical media which is paid for on a title by title basis and/or (ii) data which is paid for on a title by title basis and is transmitted to an end user for permanent storage and/or use. Such additional license may be obtained from MPEG LA, LLC. See http://www.mpegla.com for additional details. This product is licensed under the MPEG-4 Visual Patent Portfolio License for the personal and non-commercial use of a consumer for (i) encoding video in compliance with the MPEG-4 Visual Standard (“MPEG-4 Video”) and/or (ii) decoding MPEG-4 video that was encoded by a consumer engaged in a personal and non-commercial activity and/or was obtained from a video provider licensed by MPEG LA to provide MPEG-4 video. No license is granted or shall be implied for any other use. Additional information including that relating to promotional, internal and commercial uses and licensing may be obtained from MPEG LA, LLC.

Noticing a pattern? You have a license to use their software, provided you don’t make any money, your friends are also all correctly licensed, and you only produce content that complies with the MPEG standard. Using video for a commercial purpose? Producing video that isn’t within MPEG’s parameters? Have friends who use unlicensed encoders like x264, ffmpeg, or xvid? Too bad.

This last thing is actually a particularly interesting point. If you encode a video using one of these (open-source) unlicensed encoders, you’re practicing patents without a license, and you can be sued. But hey, maybe you’re just a scofflaw. After all, it’s not like you’re making trouble for anyone else, right? Wrong. If you send a video to a friend who uses a licensed decoder, and they watch it, you’ve caused them to violate their own software license, so they can be sued too.

Oh, and in case you thought this was specific to Apple, here’s the matching piece from the Windows 7 Ultimate License:

18. NOTICE ABOUT THE H.264/AVC VISUAL STANDARD, THE VC-1 VIDEO STANDARD, THE MPEG-4 VISUAL STANDARD AND THE MPEG-2 VIDEO STANDARD. This software includes H.264/AVC, VC-1, MPEG-4 Part 2, and MPEG-2 visual compression technology. MPEG LA, L.L.C. requires this notice:
THIS PRODUCT IS LICENSED UNDER THE AVC, THE VC-1, THE MPEG-4 PART 2 VISUAL, AND THE MPEG-2 VIDEO PATENT PORTFOLIO LICENSES FOR THE PERSONAL AND NON-COMMERCIAL USE OF A CONSUMER TO (i) ENCODE VIDEO IN COMPLIANCE WITH THE ABOVE STANDARDS (“VIDEO STANDARDS”) AND/OR (ii) DECODE AVC, VC-1, MPEG-4 PART 2 AND MPEG-2 VIDEO THAT WAS ENCODED BY A CONSUMER ENGAGED IN A PERSONAL AND NON-COMMERCIAL ACTIVITY OR WAS OBTAINED FROM A VIDEO PROVIDER LICENSED TO PROVIDE SUCH VIDEO. NONE OF THE LICENSES EXTEND TO ANY OTHER PRODUCT REGARDLESS OF WHETHER SUCH PRODUCT IS INCLUDED WITH THIS PRODUCT IN A SINGLE ARTICLE. NO LICENSE IS GRANTED OR SHALL BE IMPLIED FOR ANY OTHER USE. ADDITIONAL INFORMATION MAY BE OBTAINED FROM MPEG LA, L.L.C. SEE WWW.MPEGLA.COM.

Doesn’t seem so Ultimate to me.

My advice: use a codec that doesn’t need a license:

Q. What is the license for Theora?
Theora (and all associated technologies released by the Xiph.org Foundation) is released to the public via a BSD-style license. It is completely free for commercial or noncommercial use. That means that commercial developers may independently write Theora software which is compatible with the specification for no charge and without restrictions of any kind.

84 thoughts on “No, you can’t do that with H.264

  1. @David Gerard
    You mean this one?
    http://web.mit.edu/xiphmont/Public/theora/demo7.html

    Notice the thing about the bug in ffmpeg. So basically even the xiph people are admitting x264 is far ahead even on its default settings. (Let alone the stronger settings. AFAIK, Theora doesn’t even have stronger settings.)

    And here’s what happens when x264 people test stuff:
    http://x264dev.multimedia.cx/?p=102
    http://x264dev.multimedia.cx/wp-content/uploads/2009/08/quality_chart1.png
    Yes, that is slightly older Theora (older post) and it did have trouble on that video, but do notice the “MPEG-2 beats Theora”, which just should never happen on any video.

    But here’s one test encode on the new and awesome Theora to give some direction:
    http://underwater.nyaatorrents.org/bin/comparison/theora/4-h264.png
    http://underwater.nyaatorrents.org/bin/comparison/theora/4-theora.png

    @Original post
    This is a clear case of IANAL.
    First of all Theora most likely does have patented things in it because it’s pretty much impossible to make a video encoder without breaking patents accidentally.

    The only real question is that will you get sued? Yes, it is a lot more likely that MPEG LA would sue for using H.264 than some random company for using Theora.

    But with the same line of thought, MPEG LA is totally fine with open source stuff and the talk about “unlicensed” encoders is total crap. It’s not about the encoder, but what you do with it.

    Using or distributing ANY software is illegal. Period. EVERYTHING is patented. It doesn’t matter if you know about the patents when making the software or not. And again, the only question is: will you get sued? Well, IANAL but using x264 will NOT get you sued because it’s “unlicensed”.

    Besides, even without the royalty free internet video, the license is free until you ship more than 100000 products. (Or similar.)

  2. @David Gerard

    It’s come up about every time it’s been mentioned, but a brief rundown:

    1. The Theora video uses a keyframe interval about 10 times larger than the H.264 version
    2. Google’s homegrown H.264 encoder is pretty terrible to start with, and comparisons against it aren’t relevant for anyone but Google.
    3. Big Buck Bunny requires very little bitrate to reach transparency due to it being CG and the complete lack of noise. At the bitrate tested, most any encoder with any format newer than mpeg-1 would be pretty close to transparency.

  3. That’s why I use YouTube, or vimeo if I really had to. I don’t have to worry about embeding or sending a video (as long as it respects copyright). It’s also why firefox does NOT support HTML 5 demos from either video site due to use of the codec.

  4. “Theora is catching up in terms of quality to H.264, so it is a matter of time before it surpasses it.”

    This slipped my attention first, but it’s a statement that needs some attention because apparently lots of people think that way, but it shows a clear misunderstanding on encoders and coding standards.

    H.264 and Theora are STANDARDS for video compression. They are not improving anywhere. If they did change, all the decoder support would break.

    What CAN improve are the encoders. Say x264 and libtheora for example.

    But the problem is that Theora is technically inferior to H.264. (This is not a biased opinion. It’s simply because Theora is based on somewhat obsolete VP3 and all the good stuff is heavily patented, so they have to avoid them.)

    Making an encoder is like building a house. The compression standard dictates what materials you get. You can make a better or a worse house with the same materials, but at some point you hit a point where improvement starts becoming harder and harder. libtheora might’ve fixed the gaping hole on the roof, but they will definitely not be able to keep up with that speed of improvement.

    So bottom line is this: Theora will NOT catch up with H.264 because neither will really change. And it is very unlikely that libtheora catches up with x264 (which is also under heavy development by very skilled people, by the way) or any other of the better H.264 encoders.

  5. Sigh, why do I keep forgetting these things…

    Just a quick note on testing x264 with default settings. (Like xiph did.) First of all, ffmpeg’s default settings are broken for x264 encoding, so I hope they meant x264′s default settings. (A preset in ffmpeg.)

    And these default settings are not tuned for PSNR. They’re tuned to get the best possible visual quality even at the cost of PSNR drop. (psychovisual optimization) As far as I know, libtheora doesn’t really do that.

    PS. SSIM is a better metric anyway.

  6. The author of this piece is an idiot. The license clearly states that you can’t without an additional license encode or decode into the H.264 format for profit, meaning in your own software or as a service. (i.e. send me your vids and I’ll encode them into H.264 format) In other words, you can buy software with the encoder in it, and encode your own video of your kid, your vacation, wedding, whatever, and post or sell that video for a profit or even post it on your website in which ads revenue is generated, even in the H.264 format. This blogger doesn’t understand what he is reading, and is trying to make news by posting something he is so diluted to think he “discovered.
    Is this author really this stupid I mean I know the license has big words in it and it is in legalese, but they are in English. Did the author bother to fact check with Apple to make sure the license means what he thinks it does Does he really think that all the people who are posting video to video sharing websites have a separate license encode in that mpeg or H.264 format.

    Apple wants people to use the format. They want that because they are selling software that encodes and decodes to and from that format. They sell hardware to display the video as well as software to create, edit, and even post the resultant video output. I could go on, but the blogger just needs to do some basic journalism and call up a rep of Apple and fact check.

    1. Talk about idiots! Here’s one that uses two malapropisms to further confuse the issue! … “diluted”, when he probably means “deluded”, and “… you can buy software with …” when he probably means “can’t”. This guy lives in the world of quantum physics where things both exist and don’t exist, both at the same time. For engineers, this is not useful information.

  7. “As far as I know, libtheora doesn’t really do that”, well consider yourself educated now: The current libtheora development does have PSNR hurting psychovisual optimization; but what libtheora doesn’t offer is a benchmark cheater knob to turn it off— it wouldn’t serve any legitimate purpose, excepting making misleading tests. Good testing is hard. Surprise!

    Bucky says that mpeg-1 would be transparent on big buck bunny at that bitrate— Try it. It’s not. Or, if you like, look at the comparison with YT’s H.263 at a somewhat lower resolution and bitrate… Or Maikmerten‘s 720p YT comparison. As an aside, I think the noiselessness of BBB is something of an additional challenge, since there isn’t as much noise to mask artefacts… but the point was making a order-of-magnitude comparison, not a detailed shoot-out.

    Theora wouldn’t, and shouldn’t, win a detailed shoot-out against x264 operating with a much more cpu and memory hungry profile than Theora. It just wouldn’t make sense for it to. Would get it into an order-of-magnitude range where many people would rather pick their codec based on other factors? I would expect it to, it certainly does against lesser H.264 encodes at reasonable bitrates.

    It’s true that higher bitrates tend to make all codecs look more similar, and that point is also part of what that comparison is trying to show. It’s also the case that you can make just about any codec look bad compared to a slightly better codec by picking rates that straddle the knee in the rate-distortion curve. Personally I think that’s an exaggeration of the differences, you might think that picking a rate *away* from the knee understates the differences. In the comparisons against youtube we didn’t pick the rates, however, youtube did. I think thats a pretty objective way of settling the near-the-knee vs away from the knee testing questoin. ::shrugs::

  8. @Greg
    Ok, my bad, but what?? So you mean you encoded it with psy opts on for both encoders?!?

    There goes hope of any validity for the test. There is really no way of telling accurately which encoder has stronger psy opts. (PSNR hit vs. perceived quality boost) I thought you discussed this with Dark_Shikari?

    And I don’t mean even in the sense of “Well, it’s just a quick check for order-of-magnitude”, I mean totally wrong. People will link to it and use it as reference.

    Calling turning psy-opts off “cheating” is just stupid. Sure, metric-based testing is always a lot worse way to do it than subjective testing, but at least give the encoders a level playing ground! (Although in this case a fair test can’t even be made because of the missing toggle?)

    Think about two encoders… A has strong psy opts and B has weak/no psy opts. Because of this A has better visual quality, but with psy opts on, B will beat it PSNR. (Without psy A will win) By definition there’s no way of testing the quality of the psy opts of the encoders, so any metric-based test should focus on the non-psy part. Sure, it won’t and can’t tell you the whole picture, but again, that’s impossible in the first place, and you should just make the objective testing as fair as possible and leave the final word for subjective testing.

    Or how about an analogy? Car A and car B are getting reviewed by a car magazine. Car A comes with options: air con, electric windows, multiple airbags, etc. Car B has no such options. (Or maybe just air con, or something.) But the magazine in this case is interested in only the speed. And even though the options make car A a lot more comfortable to drive and thus people who buy it will be happier with it, because of the added weight, car B beats A in a drag race. So the should the magazine say “If you want a fast car, buy B” WRONG! What they should say is “If you want a good car, buy A with options, but if you really only want a fast car, buy A without options.”

  9. Look this isn’t a issue with the technical quality of Theora we could switch to dirac. This is an issue with licensing and how you can legally use FILES produced by various decoders and how you can distribute them.

    The issue here is freedom and h264 just doesn’t have it.

    Also don’t submarine patent me, h264 is just a patent pool, someone else could be sitting with h264 patents ready to surface when mpeg-la starts raking it in.

    Even worse, both standards are based on ancient technology. At least dirac is wavelet based. Haven’t we DCT’d the world to death enough by now. There are other techniques.

  10. I think we’re confusing things. There is absolutely nothing wrong with psycho-visual optimizations in a purely subjective comparison, which is what the order-of-magnitude YT comparison I was talking about that people frequently refer to.

    (the older PSNR graphs that I was involved with were prior to the ‘psy-opt’ in x264 being on by default, as I recall)

    I think you’re taking a rather myopic view on this: There is no “non-psy” part of these modern lossy video formats, their entire structure is intended to produce output which looks good. If the goal was to maximize PSNR, the design might well be rather different. (say… a wavelet codec ;) )

    There are many things you can do to boost encoder performance for particular test cases or for particular measurement metrics, but doing those things makes the comparisons even less related to the real-world results, gives an unreasonable gain to whatever codec has over-optimized for that particular case, creates traps for the users (more switches to set wrong), and more code to maintain. x264 has a “touhou” preset, micro-tunings for a single video game frequently used in some circles for codec comparisons. I’m waiting for the “german BSDM porn” preset next, It’s a bit ridiculous. It’s none of my business what knobs the x264 developers waste time maintaining, but it doesn’t seem like a good use of resources to me.

    I think you got the car analogy all wrong. Here is the right one: The magazine wants to test speed, but the nearby speed limits are all 75mph (subjective tests are ‘costly’ to perform), so they drop the cars in neutral and measure the readline rpm then multiply by the gearing ratio in the highest gear. You’ve measured something. It’s probably something useful. Speed it is not.

    The objective tests are a lot more useful within a codec than cross codecs, where they can still be a useful basic yardstick. Which is all I’d ever expect to use them for in that context. If you’re niggling too much about different optimizations across codecs, you probably have an unreasonable expectation of how much these tests can tell you.

  11. Dark_Shikari is quite possibly the only person who has ever used Touhou for codec comparisons. The tuning option is there as a joke about how obsessed with encoding Touhou videos he is.

  12. This is just silly.

    The license states that you cannot use the encoder/decoder for commercial purposes. That means you cannot open “Ben’s Emporium of Encoding” and accept payment to encode videos in the H.264 format and send them back to people. Or, you could not use the encoding library provide with your copy of the software to then create your own video encoding software. You cannot sell access to your personally licensed encoder.

    However, if you sell a video that you encoded with Final Cut Pro, you have not violated any of the terms that you quote. I don’t know if there are other stipulations, but the ones you highlighted don’t even remotely express what you are claiming.

  13. Of course everything is structured to make visual quality better, but almost all features that do that also raise PSNR and SSIM. (And are tested using those metrics.) Mainly because small feature tweaking is usually impossible to see. (But of course the small improvements of different features add up to a real visual benefit.)

    But I just realized how accurate the car speed analogy was. You’re saying it’s ok for the magazine to compare the speed of two cars and say car B is faster even though A would be faster without the options. And like PSNR tests, speed tests have really no actual benefit for normal people, only in some special circumstances. (Racing?) But the main problem here is: people are stupid and will buy cars based on the maximum speed because it’s cool and easy to measure, even though what they really want is comfortable ride. So the magazine is misleading some people to buy B by saying “car B is faster” even though with similar options it isn’t. And car A is a better car anyway.

    If you followed me through that, what I mean that we both know objective testing is always flawed, but the problem is that is often the only thing people check and make real decisions on that information because they don’t know the problems. (The flawed test got slashdotted, which should tell you something.)

    Not to mention stuff like this:
    http://www.osnews.com/story/22812/MPEG-LA_Further_Solidifies_Theora_as_the_Only_Video_Tag_Choice
    “I’m still not entirely sure about which one is better from a pure quality standpoint (a lot of contradicting reports on that one, and I’m not skilled enough to perform my own tests)”
    This is exactly the kind of statements that start appearing if you’re not careful with testing. People WILL put too much emphasis on them.

    Now, it’s perfectly possible that it was x264 which benefited from the psy opts, but that doesn’t make my main point any less valid: when testing PSNR or SSIM, psy shold be OFF. So basically I’m defending you here? (Meaning if what you say is true, theora should have higher PSNR.)

    So let’s try once more. Encoder A is competing with B. A has better subjective and objective quality. New version of A comes out with psy opts on by default. Subjective quality, which is the only thing that really matters, goes up, but objective goes down. Comparison gets made, B beats A in SSIM and people start using B because of that comparison. A feature that makes better quality should NOT have a negative effect on a comparison that people will look at. You can’t POSSIBLY claim that’s fair.

  14. Oh yea, forgot about the –tune touhou thing.
    Yes, I have NEVER seen anyone else use Touhou videos as a comparison. I was there when they discussed adding the option and they were giggling like little girls.

  15. It’s old now, but I was pretty tired of being bludgeoned with this, back when it was new. I’m glad to know that other people think the preset is as silly as I do! :)

  16. Oh saintdev used touhou too. He didn’t use the tuning option tho. (Or set the same settings manually.) x264 just happens to do very well on the areas that are needed for something like that Touhou video. (Same things are needed for other videos too, of course! For example the static right side of the video.)

    But in any case, the point is silly. The touhou tune doesn’t even enable anything you couldn’t do from command line, and tuning from command line is something you could (should?) do to get maximum quality anyway.

    Of course it’s unfair to tweak some encoder but not the other, but tweakability is a good thing and should be rewarded. I guess the fairest way would be to test on multiple settings. For example default, hq-preset, manually tweaked.

    It’s a same problem when comparing programming language speeds: some people are better at some languages so making a fair comparison is really hard.

  17. Aside of licenses concerns and visual comparison taken from various sites, have you ever tried to encode a video with x264 and libtheora (ffmpeg) by your own?

    Well, I’ve done it and, even if I poke with theora settings (and, believe me, I had tried every combination available), the result I obtained is far beyond x264… What is really messy is that Theora is even far beyond XviD! Transitions between key frames and intra frames seem, sometimes, like Divx 3 Low motion and the quality of frames is lower than Mpeg4 codecs!

    In my opinion, Theora is not so mature to challenge other codecs yet, so it will not spread into the Net. And when maturity will be really reached, probably it will too late to compete with H264, even if the licensing problem will persist (but will be also ignored by the masses!).

  18. @Franz Hilberger

    Those patents you mention are highly questionable, for each of them there are large bodies of prior art, and it seems the holders are aware of this, as there have been many commercial applications and no-one has been taken to court yet, apparently because they know if they take someone to court then their patents will be nullified.

  19. For the people saying that Ben’s legal description is wrong (@Kuukunen, @VC, @Mac), I’m not so sure.

    I suspect Ben’s being a bit melodramatic, but it seems clear (to me at least) that the MPEG-LA has a distinction between encoder/decoder licenses and provider licenses (that includes service providers, but also content providers)

    See: http://www.mpegla.com/main/programs/avc/Documents/AVC_TermsSummary.pdf

    Note that they explicitly indicate that there is a class of license for free-to-air content providers, and that includes internet broadcast.
    That is separate to the encoder license that Final Cut Pro provides.

    If you are a TV station, you can’t simply buy a copy of Final Cut Pro, encode your videos in H.264 and then broadcast them. You need to purchase a broadcasting license for that.

    The same applies to people who sell movies, and it applies to people who broadcast over the internet.

    It seems to me that the MPEG-LA is of the opinion that Google + Vimeo (for example) need to have a licensed encoder to produce their videos, and a provider (participant) license to broadcast them. Specifically it has the words: “content … provider sublicenses … where remuneration is from other sources … AVC video that is delivered via the Worldwide Internet to an end user”
    Currently that license (internet broadcast) is royalty free – “there will be no royalty during the first term of the License”, but that doesn’t make Ben’s argument bogus – it just means the MPEG-LA has made a business decision to get traction on the internet before they start charging fees.

  20. hello,

    sorry to nag but why is it that the phrasing ‘…LICENSED HEREIN ONLY FOR THE PERSONAL AND NON-COMMERCIAL USE OF A CONSUMER TO (i) ENCODE VIDEO …’ does not rule out profit-making from the sale of self-made videos encoded with this software ? i see that the commercial activity isn’t the encoding process itself. however, it would be a commercial use of something (the video) that only exists grace to employing the encoding process using the software in question. am i totally off here ?

    thanks for your patience.

    regards, carsten

  21. @Tim Vernum
    But the thing I was complaining about was this:
    “This last thing is actually a particularly interesting point. If you encode a video using one of these (open-source) unlicensed encoders, you’re practicing patents without a license, and you can be sued.”

    Then why does Facebook use x264?

  22. Facebook has acquired a license (see 200.) and is presumably paying $5million a year for the privileged of not being sued. I don’t think it’s unreasonable for Tim to have made the assumption that the “you” he was speaking to isn’t a large corporation. ;)

  23. As Greg points out, Kuukunen, large video sites are already paying the blackmail, err, I mean, licensing fee for this BOGUS “open standard”. Most people believe an open standard is one free of licenses, patents, and proprietary components. Saying it’s “open” to look at is complete crap if you can’t use it. Under that definition, I could go to the USPTO website and look at any patent, but I’m unable to use them. Does that make them all open? No way!

    Listen, it’s proprietary. That’s fine, so long as they label it as such. But this Flash vs h264 being some soft of Closed vs Open is complete bunk. It’s proprietary vs proprietary. It’s Adobe vs Apple. And quite frankly, once this becomes evident, Apple’s going to get slapped with anti-competitive practices for not allowing Flash on their iPhone and iPad. Adobe says they have it all ready to install, but Apple is blocking it to force companies to use their encoder, or face being empty blue boxed on Apple devices.

    So, please, everyone, save the sanctimonious “h264 is open, down with proprietary flv” crap for Mac Fanboy conventions.

  24. This is obnoxious. You report that the bogey man is coming and you have no understanding with licensing. I guess that’s OK in this age.

    The license agreement you are reading largely has to do with how APPLE has licensed H.264 from MPEG LA. That licensing is for those who implement codecs. Apple’s license does not cover the end user’s use of the product (that’s to say, just because Apple makes Final Cut does not mean that NBC can use H.264 for free because they used Final Cut).

    The AVC licensing terms are spelled out here:

    http://www.mpegla.com/main/programs/avc/Documents/AVC_TermsSummary.pdf

    Not this part for any Internet video you stream:

    “In the case of Internet broadcast (AVC video that is delivered via the Worldwide Internet to an end user for which the End User does not pay remuneration for the right to receive or view, i.e., neither title-by-title nor subscription), there will be no royalty during the first term of the License (ending December 31, 2010), and after the first term the royalty shall be no more than the economic equivalent of royalties payable during the same time for free television.”

    You may have missed it in all “The Sky is Falling” diatribe, but this was recently extended until 2015 or thereabouts. Thus, there are no fees for encoding Internet content for the next 5 years.

    Still, the desire to paint H.264 and its licensing as a bogey man in order to push the Theora agenda is something you can try. If the codec weren’t crap, and if any devices on earth outside of a computer could actually PLAY the content, it might be more interesting. As is, H.264 already has a colossal lead and the consumer is expecting H.264 as that’s all that works on mobile devices.

  25. This is the big monster in the closet that never actually appears…

    Also, Theora may have patent liabilities. Just because no one has sued in the last 6 years of Theora’s obscurity doesn’t mean that someone can’t, or won’t. And, Theora is devastatingly lacking in quality and speed compared to H.264.

    http://x264dev.multimedia.cx/?cat=54

  26. Yeah.
    The original poster is right.

    There are different licenses for different purposes with H.264. Also MPEG-2 and MPEG-4 are all licensed under the same terms.

    If your publishing for the web vs publishing for TV vs publishing for sale in retail… these are all have their different licenses. And end users need all their own licenses.

    You cannot use the patent license provided with Final Cut Pro because the developers of Final Cut Pro did not want to pay your broadcast license.

    *DUH*.

    ———————————–

    And as far as the ‘Patent FUD’ levelled at Theora:

    Your just as likely to violate a patent using H.264 as you use Theora. Popularity is NOT a defence. ‘Doctrine of Laches’ is not going to give you a advantage…

    Theora is OLDER then H.264 by 3-4 years and H.264 has had various revisions/new profiles developed with the latest wide-spread one released in 2007. Theora may not be the most popular codec out there, but it’s (and VP3: which it is based) has been widely distributed and used for almost a decade now.

    When you pay licensing fees to MPEG-LA you only paying money to protect yourself against the patent holders that have contributed their patents to the MPEG-LA patent pool.

    _THAT_IS_IT_.

    Your just as 100% liable for any unknown patents if your using H.264 as if your using Theora.

    In fact, if you think about it, what is the likelihood of a unkown patent affecting Theora and NOT affecting H.264?

    Imagine somebody going to court and telling a judge:

    “Oh I can’t sue those billion-dollar companies using H.264 for violations because it’s used in the iPhone and, gosh darn, I am just too late. All that huge amount of money they have is just something I’ll have to give up on. But all these Theora users that are not engaged in any sort of commercial exploitation of their users… Those are the guys I want to sue!!”

    I am sure that going after Youtube.com, Google, MPEG-LA, and Apple would not be worth it for these folks and going after Xiph for the $10K they have saved up for website stuff and conventions is just going to be such a more jucy target.

    Avoiding the MPEG-LA patents are relative easy for Theora folks because MPEG-LA has to publish everything publicly in detail about their patents. It’s the _unknown_ patents that are the threat and that threat is just as real for H.264.

    Possibly more because it’s a younger, more advanced codec. And the amount of software patents out there have grown exponentionally since the original VP3 stuff was open sourced.

    ——————————

    The advantage of Theora will always remain:

    You can use that encoding for whatever you want whenever you want and however you want. You can open source it, you can close source it. You can include it in your product. You can use it to publish your stuff.

    And _REMEMBER_… The difference between H.264 and Theora is that H.264 can compress videos further and retain the same quality. You can save money on disk space and bandwidth by using H.264, but Theora is capable of same quality and you pay the price of larger file sizes and fatter streams.

    So if your somebody like Facebook and the idea of paying X hundreds of thousands of dollars on a license for H.264 will easily be offset by the cost of bandwidth…. then the chances of Theora displacing H.264 is about nill.

    But if you want to do something like freely distribute and use software for streaming TV from your house to your laptop at work or creating your own websites that include streaming videos then Theora is going to be the way you want to go if you want to remain legal.

    Also do not forget that Theora keeps H.264 prices low. I seriously doubt that the MPEG-LA folks would have extended their temporary low-prices a few more years if Theora was not around.

    It is the _stated_goal_ that the current prices are there to get people using it widely and later on they will start to rise the prices based on what they perceive as a the value of H.264 to the companies and users that will be locked into using it.

    And remember… Don’t beleive anything Apple has to say on the subject. They own patents on AAC and H.264 and the more people that depend on those technologies the more money they have to make from it.

    ————————————-

    And, of course, the MPEG-LA tolerate open source software like x264.

    x264 and ffmpeg folks produce _EXCELLENT_ top-notch software. Much better then the majority of proprietary encoders/decoders out there. We are completely fortunate that they exist.

    But the MPEG-LA folks will still sue you if you don’t pay for the patent license if you use them commercially. If your too small for them to give a shit then you’ll probably wont’ get targetted.

    Google, for example, uses parts of ffmpeg bundled with their Chrome browser. They pay the patent fees for using it, too.

  27. What a crap. So after I pay some small $999 fee I’m not allowed to make moneys. Kinda interesting idea. Probably I’m expected to buy $999 software just to encode family movies for myself? What a moron, we clearly need a better licensed codec without idiotic licensing restrictions. Currently licensing terms appears to be a BS.

  28. Stephen Shankland got a response from Allen Harkness, MPEG LA’s director of global licensing, in regards to this. http://news.cnet.com/8301-30685_3-20000101-264.html

    It’s OK for a professional to use the software to make a living. It’s not OK for big corporations without an explicit license agreement.

    The article includes comments from Dan Homiller, a patent attorney with intellectual property firm Coats and Bennett. “The purpose of the provision in the MPEG LA license is to ensure that the license doesn’t cover commercial distribution of H.264-encoded video,” Homiller said. “It would be nice if there were a ‘gentler’ way to convey this, but it might be challenging to do so without opening up some loopholes that the licensers would regret.”

  29. Can someone help me out. I’d like to create a demo “reel” of a kids’ program, but I am unsure if I can us a video enabled DSLR due to the encoding issue.

    From my understanding since I’m not a professional and I’m not making money off it (yet) I can legally send the demo to prospective clients/upload for others to see.

    Am I correct with this train of thought? If there are legal issues, then how can I record a “demo reel” without breaking any laws?

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>