[Anders]

I have upgraded my computer to an EMU 1212M sound card. I am very satisfied with it and it is fun, but my only earlier experience is my previous standard card (built-in on motherboard) which I hardly used. The EMU sounds good with my EarMax Pro amplifier and Grado RS-1 (equalised to get rid of the upper midrange peaks).

I am now trying some software and wonder about where things happen in this chain of software and hardware. The quality of mp3's seems obviously to depend on the quality of the ripping software and the bitrate. When that's done it's done. But what happens after? Does the software player (Winamp, Windows Media) etc. process the sound (except equalisation etc. that you may have set yourself)? Or must the audio stream be converted in some way before it enters the soundcard/DAC?

 

[sgrossklass]

Not all MP3 decoders are equally accurate - check out the MAD website for a comparison. That said, today that's no longer that much of an issue, with the biggest errors being simply rounding related (remember that MP3 is lossy so the PCM samples put out are not exactly the same as in the original); more advanced decoder libraries like libMAD or the Shibatch MPG123 thingy can output in >= 24 bit or use dithering with noise shaping in case of 16 bit output.

 

[wistily]

It has been proven with double blind studies that foobar had the best mp3 decoder. However, this doesn't make much of a difference.
FIY, the audio chain is :
CD->ripper->WAV file (pcm)->encoder->MP3->decoder->PCM->soundcard

The limiting factor is the quality of the encoder, and then perhaps the quality of the soundcard followed far behind by the quality of the ripper and mp3 decoder at the same weight imho.