https://www.heise.de/hintergrund/30-Jahre-MP3-Als-die-Musik-ins-Internet-kam-10485346.html
I guess we did a good job already, 10 years ago such an article would very likely not have included PASC.
1 Like
In English as well.
Anniversaries in technology are tricky – because there are usually several dates that come into question as a birthday. This is also the case with MP3: On July 14, 1995, the project participants at the Fraunhofer Institute for Integrated Circuits (IIS), which was in charge of the project, decided on “.mp3” as the file extension for audio files compressed using their method. Up to this point, the files were called “.bit” in-house. However, the sophisticated system, which some even criticized as overly complex, was already fixed at the MPEG meeting in London from November 2 to 6, 1992. The development history even goes back to the 1970s – so the “pregnancy” lasted more than a decade.
MP3 is short for MPEG 1/Layer III – The abbreviation MPEG (Empeg pronounced) stands for “Moving Picture Experts Group”. MPEG dealt with data reduction processes for images and therefore also had to come up with something for sound. The aim was to reduce the comparatively high data rate of PCM (Pulse Code Modulation) digitized audio signals to quantities that would enable digital radio broadcasting or use on data carriers that have weniger capacity than the CD.
This was being researched in many places around the world – Sony was working on Atrac for its MiniDisc, Dolby on AC-1, the precursor to AC-3, which was later marketed as Dolby Digital. In Germany and the Netherlands alone, there were initially three camps: the first group was formed around Philips and the Munich Institute for Broadcasting Technology (IRT) and worked on a system called “Musicam”. Karlheinz Brandenburg was initially a lone wolf, researching ways to tame the flood of data from digital audio signals in a subjectively sound-neutral way at the University of Erlangen. “We wanted to store the signal in such a way that it contains everything that is transmitted from the inner ear to the nerve pathways,” says Brandenburg in an interview with heise online. Team three was formed with Detlef Krahé at the University of Duisburg and Ernst F. Schröder from Thomson (then Telefunken, now Technicolor). Ultimately, it was teams 2 and 3 whose work resulted in MP3.
Patent denied – “It can’t work!”
Brandenburg was on a mission: in the 1970s, his doctoral supervisor, Professor Dieter Seitzer, later founder of Fraunhofer IIS, had the idea of distributing music in hi-fi quality via the emerging ISDN telephone network. He wanted to protect his initial ideas – but the patent was initially not granted because the examiners were of the opinion that it was not possible to transmit music at the targeted bit rates using the current state of the art. Seitzer then found a doctoral student in Karlheinz Brandenburg, who took on the subject. Brandenburg soon moved to the newly founded IIS, where a group headed by Professor Heinz Gerhäuser continued his research.
Initially, all researchers were slowed down by the available computing power – and it could take hours to encode a sound sample with the required parameters. In Europe, the technology gained momentum with the award of EU project 147 (“Eureka”), which was intended to help digital broadcasting achieve a breakthrough in the EU. At the same time, the first real-time systems were available, which significantly accelerated research. MPEG was founded in 1988, and after various listening tests, MPEG 1/Layer I was first used briefly as a slimmed-down Musicam version on the long-forgotten digital compact cassette (DCC).
Long dry spell
MPEG 1/Layer II is the somewhat more complex Musicam variant. It is the standard for SD digital TV and radio via cable and satellite and is used for terrestrial digital radio DAB (Digital Audio Broadcast). In Germany, however, the first DAB was not able to establish itself. This was not achieved until 2011 with DAB+, whose codec is based on AAC (Advanced Audio Coding). MPEG 1/Layer III – also MP3 – is based on the Thomson/Fraunhofer development ASPEC (Adaptive Spectral Perceptual Entropy Coding) and the modified discrete cosine transform (MDCT). The latter is one of the MP3 key technologies – and came into the system after the University of Hanover joined the project in 1988 with Hans-Georg Musmann and Bernd Edler.
While Layer II won the race in broadcasting, the Thomson and Fraunhofer teams were initially left behind. But their patience paid off. Thomson man Schröder told heise online: “Brandenburg had foresight and perseverance.” He looked for applications outside of traditional broadcasting and relied on the rapidly growing computing power of PC processors.
Then several things happened almost simultaneously: ITT-Intermetall (now Micronas) supplied the MASC 3500, the first single-chip signal processor that could be programmed as an MP3 decoder. The – failed – digital satellite radio “Worldspace” used MP3 as a codec, and the US company Telos Systems built MP3-based technology for feeding outside broadcasts to radio studios. Brandenburg: "Their founder, Steve Church, gave us tips on how we could market MP3. Among other things, he told us: 'Look at the Internet!
2 Likes