The Static Page

Patent Audio

One of the perennial shadows hovering over MP3 (MPEG 1 Layer III audio) is the patent issue. Fraunhofer, a large German technology company, own one or more patents covering fundamental aspects of how MP3 works. And despite the format's popularity - or perhaps because of it - Fraunhofer staunchly refuse to cede defeat. It's understandable, really. After all, patents are designed to provide the inventor with some sort of means to recover his cost in creating it. This doesn't go down well with a lot of the enthusiasts.

While MP3 has a lot of momentum behind it, it is not the last word in lossy audio compression. Other formats do exist and do compress better and with better fideilty, more flexible channel coupling etc etc. And, naturally, patented. The music industry charge for their newer audio compression, per song if they can get away with it. And then they wonder why thay can't seem to unseat MP3.

At the moment, Fraunhofer have stopped short of requiring licenses for all MP3 activity. Many would argue that it is too late, anyway - the horse has bolted, as the saying goes. It would require some intense and probably financially unviable activity to police all the MP3 trading, but they could probably lay into the decoder and encoder market again. LAME think they have a development model that should preclude them from being successfully prosecuted by Fraunhofer. The LAME encoder was originally patches to the ISO sample code. Or perhaps the answer is to move away from MPEG 1 Layer III entirely...

An enterprising set of coders began a project many months ago called Ogg Vorbis. This is a lossy encoder (like MP3) but developed using ideas that aren't patented. The code is deliberately GPLed and the ideas public domain. The idea is to have a totally free, gratis audio codec that performs better than MP3. At the moment, Vorbis is nearing a 1.0 release. It generally sounds pretty good and plug-ins are available for some of the most popular media players, allowing near seamless transition to .ogg files.

As some of you will be able to guess, I've been experimenting with Ogg Vorbis over the last few days. In short, I'm impressed. I haven't found any audio artifacts and it makes smaller files than LAME in MP3 mode. Unfortunately, I seem to have hit a bug cycle in XMMS, and needed to recompile it from the latest source for it to successfully load all my ogg files. (Well, most of them. There are still a few it doesn't like.)

The prime disadvantage of Ogg Vorbis files at the moment is that MP3 files and support are currently everywhere. This will be hard to change. Re-encoding tools are a start, but the current wave of new portable audio players are MP3 devices. And probably licensed properly from Fraunhofer. But since Ogg is free, there is very little stopping a hardware maker from supporting Ogg Vorbis instead. Their real selling-point will be smaller files, and hence more songs per solid-state storage.

Note that Vorbis doesn't solve content distribution. It would be still be copyrighted music, even in an .ogg file. But it would take the industry patent licensing out of the picture. I wonder if a free codec with better performance than MP3 would scare Fraunhofer at all? What do you think? Would you be willing to re-encode your MP3 collection as Ogg Vorbis?

Wade Bowmer, aka Static

Comments? Email me at static dash page at yceran dot org.