Stereo VS Joint Stereo

[Anakin]

Which one will sound better at 128kbps on an mp3 player. My friend and I were discussing this at work.

[jagabo]

At 128 kbps joint stereo should sound better. To some extent it will depend on how much difference there is between the two channels.

[turk690]

MP3 compression provides for a number of ways with which to produce the same quality on ever-slower bitrates (or smaller files, for the same bitrate). Normally, an uncompressed stereo *.wav file is in turn composed of two independent parts (L & R or ch1 & ch2). Conventional treatment when compressing to MP3 will treat these parts independently and produce a file of x size when compressed using y bitrate. Most modern stereo material, however, has a high degree of commonality between the two parts (mono content, such as vocals and bass). If chosen in the encoding parameters, the joint stereo feature attempts to seek this common material and assigns bits common to them so that now, even still at y bitrate, resulting file size is lower than x compared with if joint stereo wasn't chosen.
If, theoretically the encoder did its job well, there should be NO difference in sound quality between two files of the same bitrate one conventional MP3 the other joint stereo MP3 encoded from the same uncompressed *.wav source. The file size of the joint stereo MP3 file will be lower than the other; just how much lower will depend on how much mono content the original uncompressed file had.

[jagabo]

two files of the same bitrate one conventional MP3 the other joint stereo MP3 encoded from the same uncompressed *.wav source. The file size of the joint stereo MP3 file will be lower than the other

If the file size is lower the average bitrate is lower.

[Cornucopia]

No, turk690,
since FILESIZE = BITRATE * TIME, with the same source (time) and choosing the same bitrate MUST give the same (give or take) filesize. Assuming CBR encoding.
What's different is the encoding efficiency. This translates directly to quality. Hence, at lower bitrates (128 would be in this range), joint-stereo is higher quality than stereo. Unless you had material that was very different between channels, OR those slight differences were VERY important (read: BINAURAL, etc). In those cases, you should use either stereo or dual-channel encoding.
The difference between the 2 there is in bitrate allocation. DualCh always keeps the same allocation to both channels evenly, regardless of the material; whereas Stereo will allocate more bits to one channel over another (temporarily) if it needs finer quantization at that point in time.

Scott

>>>>>>>>>>edit: you beat me to it, jagabo!