MPEG LA, the one-stop shop for motion video patent licenses, yesterday announced a call for patents essential to the VP8 video compression algorithm — the algorithm that is fundamental to Google’s WebM video format. MPEG LA is asking organizations that hold patents believed to cover integral, unavoidable parts of the VP8 algorithm to come forward and submit those patents to the licensing company. The patents will in turn by analyzed by MPEG LA, and those deemed to be relevant will be pooled together. The pooled patents will then be available to license as a single convenient bundle.
In its promotion of WebM and VP8, Google has insisted that all the relevant patents were developed by codec company On2, which Google purchased last year. The patents can be licensed from Google without payment of any royalties or any restrictions on usage. Google has been heavily promoting WebM for use with the HTML5
<video> tag, which allows plugin-free video to be embedded in webpages, and the royalty freedom is a key part of WebM’s value proposition.
Competitive codecs such as the open and industry standard H.264 require royalties to be paid by software and hardware developers. Companies like Opera and Mozilla, as well as the W3C group that is developing the HTML5 specification, deem these royalties be an unacceptable impediment to their usage. They have no such qualms about the royalty-free WebM.
If MPEG LA is successful in assembling a patent pool, that royalty freedom could come to an end. The company is soliciting patent submissions until March 18th. Once the submissions have been made, it will determine which patents are essential to VP8; only those patents that are unavoidable can form part of the patent pool. The owners of those selected patents will then decide on the license conditions they wish to impose, and these conditions could include royalty payments.
Whether this will happen, of course, is the big question. MPEG LA might fail to form a patent pool altogether: it may receive no relevant patent submissions, in which case the patent pool process will likely end. Such an outcome still won’t mean that WebM is in the clear — a company may feel that it’s more lucrative to avoid a patent pool and allow WebM usage to become more widespread before asserting claims — but it would probably imply that there aren’t dozens of potential claimants just waiting to come forward.
This sort of outcome might well see Microsoft’s current neutral stance towards WebM (it will work in Internet Explorer 9, just as long as a suitable third-party codec is installed) become more overtly positive. Redmond might start shipping a WebM codec of its own, for example.
If MPEG LA does form a patent pool, the license terms will be critical. MPEG LA exists to monetize patents, however, so it’s unlikely that any patent pool would permit the kind of indiscriminate royalty-free license that Google currently offers. More likely, they would choose terms similar in kind to those of H.264; Web video may be free, but decoders still incur a royalty. This would put WebM implementors in a difficult position — either drop WebM support, pay up, or risk going to court to fight a patent infringement suit.
An infringement suit is an unappealing prospect: even if you win, the drain on your financial resources can mean that ultimately, you lose. This is especially problematic for organizations like Mozilla, since Google offers no indemnification for users of WebM — if Mozilla gets sued, Google won’t step in to help. As such, the safest, most conservative option for Opera and Mozilla would be to drop support. Google has deeper pockets and can better sustain a legal attack, but even there, the company has to weigh its options carefully. A lost court case could cost tens of millions of dollars. Paying up just to avoid the problem may very well be the better option.
But paying up is problematic too. VP8 is, for most purposes, inferior in quality to H.264. H.264 is much more widespread in software tools, hardware accelerators, and so on: it’s enormously widespread already. If VP8 loses its key feature — royalty freedom — implementers may very well decide that, since they have to pay anyway, they’d be better off paying for the superior, more widely used H.264 license, and abandoning WebM entirely.
Whatever happens — and it will probably be many months before we find out — this is bad news for WebM. The formation of a patent pool directly undermines Google’s claims about the codec — and yet, even if MPEG LA fails to create a pool, question marks surrounding the codec will remain.