MP3 and Patents (editorial)

Maciej Stachowiak has written an editorial on MP3s, and the major problem with the format. Now I'm not a huge MP3 freak (althought I have d/l 'em, play 'em, and bought many CDs because of 'em :) but I know a lot of you folks are, so read what Maciej has to say and let's hear your voices on the issue.

The following is an editorial by Slashdot Reader Maciej Stachowiak

MPEG Layer 3 and Software Patents

MPEG Layer 3 format audio (MP3) is seeing a lot of interest and exposure in the free software community. It is a remarkable technology, allowing CD-quality audio to be heavily compressed without loss of sound quality, allowing reasonable storage of large amounts of music on digital media. However, the MP3 format is protected by strong patents. Although the patent holders have allowed certain decoders for the format to be freely distributed, the licensing terms under which the patent is available make MP3 a dangerous standard for the free software community to adopt. This article investigates the broad scope of these patent claims, and explains why we in the free software community, especially those of us interested in music, should seek out alternate formats.

Many people believe that because MPEG layer 3 is a so-called open standard, anyone can implement it. Unfortunately, this is not the case. THOMSON multimedia and Frauenhofer IIS hold a number of patents related to MP3 technology, and have pooled them to be licensed as a unit. Patent restrictions on what is supposed to be an international open standard is a travesty if these patents are not freely licensed to all implementors of the standard - it is irresponsible for any standards body to accept such technology as a standard. The existence of such patents means that even those implementing the technology straight from the standard, without examining the original source code, must pay the licensing fees. But examining the scope of the claims makes it clear that the situation is even worse than it might initially appear. The following discussion is culled from the MP3 Licensing FAQ available at href=http://www.mpeg.org/~tristan/MPEG/mp3-licensing-faq.html

THOMSON and Frauenhofer are relatively generous with the patents for MP3 decoding technology. The licensing FAQ states: "For the FREE distribution of decoders we do not charge a royalty. At the Fraunhofer IIS and OPTICOM web-sites you can find the players we have developed and which may be downloaded for FREE also. Fraunhofer IIS and OPTICOM do not give any technical support for the free players. ... More in general, as long as desktop software decoders are distributed free-of-charge for personal use, no license fee is expected." However, if any money is charged for distribution, or if the decoder is used for more than personal use, a royalty structure automatically kicks in. This leads to two obvious problems. First, no MP3 player may distributed as fully free software; it can only be "free for non-commercial use". This is a big problem in itself - it means that many distributions of Linux or other operating systems will not be able to distribute such software on principle, and that projects like KDE and GNOME that aim to be free software cannot provide fully integrated MP3 support on the desktop. Worse yet, users will never be able to get any MP3 player on CD, even a semi-free one, because the royalties kick in when the software is distributed for any fee, even a copying fee, and even if the software is freely redistributable.

The situation with encoders is even worse. A lot of people have talked about the fact that there isn't any MP3 encoder available for Linux at a reasonable price. The reason is this section of the Licensing FAQ: "Q. OK, I understand that most of the distribution of software decoders is free of charge, is this the same for the software encoders? A. No, while we basically support the freeware concept for software players, for all encoders a license is needed." The royalty for distributing encoders starts at $25 per unit, and carries a $10,000 yearly minimum. Suffice it to say, we will not be seeing encoders at a reasonable price ever. Making music to be distributed in MP3 format will be limited to those who are willing to shell out money, and who are willing to use proprietary software. Composers who are also free software hackers, such as myself, are naturally alarmed.

But the worst part of the MP3 patent is that the patent holders claim the right to receive royalties for any MP3-format data that is distrubuted for a charge, that is broadcast through any kind of "Internet Radio" application, or that is distributed on CD-ROM. I refer the reader to the FAQ for the specific fee schedules, but they are quite high, especially considering how many MP3's you can fit on one CD. This is a serious problem for any hobbyist musician. If MP3 becomes an established format for audio distribution, THOMSON and Frauenhofer will be collecting money on every sale, and only those with connections to a serious record label will be able to distribute their music in a standard format.

Because of these concerns, I urge people in the free software community, any programmers concerned about the increasing danger of software patents, those who believe in the integrity of open standards, and musicians who care about their freedom, to boycott the MP3 format. I also urge everyone to lobby the MPEG group to refuse to standardize any other media formats protected by restrictive patents (this is a serious danger for the upcoming MPEG 4 audio standard as well); to fight software patents in general; and to work on a truly open standard for compressed digital audio.