Re: BBC R3 gone from DAB multiplex

Roger Wilmut wrote:
In article <cml2v21g5gpuv18ho73b561mlseenkgs06@xxxxxxx>, Roderick
Stewart <escapetime@xxxxxxxxxxxxxxxxxxxxxx> wrote:

On Thu, 8 Mar 2007 14:28:15 -0000, "Serge Auckland"
<sergeauckland@xxxxxxxxxxxxx> wrote:

In joint stereo, the stereo sum and difference are
coded separately, and in FM, L+R and L-R are transmitted seperately and in
both cases reconsituted in the receiver. Also in both cases the L-R channel
carries less information than the L+R channel.
Less of what? Less bandwidth? Goodness knows what abominations are
perpetrated on digital radio signals but I'm sure I have a BBC
Engineering Information *** somewhere describing the stereo analogue
signal, and I don't recall any mention of filtering of the L-R signal.
Is this really what you're saying?

Rod.

He is wrong about this: see my earlier post. The FM ndifference signal
is used to modulate the 38kHz subcarrier and has pre-emphasis, but is
not otherwise tampered with.

It's the left and right channels that have pre-emphasis - i.e. it's
applied to the channels before matrixing into M & S.

The greater noise in the S channel is a consequence of the triangular
noise spectrum of FM: the noise power in the demodulated signal is
proportional to the noise frequency.

It's interesting ... in an absurdly anoraky sort of way ... to speculate
how the noise in the S channel would compare if the same multiplex
(Zenith-GE pilot tone) was transmitted by, say, VHF AM. In this case,
because there are two sidebands in the subcarrier, I reckon the noise in
the S channel ought to be 3 dB *lower* than in the M channel. (Signal
sidebands add in phase, noise sidebands add in quadrature).


--
Richard Lamont http://www.lamont.me.uk/
<richard@xxxxxxxxxxxx>
OpenPGP Key ID: 0x5096714C
Fingerprint: F838 740C 76B4 6EC6 9ECC 1C4D A4DE 3322 5096 714C
.