I don’t really understand how you could gain quality from a 128kbps or 192kbps mp3 when re-encoding it to 320kbps. The “information” is already lost and cannot be re-obtained. In fact, these “fake” 320kpbs (re-encoded out of lower kbps) are th emain reason many tracks found on the net, supposedly at 320kbps are not really sounding as such.
edit: unless I misunderstood the meaning of your posting, I read it as if you are stating that you re-encode mp3s to gain quality – if you just state that a 320kbps of unknown source could be a re-encoded 128/192 mp3 hence it sounds bad, then I agree 100%
Now, to answer the OP, the dual stereo format is the higher quality of stereo seperation you can get – always go for that. Joint stereo provides smaller file size but compromises the stereo seperation of the sound. The difference is mainly audible in quality speakers/monitors but not so much in clubs where the “stereo image” is neglected by the design of the club in the first place.
Another thing that supposedly could provide different results of a wave file when encoded to 320 kbps mp3 is the algorithm used – the most famous ones being “lame mp3 encoder” and “Fraunhofer mp3 encoder”. frankly, I have tested both and I wasnt able to detect noticeable differences in the final result (at least audibly). I never bothered to “null test” the resulted mp3s to see if there are any differences.
.
