post #1 of 1
5/22/11 at 9:01am
Thread Starter 

I've been thinking about the topic of audio compression, and if sound waves are really just sine waves, would it be practical to have the studio record music with a "vector graphics" system reformatted for audio? The end result would look something like this for 12 seconds y = 44.3sin(x) + y = 20 sin (x). On the way in, the computer would perform some calculations and break the audio signal down into smaller waves which are added together to get a specific sound. On the way out, the computer would just add the pieces together to achieve a hypothetically infinite sample and bit rate, potentially taking up much less space in the process. Then, the user could just specify how high he / she wants the "sample rate", which in this scenario is related to hardware capabilities. Does this make sense? I'll re-explain if it doesnt, but it sounds like with the right hardware you could scale up to a sample rate in the gigahertz range without "guessing" at what was on the original recording.

Reply