Re: [Discuss-gnuradio] Stereo gain "hacked"

[Robert McGwier]

Al, Matt:

Thanks for your input.  I am pretty ignorant about these FM standards 
and basically I googled my way to a reasonable spec,  drew the flow 
graph and coded it.  The stereo demux was definitely not working with 
DEemphasis applied to the composite signal.  And it was clear why.  The 
taper was awful and the lower sideband was approximately 10 dB higher 
than the USB on the L-R, DSBSC stereo signal.  I removed the deemphasis 
and voila, stereo.  That said,  when the pilot/carrier is 20 dB out of 
the noise in the 512 pt FFT, that should be strong enough that you do 
not have hiss that is irritating.  But the COMPOSITE signal is what is 
FM modulated but I bet they do not preemphasize the composite signal. 
They might do preemphasis on the L+R baseband signal as Matt suggests 
but I just don't know about the others.  I am trying to understand what 
about that would make sense.  It would be very interesting to read a 
real spec.  I have googled and not really found one a  useful one.

I hate this RDS Top Secret MONEYWORD specification.  Frank and I are 
determined to fix their wagon.

I did not know about Achilleas work.  I will search the archive.

Bob



al davis wrote:
> On Monday 13 February 2006 20:06, Matt Ettus wrote:
>   
> > I seem to remember that the preemphasis on stereo signals is
> > not performed on the multiplexed signal.  It is actually
> > performed on the audio components separately, before mixing
> > with the stereo subcarrier. Therefore deemphasis needs to be
> > done after the stereo part is mixed back down to baseband.
> >
> > To me, this is backwards, and is not useful, but I think that
> > is the standard.  Achilleas sent an email to the list several
> > months ago on this subject.  He also had a simple
> > implementation that did stereo.
> >     
>
> Actually it is very useful the way it is.  It would be bad to 
> preemphasize the composite signal.
>
> To preemphasize the composite in effect converts the system to 
> phase modulation.  The subcarrier would then consume most of 
> the bandwidth, and the baseband (mono) would be drastically 
> reduced, resulting in a significant reduction in apparent mono 
> signal to noise ratio. Stereo signal to noise ratio would be 
> about the same as it is now, but there would be no advantage to 
> mono.
>
> In the existing system, if stereo SNR is not good enough, you 
> can switch to mono and get almost as good SNR as if stereo 
> wasn't there.  "Almost" means about a 6 db degradation, which 
> is due to the fact that baseband modulation must be lowered by 
> that much.
>
>
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
>
>   


--
AMSAT VP Engineering. Member: ARRL, AMSAT-DL, TAPR, Packrats,
NJQRP/AMQRP, QRP ARCI, QCWA, FRC. ARRL SDR Wrk Grp Chairman
Laziness is the number one inspiration for ingenuity.  Guilty as charged!