Back to the topic again, and to repeat myself, the actual format of the identical source material can make a big difference to the apparent quality. I'm playing quite a bit with this decompression thing at the moment, and I'm using Bruce's generously made available track of dynamically uncompressed Latin music to play with. Thanks, Bruce, and sorry, I'm doing some terrible things to the content, in the name of science!
The track, called Bronze, downloaded in MP3, was obviously very clean and dynamic. So, in the playing around with the track it's now in standard WAV, and 24/192 WAV: in the process the size of the file went from a demure under 8 Meg size to a whopping 266 Megs, obviously that with the 24/192 version. And you know what's coming: on playback on a completely normal desktop PC, using the built-in DAC, to completely standard PC speakers; absolutely no tweaking or funny business indulged in, the worst playback is the original MP3, next is the normal WAV, and the best by a big margin is the 24/192. This is in all the key areas of treble sparkle, sense of the acoustic, space: there is an old Wurlitzer (I think!?) electric piano in the mix, and the difference in the tonal quality of that is quite dramatic.
So what's this saying then?
Frank