QUOTE (Leo 69 @ Mar 10 2005, 10:11 AM)
To my mind encoding from wav to, for example, mp3, then decompressing mp3 back to wav and then
again encoding into mp3 (and again, and again) will face the mp3's flaws and they will be more and more evident to the listener with each reencoding. Maybe this will help LAME developers to find and fix some weaknesses at an "early stage", so to say, won't it ?
Its not that flaws multiply, rather the "desired" quantization noise multiplies. MP3 is SUPPOSED to insert noise into a recording, so this wouldn't help LAME developers insert less, or anything like that.