[BC] AM HD power levels

[Robert Orban]

It has always been practical to transmit audio frequencies up to 
about 18 kHz in stereo FM. but the advent of overshoot compensated 
filters tended to move that bandwidth down closer to 15 kHz because 
it was far easier to control overshoots.

That being said, there have been at least two relatively recent 
published double-blind studies indicating that 15 kHz audio response 
was good enough for high fidelity because people with normal hearing 
to 20 kHz (tested with sinewaves) could not detect the difference 
between 15 and 20 kHz limits in otherwise identical program material 
in a statistically significant way. The reason has been well 
established by psychoacoustic studies: The shape of the frequency 
masking curve of a psychoacoustic critical band has a very sharp 
slope below the masker and a rather gentle slope above. So, with 
program material, the stronger frequencies at 15 kHz and below were 
masking the energy between 15 and 20 kHz.

The great Dr. Harry Olson of RCA Labs always maintained that 15 kHz 
was good enough for high fidelity reproduction. The classic "high 
fidelity" three-channel stereo concerts arranged by Bell Labs over 
phone lines in the '30s used a 15 kHz limit, which was based on 
Fletcher and Snow's work in basic psychoacoustics at Bell Labs. 
Ampex's frequency response spec for their classic recorders (300, 
350, 351...) at 15 ips did not extend to 20 kHz. And so it goes...

Bob Orban

At 10:01 PM 8/31/2008, Phil Alexander wrote:
>In AM we have something called NRSC-2 a/k/a Sec. 73.44(a).
>That rule prohibits anything approaching good quality.
>50 years ago both AM and FM stations were routinely
>transmitting audio from 30 - 15,000 Hz. WLW was certified
>by MacIntosh as transmitting 20 - 20,000 Hz at very low
>distortion. Today we have technology making AM capable of
>transmission beyond that, but alas, stereo prevents FM
>from duplicating this feat.