Last month I posted about WebM, the free-for-reuse video codec that has been released by Google. In that post, I mentioned the reservations that Apple and others have about patent status of the codec. There are various groups of patents held by industry players that concern the encoding and decoding of video. To make the process of creating hardware and software that encodes and plays video easier, the MPEG LA was created, a body to administer the collection of royalties for the use of up to date digital video and audio patents. So if you’re writing a program that decodes, say, H.264 video, you could pay a fee to the MPEG LA to cover your use of the patents. In turn the MPEG LA would split the fees between the individual patent owners in the ‘patent pool’.
Now as detailed in a previous post, the issue of video patents is becoming increasingly controversial as the web becomes more and more video-oriented. The new HTML 5 specification includes a ‘video’ tag along the lines of the current ‘image’ tag which allows the embedding of video media direct into web pages, without the need to specify a helper object (like say a Flash video player) to actually do the decoding. This means that web browsers themselves now need to be able to handle the decoding and displaying of videos, and that would seemingly mean the payment of money to the MPEG LA (unless you used very old video standards that have the double disadvantage of poorer quality and larger file size).
So the arrival of WebM was seen by many as a blow for the open web, making the creation of open source and free software web browsers and video encoders possible without the levying of patent fees by the MPEG LA. Not unnaturally, the MPEG LA were not too pleased about this prospect, and began making comments about starting a VP8 patent pool almost immediately. On February 10th the MPEG LA went public with their plan to create such a pool.
So it seems that battle lines are being drawn. When asked about the situation Google told The Register that they are in the process of building a coalition of technology players who agree to not assert their IP against the WebM standard. Indeed the VP8/WebM code made available by Google is under a permissive (BSD) licence with an additional patent licence that is conditional on the licensee not alleging that VP8/WebM infringes on any of their patents. While the terms of the WebM hardware codec designs are not publicly available, it seems very likely that they too carry this kind of stipulation.
It’s not surprising then that the list of firms who support the use of WebM has no cross-overs with the list of firms that hold patents in H.264, the standard with which WebM competes most squarely. If WebM does implement some of the patents from the H.264 pool, and it’s quite possible that it does, Google should in theory be paying a lot of licence fees into this patent pool, to cover the entirety of the community use around WebM. In practice, if Google can force uptake of WebM rapidly enough, hardware and software producers will need to serve user demand and join the Google non-assert posse, even if they are also H.264 licensors.
This strategy is helped by the fact that WebM is less of a standard and more of a solution. Formal video standards like H.264 are described in official published documentation, and implementers are allowed to write the encoders and decoders according to their own ideas of what is most efficient. WebM does not have the same kind of official standardised documentation that H.264 has, which makes it much harder to implement without just using the code Google makes available, with its accompanying non-assert patent commitments.