[BC] AM audio bandwidth - AES Journal article

[Robert Orban]

At 08:09 PM 6/18/2009, Peter Smerdon wrote:
>Thanks for the reply Bob.
>
>Reading the paper, I didn't absorb that the testing it was an NRSC AM
>Study Group project.
>
>Now that I read the abstract and paper more carefully, I can see that is
>mentioned tangentially in both.
>
>Is this work going to result in a new NRSC recommendation?

This testing resulted in "Guidelines," but not recommendations. 
Guidelines are strictly voluntary.

http://www.nrscstandards.org/SG.asp

See NRSC-G100 on this page.

The short summary is that with today's receivers, the optimum 
transmitted audio bandwidth is somewhere around 7 kHz when one 
balances audio quality and interference. Indeed, even when one does 
not consider interference there appears to be no improvement in audio 
quality with typical radios when the transmitted audio bandwidth is 
increased beyond 7 kHz. There may even be a seemingly paradoxical 
reduction in audio quality, probably because the transmission audio 
processor's final clipper is generating difference-frequency IM 
distortion from power between above 7 kHz. The radios are essentially 
deaf to the power above 7 kHz but can reproduce the IM distortion 
quite nicely. In addition, any IF asymmetry can cause IM distortion 
in envelope detectors that is lower in frequency than the modulation 
causing it.

Bob Orban