Scientists at the University of Rochester have revealed that they have developed a technique to reproduce digital music that is 1,000 times smaller than a typical MP3 file. The new method uses physical modeling technology to synthesize the sound rather than recording it directly. Professor Mark Bocko, head of electrical and computer engineering worked with two of his students (Xiaoxiao Dong and Mark Sterling) to model the characteristics of a clarinet. After studying and replicating all the physical aspects of the musical instrument, a 20-second MP3 recording was reduced to less than 1 kilobyte (1K) using the new technique.
The University of Rochester's website has a demonstration that is interesting to listen to. Although the new compression method doesn't replicate the recording perfectly, it is still impressive.
On the subject of formats, did you know Apple's iTunes software can convert to different audio formats? To find out more, read the article on how to convert audio formats using iTunes
Comments
There is no comparison between the human sample and the virtual performance. One sounds natural and airy and the other sounds as if it is computer-generated; the tone quality sounds as if the piece was being played on an inexpensive synthesizer. There is no substitute for the real thing — analog!
Well “G. Ross”….I’m pretty sure they only used the clarinet in here to test out the new technique, not to replace anything. If/When something is replaced, it would be the old format we are used to now, with smaller sized files.
I proved them wrong by not only attending university, but have been a great success in my professional life. ,
Part of maturity is being able to appreciate, or at least tolerate, the success of others. ,