I am rather new to the concept of joint stereo (only just found out what it probably means). As I understand it, the data that is different between the two channels is put into a side channel while the (usually much larger) information that is common to left and right channels goes into a "main" channel. Proponents of joint stereo say that in compression, this makes a higher absolute error on the side channel, but a lower proportionate error than with stereo compression, because the data size is much less. They then say the absolute and proportionate error with the common track information is much less than with stereo, because far more resources are devoted to it, so it follows that unless left and right channels are completely dissimilar, joint stereo must give a closer approximation of the original file, and with a smaller file size than stereo when decompressed into left and right channels again The main difference I can personally tend to hear in joint stereo compared to stereo is not in the stereo separation but in a less bass orientated equalisation. There is also some greater transparency, which may partly be as a result of lighter bass. Of course this could be merely the implementation of joint stereo in the editors I use. My point in writing this is that the joint stereo proponents suggest the Lame MP3 encoder should always be used to compress with joint stereo, not stereo. Other lossy formats which might seem better than MP3 (AAC , MPC ...) don't seem to offer joint stereo options in the encoders I have come across. Is joint stereo yesterday's issue now?