[BC] Nominations For Best sounding AM
Robert Orban
rorban at earthlink.net
Mon Feb 22 15:05:09 CST 2010
This is a hard question because the answer is "it depends on the
radio you are listening to and the amount of noise and interference
on the channel at the location where you are listening." As the
recent NRSC study showed, AM radios vary widely in frequency
response, although their mean -3 dB frequency is about 2 kHz. A
station that sound "harsh" on a wider-band radio may the station that
best punches intelligibility through on a narrowband radio.
This is why creating processing presets for AM audio processors is
such hard work. (It's Greg Ogonowski's and my least favorite task in
developing AM processors.) No matter what you do, AM audio cannot
compete with modern, wideband standards (like CDs) because of the
radios. The presets have to represent a compromise that sounds decent
on radios that are both narrower and wider than the NRSC mean. For
the narrowband radios, intelligibility is a real problem because they
are significantly narrower than the 3 kHz usually considered to be
necessary to preserve speech intelligibility, at least according to
the many tests at Bell Labs that created the standard for
"toll-quality" telephony.
The other problem is the noise and interference environment on the AM
band. This requires minimizing the peak-to-average ratio of the audio
without driving listeners away due to artifacts. Audio with a lot of
dynamics compression is preferred by audiences when the main goal is
easy comprehension of speech when there is a lot of noise and
interference. The same amount of processing may sound over the top
where field strengths are high compared to the noise and interference.
There are, of course, potential audio problems that will take away
points from any audio quality assessment. Examples are codec
compression artifacts caused by too many cascaded codecs and/or too
low a bitrate, and obvious hum and/or noise in the audio. Clipping
mic channels and program busses are another problem. Clipping should
be done in one place and one place only -- the audio processor -- so
that it can be controlled. But assuming that the audio path is clean
before the transmission audio processor, all you can say is "the
best-sounding AM is the AM that is best processed for a station's
target demographic, for the radios that that demographic is likely to
own, and for the noise and interference environment where they are
likely to listen."
Bob Orban
More information about the Broadcast
mailing list