It's an interesting point. Both MP3 and Vorbis work by discarding what they believe to be inaudible details in the music. Now, once they're gone from the source after the first encoding, how much additional information will be lost on subsequent decode/re-encodings? That detail has already gone. Now I'd expect some artifacts as both encoders are designed to work on pristine input and won't be tuned to work on already processed data, but I wouldn't expect it to be that significant a drop, at least for a single, one time conversion, though clearly there is bound to be some degradation.
Having said that, I'm one of those people who only use flac because I can here the difference between lossless and lossy encoders, or at least I believe I can. Not during normal background playing situations, but when I have my headphones on and I'm actually sitting 'listening' to it. For the rest of the time, I find anything above 192bps mp3/vorbis good enough. I prefer vorbis because it does a better job on gapless tracks where as mp3 always stutters between tracks, but vorbis support in portable players is few and far between so sometimes you just have to use mp3.
I'll be looking to bulk convert my flac collection to mp3 shortly myself for when I go out and about and want a little music, so I'll also be checking out the perl program listed above when the time comes. Thanks for the suggestion.
|