Hi,

I encode with lame 3.90.3's alt-preset insane (320 cbr), and aim to output the highest-quality sounding MP3 file. Technically speaking, would encoding in stereo (-m s) or joint stereo mode produce a higher quality file (no matter how slight or negligible the difference)? I know the difference is insanely subtle, I just want to know technically.

Joint stereo would provide the higher quality, as in this mode the encoder is free to choose between left/right stereo and mid/side stereo on each frame.

So the reasoning is M/S stereo takes less Mb, leaving more room for sound quality?

Exactly.

When stereo is used, there's a risk of more ringing, pre-echo etc. Makes you wonder why rippers that use LAME expose the stereo switch at all.

It's your choice, but I really don't see there being any reason for using anything higher than APS when encoding with Lame.

And you should stick to using Joint Stereo, which is part of the alt switch you are using.
J

I don't see any reason to use anything less than -V0 --vbr-new, I have enough space and still can abx some -V2 files ...

Depending. In the event that M/S representation requires more bits than L/R, it won't hesitate to use L/R, at least so I assume, since it makes sense.

The only lossy joint stereo encoding is intensity stereo, which is typically used only for low bitrates (112kbps or less), and is not implemented by LAME anyway.