[Don Pearce[_3_]]

On Sun, 30 Oct 2011 22:43:08 -0700, (Dave Platt)
wrote:

In article ],
isw wrote:

As indicated by iTunes, an increasing proportion of mp3s are being
encoded in "joint" stereo. Does it actually "save storage" and
"improve performance" as it is touted to?

IIRC, it's a "sum and difference" process (L+R, L-R), which results in a
lower bitrate for the same quality as compared to encoding the two
channels individually -- the L+R channel is just as easy/difficult to
encode as either L or R alone, while the L-R channel contains a whole
lot less information *in most cases* and so takes a smaller bitrate to
encode. If well done, there's no reason why there should be a reduction
in quality; no information is lost by that process.

It's perhaps worth noting that LP records used what amounts to a
"joint stereo" encoding. The L+R signal is encoded as a horizontal
motion of the stylus, and an L-R difference signal is encoded as
vertical motion.

There were several reasons for doing this - compatibility with older
and less-expensive monaural playback turntables, improved
trackability, etc.

This form of joint stereo coding - called mid-side - is really for the
convenience of mono compatibility. It does not cause any information
loss and permits no compression.

The kind we are talking about here is intensity coding. in which the
high frequencies are combined into a single channel, with just a
little bit of side information and some panning instructions for the
codec. Low frequencies are pretty much left alone.

This does give a once-useful data reduction, but doesn't sound
particularly convincing for critical listening. It can also go
horribly wrong with some audio cues.

d