Quote:
Originally Posted by Gogar
As a side note to Todd's comment I'm really amazed when working with people in the industry how many people confuse data compression with audio compression; also how many people confuse recording bit depth vs. the processor depth in their computer.
|
Just to expand on that:
data compression - you throw away data that an algorithm decides isn't important. In other words, the computer does some math to determine what is "good enough." That happens in music (e.g. mp3 and aac), photos (e.g. jpg, png), video (e.g. h.264). It is a strictly digital process.
audio compression - takes a waveform and reduces the dynamic range, mostly by limiting transient spikes (though can also expand quiet passages as well). This allows one to "even out" the sound of a recording and optimize for a particular medium or application. In extreme cases it can be used as an effect (breathing/pumping). Old school this was done analog, both by compressors (by design) and by recording tape (an artifact that people liked). It is often done in software now, but it is looking at the audio waveform as opposed to the data structure.