MPEG LA – Licenses and fees


So, because of a request i got in my job, i had to contact the famous MPEG LA.

That is the company that owns the patent for mpeg2, mpeg4, h264 codecs and for some other codecs. In one word for MPEG video codecs.

Basically if you deal with videos there is a chance that your project, either web application, stand alone software or video hardware is using a codec that has been copyrighted by them.

Their patents cover four categories and you can find infos on their website www.mpegla.com .
That’s where you can find also a contact page with a phone number. First of all, they list numbers for several US areas but the same person answers for each of them, they require you to disclose at least your first name before they can handle the call and transfer you to the proper department. I bet the department is pretty much one big incoming call basket for companies using their codecs without even knowing that, i think big manufacturers (windows media player for microsoft, video players in general, etc.) have some kind of special channel, plus of course some other offices.

My situation was easy, i called on behalf of someone that was working on a project that uses their h264 and mpeg2 codecs. I explained to them that the patent has been initially ignored but in good faith and that there was the willing to disclose the books for the past years in case they needed that. On the other hand we were talking more about a problem of quality of the violation rather than quantity.

What i understood is the following:

  • with the exception of mpeg2, all the other licenses they can give you, will trigger royalties only if the commercial distribution goes over the 100,000 pieces per year. For mpeg2 there is no threshold and you must pay for each unit supporting that codec.
  • You are supposed to pay reotractively, $2.50 per unit that utilizes the codec and that you distributed commercially before you signed the agreements that will give you access to the licenses
  • you have to pay a royalty of $2.00 for each unit that you will distribute after the signature
  • a discount on the past royalty is possible, can be discussed, but there is no guarantee until they know who you are, where you live and how much you own them
  • they rely on the licensed business to report royalties

My personal comment is that $2.00 is a lot and is a royalty very likely established with hardware in mind. They have a lower fee for video streaming, few cents or so, but in terms of commercial unit, either software or hardware, you are supposed to take out $2.00 out of your sales per each unit. It’s a lot.

Beside that, i have been constantly thinking of what i heard in the past about mpegla, their legal fire power, their capacity to run after businesses that are not licensed, hunting them till they pay. Sorry guys, your reputation might even make you feel proud but will not help your business.
Surely people call in also because of such a reputation, but that does not help business relationships.

Personal impression: for the contact itself is that they sound reasonable, flexible, open to discuss, but that’s how they sound, not how they look like when they ask you 5 times in a row in 2 days for your business mailing address… But getting to the point is not that hard, if you know what you are doing and what you need it is simple to get licensed.

I will add some more info in the next days, can’t disclose communication for a mere business fairness but can surely publish common interest information.



This entry was posted in ecommerce, web. Bookmark the permalink.

4 Responses to MPEG LA – Licenses and fees

  1. Lee Carré says:

    … a codec that has been copyrighted by them.

    Unless you’re using an implementation they had authored (in which case copyright would apply), then I suspect you actually meant “patented”. The difference is between the idea/technique, and the specific implementation, which is significant. The difference is that of the Shakespeare’s text of Romeo and Juliet, and the idea of a love story (perhaps with specific themes). This is why (software) patents are more harmful (to everyone) than proprietary copyright.

    More generally, this is why the (propaganda) term “intellectual property” is so misleading/confusing (and designed to be so: gnu.org/philosophy/not-ipr.html)
    There are other terms worth avoiding, for similar reasons.

    … owns the patent for …

    Actually, it’s patents. There are many of them, often overlapping (which patents aren’t supposed to).
    Sadly there is no longer the concept of a single patent on an ‘invention’, particularly when patents are permitted on abstractions, such as software (which is really just maths and apparently unpatentable).
    “Own” is yet another term to avoid; it implies some kind of natural right (as with real physical property; hence the confusion “intellectual property” brings). A patent is really just a government-granted monopoly for a fixed/limited term.

    Sorry guys, your reputation might even make you feel proud but will not help your business.
    Surely people call in also because of such a reputation, but that does not help business relationships.

    That’s the thing; they’ve already been granted a monopoly on the encoding technique, so they hardly need to play nice to be accepted; they’re the sole licensor and have all the power.

    • technonsense says:

      I remember i read somewhere that they are artificially extending the duration of the patents, something that you undirectly confirmed when you said that “A patent is really just a government-granted monopoly for a fixed/limited term.”
      Do you know if and how they are extending that term?

      • RxLex1 says:

        You can go to the USPTO . gov website and check on all their patents.

      • Lee Carré says:

        I remember i read somewhere that they are artificially extending the duration of the patents, something that you undirectly confirmed when you said that “A patent is really just a government-granted monopoly for a fixed/limited term.”
        Do you know if and how they are extending that term?

        I’m glad that you found my previous comment(s) useful ☺.
        The kind of people who hold these patents will always want them to last as long as possible, and, thus, will seek some way to make that happen. So, I’d expect that’s exactly what they’re trying to do; even if it’s difficult and not supposed to be possible. They have a vested interest in maintaining their monopoly, after all.
        One, simple, way, is to re‐apply for a new patent, so that the next continues from the end of the previous one; effectively an extension. Patent re‐application can’t be exactly the same as a previous patent (a simplistic attempt to prevent applicants from effectively obtaining a perpetual patent, by continually re‐applying); however, changing some trivial detail, even slightly, usually gets around this requirement (it’s no longer exactly the same, after all). This is what many (American) universities do with patents they hold.
        Another ‘technique’ I’m aware of, is causing the processing of a patent application to be delayed (bogged‐down with bureaucracy/negotiation, etc.); because a patent’s duration starts from when the application is (first) submitted, not (stupidly) when it is (first) published (as with copyright). This results in ‘submarine’ patents (appearing by surprise), which are particularly disastrous with then‐established technologies (determining if what you want to do is covered by patents is rather difficult with abstract concepts which can be described many different ways; this is why many (software) patents overlap, which patents aren’t supposed to). Consider what happened with MP3, in that the patent holders initially adopted the strategy of encouraging implementation and use of the encoding technique, to then ‘bait and switch’ by enforcing ultra‐proprietary licensing requirements in order to leverage maximum profit from licensees; leaving many with the choice between not supporting the (then) de facto format, or paying the considerable fees to the patent holder(s) for permission to support it in their software/device. Consider, also, that even if a hopeful licensee has the demanded fees, the patent holder has no obligation to grant a license, as is usually the problem for Free Software, such as the LAME encoder (which has been pushed underground, just like DeCSS).
        Yet another way would be to stagger the applications on related techniques, so that the total time from the start of the first patent, to the end of the last, is longer than the intended (currently 20 year) duration. This, I believe, plays on the fact that it’s easy to make patents on abstractions overlap, and that software (especially) requires implementing many techniques in order to do most anything useful. One simple example would be the algorithm to encode a format as at least one patent (although, could also be many, covering different parts of it), and the algorithm to decode the same format being covered by at least one different patent (but, again, could just as easily be many patents for different parts).
        However, I’m far from any kind of expert on the matter. Most of what I know is from what I’ve read, myself. For truly unbiased background information, I suggest: essays and speeches (http://www.gnu.org/philosophy/essays-and-articles.htmlhttp://www.gnu.org/philosophy/speeches-and-interview.html) by either of Richard Stallman and/or Eben Moglen (General Counsel) for the FSF); Math You Can’t Use by Ben Klemens (though, be wary of his careless use of terms to avoid – http://www.gnu.org/philosophy/words-to-avoid.htm); Patent Failure by James Bessen and Michael J. Meurer; and possibly also The Future of Ideas by Lawrence Lessig.
        There are, also, various ‘patent watch’ sites on the Web, which may be to your liking for remaining apprised of recent happenings.
        I’m sure that’s plenty more than you’d, ordinarily, care to know about ‘modern’ patent absurdity.

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>