The MP3 Is Dead… Long Live the MP3?

Despite the expiration of the MP3 patents, users across the world are still going to be using this format for some time. But how long will its cultural influence last, and how are Digital Humanities projects facing issues like long-term digital preservation in view of the format’s so-called “passing away”? The MP3 audio compression format triggered some major shifts in the music industry, reconfigured audio playing devices and rematerialized the way we approach our music collections. It also became the object of many scholarly studies, such as George Yúdice’s Nuevas tecnologías: música y experiencia, Jonathan Sterne’s MP3: The Meaning of a Format, and Stephen Witt’s How Music Got Free. But the Fraunhofer Institute for Integrated Circuits recently announced that their patents for the MP3 have expired and their licensing program has terminated after 24 years.

Perhaps due to the announcement’s conclusive tone (shouting out to all who took part in the project, thanking them for their support), it has been interpreted by some media as “the death of the MP3.” As both Sterne and Witt make it clear, other “deaths” have been previously heralded, as it struggled from the very beginning with other competing formats, like RealAudio, or MPEG-2, which for some time seemed to win the match. As early as the mid-90s, Sterne reminds us, an Australian hacker reverse-engineered a Fraunhofer MP3 codec and released it under the name “Thank You, Fraunhofer.” Here is where both narratives of the format’s history diverge—whereas Witt considers the MP3 thrived on despite being hacked, for Sterne this contributed to its popularization.


In a typical reaction to the news, NPR’s Andrew Flanagan considers, “We may still use MP3s, but when the people who spent the better part of a decade creating it say the jig is up, we should probably start paying attention.”


But is this really the end of the format? 
It probably would be if it hadn’t been released to the public domain—and officially it hasn’t—but there are other “liberation” stories besides “Thank You, Fraunhofer.” Since the late 90s, LAME, an encoder licensed under LGPL, has cleared the way for open source audio compression. It is used by digital audio workstations like Audacity, CDex, and Virtual DJ. Although its developers claim it does not infringe any copyright law, since its source code is released only for educational purposes, they do remind us that, in some countries, using it can interfere with Fraunhofer’s patents—that is, until now. Version 3.99 was released in October 2011, and its most recent revision was in February 2012. The latest version, 3.100, has not been released yet, and there has not been any news from the developers after the expiration of Fraunhofer patents. However, open source software programs like freeware audio player Foobar2000 are already adding LAME to their encoder packs after the announcement.


It is true that there will be no more industrial development to the techniques that brought about its creation. But the MP3 will be circulating in Western and Westernized cultures as long as the infrastructure that supports it still works.


This is not the same as when the last company manufacturing VCR players announced they would stop producing them. People can still create MP3 files without Fraunhofer; new improvements to the format (if any) will come from open source communities, rather than the industry. AAC has been usually referred to as the “natural” heir of the MP3’s kingdom, but it will hardly have the same cultural impact. As Sterne reminds us, “To succeed, the MP3’s eventual replacement will require its own combination of technical processes, multi-industrial and transnational regulatory formations, user practices, and opportunities. Whatever it will be, we know that simple technical improvements or new business models are never enough.”


There has also been a severe “path dependence” (the tendency to using one particular standard or a technology instead of another) from audio industries on this format, which is difficult to resist. This concept, coined by Paul A. David, and studied by media scholars such as Trevor Pinch, explains the success and domination of MP3—a format that shares the histories of the music, computer, electronics, and broadcast industries—not by having the best audio quality in the market (which it hasn’t had, and probably never did), but rather by the strong inertia this standard exercised for years upon the very media that made its birth possible in the first place.


Sterne explains, “Once manufacturers and users adopt a system built around a certain standard, the standard becomes a self-reinforcing phenomenon. Both manufacturers and users have interests in the persistence of the standard (or ‘path’), since a change in standard means a transformation in manufacturing equipment and sometimes major purchases for users. So the potential advantages of a new standard have to outweigh the cost for either manufacturers or users.”

This transition to a new standard will take longer to happen in, for example, Latin America, where selling CD-Rs of USB drives full of MP3 files is still a business for informal computer stores and marketplaces. In underdeveloped or developing nations, the fading away of the MP3 infrastructure will take much longer than in Western Europe or North America.


So what will happen to MP3-based audio repositories, like PennSound or UbuWeb? In fact, this question raises the conundrum of digital preservation in general—we do not really know how to preserve files that in the long run will become obsolete (that is, inaccessible through available infrastructures). Some suggest the use of prevailing standards in digital formats, software, and protocols to tackle this question. Others consider emulation as the best option without reconverting files and losing data in the process. Format migration is not recommendable for MP3 files, since it is an end-use format, and re-codification is strongly discouraged by its promoters. However, people regularly re-code and circulate MP3, as in the case of mash-ups. The result, of course, is a loss in definition, just as when you photocopy a document’s copy too many times.

The dominance of MP3 over other files may not be that critical for already established digital audio repositories, but it certainly is for a work in process like PoéticaSonora. Created by faculty members and students from Concordia University (Montreal) and UNAM (Mexico City), PoéticaSonora is engaged with promoting and preserving sound art and sound poetry in Mexico.


One of our most ambitious projects is the creation of an online database with refined search tools to access sound art and poetry works. Sound files have been donated by several sources—cultural institutions, private collectors, artists themselves, and so on. Most of them come in MP3, a few others in lossless formats, or on CDs. In these cases we follow suggestions from digital preservation experts Lisa Goddard (U Victoria) and Kelly Stewart (SFU) and use lossless formats for preservation, whereas MP3 is intended for access purposes. This way we get a backup that is not dependent upon external supports prone to deterioration, like CDs. It also means, however, that there might not be a high-definition replacement for some files.

 

In these cases, we will have to make do with MP3s. We still do not have a definite answer for this problem, and we keep up to date to innovations in the field to figure this out.


Overall, it is good that the industry is moving forward from the MP3. For authors like Sterne, the perceptual limitations behind it made it a perfectible format. Paraphrasing John Philip Sousa’s derision of recorded sound as “canned music,” Sterne claims, “MPEG audio is processed sound for listeners who live in a processed world.” 
The AAC is not as hip as it predecessor, but it is just as processed. Until another format does not break its ties with psychoacoustics and its history of biased sound processing, we cannot claim there is a strong candidate to occupy the place that MP3 has had in our cultures during the last three decades.