MP3 is a very popular format for the encoding and compression of audio files, one that's contributed heavily to a rapid increase in availability of digital music online and helped usher in great changes in the way we consume music. Its proper name is MPEG-1 Audio Layer III. MPEG stands for Motion Picture Experts Group (MPEG), a group established in 1988 to set standards specifically for digital audio and video encoding and related technologies. MPEG is a branch of the International Standards Organization (ISO), a Geneva-based group that sets voluntary standards for a large variety of industries worldwide. MP3 can compress mono or stereo digital audio down to around one-tenth its original size, digitally speaking, for easy transfer over the Internet, or for storage of large numbers of songs onto computer hard drives, CDs, DVDs or MP3 players (such as the iPod) without compromising audio quality to any great extent (although it does lose a little quality, sound-wise).
An MP3 encoder quantizes audio data into numbers, which can be scaled down by dividing them all by another number and rounding. Individual bands can be scaled differently to adjust precision. The encoder then uses something called Huffman coding to convert these numbers into even shorter binary strings of information using search trees (in reality, tables of the possible numbers and their binary codes). These tables have the shorter, less precise numbers at the top so that they can be located first. If a sound element is easily perceptible, it will be encoded with more precision than one that is harder to hear to keep the sound quality as high as possible while reducing data file size.
Along with the above, MP3 also takes advantage of some psychoacoustic (how people perceive sound) phenomena to compress audio files down to smaller sizes. For one, it discards the data for any sounds that are below or above what the human ear can actually hear. We can hear frequencies in approximately the 20 Hz to 20,000 Hz range (although the average adult can't hear much above the 16 KHz level due to hearing damage from loud noise exposure that happens naturally throughout life). MP3 encoding also uses the Haas effect, where two identical sounds arriving at nearly the same time but from different directions are perceived as a single sound from one direction, and frequency masking, where a louder sound at a similar frequency to a quieter sound will be the only one heard if both are playing at the same time, to get rid of data. This penchant for discarding audio data is why MP3 is called a lossy compression method, although it uses some lossless methods, too, like Huffman coding. MP3 encoding uses other more traditional compression methods for simpler or more audible sounds that don't fall prey to these psychoacoustic effects, as well.
The creation of MP3 files has come in handy, allowing us to send music over the Internet with ease and to carry around hundreds or thousands of songs on our phones or digital music players. MP3 has become a pervasive music encoding standard. Read on to find out from whence it came.
The MP3 format was given its name in 1995, but its development began years before. MP3's roots were in lots of earlier digital encoding research, perhaps most notably the doctoral work of Karlheinz Brandenburg, who is often cited as the format's inventor. He, however, would be the first to tell you that he didn't do it alone.
In the early 1980s, Brandenburg studied electrical engineering and mathematics at Erlangen University, and pursued his doctorate at Friedrich-Alexander University. Erlangen-Nuremberg. Professor Dieter Seitzer encouraged Brandenburg and other students to work on methods for transmitting music files over integrated services digital network (ISDN) telecommunication lines. Brandenburg studied psychoacoustics and was able to come up with a highly efficient audio compression algorithm that he called optimum coding in the frequency domain (OCF).
Brandenburg became an assistant professor and continued his research. He worked with scientists at the Fraunhofer Institute for Integrated Circuits (a member of a collection of 56 German research institutes called the Fraunhofer Society), which had a working relationship with Erlangen-Nuremberg University. At Fraunhofer, OCF was improved by the team with some contributions from researchers at Hannover University, AT&T Bell Labs and Thomson, and the adaptive spectral perceptual entropy coding (ASPEC) codec was born.
In the late 1980s and early 1990s, MPEG took proposals from various groups that hoped for their encoding methods to be adopted. MPEG chose multiple proposals that were merged into three formats: MPEG-1 Audio Layer I, Layer II and Layer III. MPEG-1 Audio Layer III was largely based on ASPEC.
Other key engineers who worked on the project at Fraunhofer were Ernst Eberlein, Professor Heinz Gerhäuser, Bernhard Grill, Jürgen Herre and Harald Popp, along with additional supporting staff. Together, they developed the software and hardware necessary for MP3 encoding and decoding. In 1993, Brandenburg officially joined the staff of the Fraunhofer Institute for Integrated Circuits as head of their Audio/Multimedia department.
In 2000, Brandenburg, Popp and Grill accepted the Deutscher Zukunftspreis award (or "German Future Prize") on behalf of the entire team for invention of the MP3. And in 2007, the Consumer Electronics Association (CEA) inducted Brandenburg, Seitzer and Gerhäuser into the CE Hall of fame for their contributions to the format. As of this writing, the main researchers all still work at Fraunhofer in various capacities, including Brandenburg, who is director of the Fraunhofer Institute for Digital Media Technology in Ilmenau, Germany. Brandenburg and others have even received a share of the profits from MP3 licensing, as dictated by German law.
Every innovation relies on prior breakthroughs, so related work that occurred before MP3 was even a thought should not be ignored. For instance, the Haas Effect was named after Helmut Haas, who presented the idea in his doctoral thesis in 1949. That same year, Claude Shannon and Robert Fanning came up with the predecessor to Huffman coding. David Huffman was a student of Fano, and he improved upon their technique in 1952. The MP3 researchers built upon the foundational work of many others, as well.
Suzanne Vega even played a small role, as Brandenburg and the team labored to compress the a cappella version of her song "Tom's Diner" without generating distortion and background noise. A portion of the song was used as test audio to determine when the quality had reached an acceptable level.
Microsoft's Big Payout, AlmostPatent law and corporate mergers, complicated animals on their own, both played a part in a lawsuit brought by Alcatel-Lucent against Microsoft regarding the MP3 playback capabilities of its Windows Media Player. German research institute Fraunhofer and AT&T's Bell Laboratories worked on audio compression technology together to some extent after entering into a joint development agreement that started in 1989. Each ended up with patents for various related technologies. But Fraunhofer is the company that is considered the industry patent holder for MP3, and quite a number of companies have licensed the rights from them, including Microsoft (in a $16 million agreement), Apple, Sony, Yahoo, Intel and Texas Instruments, among hundreds of others. But this joint research would spawn trouble down the line. AT&T spun Bell Labs off into a company called Lucent Technologies. Lucent sued Microsoft (with whom it was also embroiled in other patent battles), claiming that it was infringing on Lucent's MP3-related patents. Alcatel, a French competitor, bought Lucent, and the suit was continued by Alcatel-Lucent. In 2007, a San Diego jury ordered Microsoft to pay $1.52 billion for patent infringement to Alcatel-Lucent, an amount calculated based on Windows PC sales since May 2003, and one of the largest patent verdicts in history. Later that year, however, a judge overturned the verdict, ruling that the damages were improperly determined, and that one of the two patents at issue was co-owned by Fraunhofer and AT&T (not owned outright by AT&T), and since Fraunhofer wasn't included in the suit, and Microsoft had a license with Fraunhofer, Alcatel-Lucent didn't have the right to sue over that patent. Of course, this leaves the door open for another lawsuit or an appeal, so as of early 2013, we may not have heard the end of the matter.
How will music evolve next? While the future is difficult to predict, MP3s will probably stick around for quite a while.
© Hemera/Thinkstock
MP3 has had some far-reaching effects that few could have predicted. In 1997, the MP3-playing computer software Winamp became available. Allowing users to easily organize their music files and create playlists, Winamp was a harbinger of players to come. That same year, Microsoft's Windows Media Player began to support MP3 playback. Expensive portable MP3 players, like Diamond Multimedia's Rio 100 and Saehan Information System's MPMAN, also appeared in the late 1990s. They could hold a limited number of songs via flash memory, but those made way for hard drive memory players with higher storage capacity, like the original iPod, introduced in 2001, which played MP3s, AIFF and WAV formats. Newer iPods play advanced audio coding (AAC) natively, which is included in both the MPEG-2 and MPEG-4 standards, and is purported to have better audio quality, but other formats are still supported. MP3s are so ubiquitous that portable music players are often called MP3 players even though most play a variety of formats. Most CD and DVD players can also play MP3s.
Peer-to-peer music sharing service Napster was introduced in 1999, and it caused the popularity of the MP3 format to explode by making more digital music available than ever before. It also caused a great deal of controversy related to copyright infringement, illegal music downloads and piracy. The Recording Industry Association of America, among others, sought legal recourse against Napster as well as individual music downloaders. Napster was shut down in 2001, but other similar services popped up in its place, although perhaps none as brazen.
This newfound availability of music whetted the public's appetite for digital music, which has caused big changes in the music industry. The studios continue to fight piracy, but eventually embraced online digital music sales. Studios began licensing much of their music to companies such as Apple, whose iTunes store paved the way for a new music consumption model. Amazon is also a huge seller of digital music. And digital music sales have overtaken the sale of physical CDs. Digital music has even allowed some artists to break away from studios by giving them a means to easily distribute their own work, democratizing the system a bit. These changes have also paved the way for legitimate music streaming services that make money from advertisements and subscriptions.
MP3 also helped change how and where we listen to our music. We can record larger numbers of songs to writable CDs than could be traditionally held on pre-recorded CDs. And modern players allow us to carry thousands of songs around with us (either purchased digitally or ripped from our own CD collections). The MP3 and its successors have made entire music libraries portable.
There is some debate as to whether MP3s sound particularly good compared to other, less lossy codecs, but despite great increases in Internet speed and storage space, MP3 is still the most common digital music format, and it shows few signs of going away. There have been subsequent MP3 innovations, such as MP3Pro and MP3 Surround, both backward compatible with original MP3. People are even working on things like embedding secret messages or other hidden information in MP3s (MP3Stego, for instance). MP3 has become the norm for digital music, and until something groundbreakingly different comes along, it may remain so for a while.
How MP3 came about and what encoding the files entails made for pretty fascinating research. I've always been a fairly heavy consumer of music, first on vinyl, then tape, then CD, and I think it's fortuitous for music lovers that MP3 and other digital music formats came along and spurred the widespread distribution of music online (sound quality debates aside). I still like owning physical copies, but back in the days when in-store purchasing was our only choice, there was always a good possibility of disappointment. I shelled out for quite a few CDs often to discover that I only really liked one or two of the songs. Many of those disks are gathering dust on my shelves.
At some point I made a rule for myself that I would only buy an album if I knew I liked at least three of the songs (unless it was by one of my very favorite artists), and that seriously curbed my music purchasing. It's not like entire albums routinely get radio or music video play. But then things like Napster came along that allowed us to sample a wider variety of music than we'd been able to encounter by chance on the air before. And then Internet radio followed. I cannot tell you how many new artists I've discovered via my Pandora stations, and how many of their songs I've downloaded from iTunes.
There is probably some charm missing because we no longer get surprised by the b-side songs (before long, no one is even going to know what that means). But it is nice being able to spread the money around on tunes I know I'll listen to more than once. And now I can also digitize those lonely songs from the poor mostly-rejected CDs and listen to them just about anywhere without having to carry a bunch of disks with me, thanks to MP3 and its cousins.
