Movement to Reduce AM Radio Bandwidth Ignores Hearing and Hardware

[W2NBC]

 This is an article that some may have seen regarding NRSC standards and brickwall filtering artifacts.
If you're daisy chaining EQ's to achieve close to a brickwall filter, you may want to think about where you start the filter and how sharp the curve ..

http://www.rwonline.com/article/opinion-39let39s-keep-am-sounding-good39/16823

[Opcom]

That might be the best description yet of group delay and why it can mess up the audio.

[w4bfs]

the choice of filter type has large results in group delay distortion .... in the analog filter days the Bessel response filter had the best group delay performance ....I don't know about DSP

[W7TFO]

The root problem is the endless loop between the Federal Clusterfrigging Committee and the brodcoopingcastrater.  

Nothing more than a continuous race to see who is the dumbest, from technical techno-midgets to insolent bean-counters.

Bah Humbug!

73DG

[k4kyv]

One thing I don't quite get in the article is what he says about the telephone audio on the talk-show sounding better than the studio audio, because the phone audio was already brick-wall low-pass filtered by the land-line network when it entered the studio mixer board, while the announcer's voice was brick-wall filtered later in the audio chain by the local low-pass filter. The audio was still brick-wall filtered, wasn't it? So shouldn't the group-delay artefacts still be there, regardless?

I can attest to the effects of a sharp cut-off low pass audio filter.  I have a 3400~ filter available for my audio, which I often use under congested conditions.  Most people say the audio sounds pretty good with the filter in, but I can easily hear a harsh rattling sound in the overtones of the voice, which disappear when I switch to the more gradual roll-off that occurs over the range between 5.0 and 7.5 kc/s. Most people agree they can hear the harshness of the 3400~ filter, once I call their attention to it and they listen for it.

The mechanical filters in the 75A-4 and R-390A cause a similar degrading of audio quality.  That's why the R-390 (non-A) sounds so much better than the "A" model.

[K5UJ]

The mechanical filters in the 75A-4 and R-390A cause a similar degrading of audio quality.  That's why the R-390 (non-A) sounds so much better than the "A" model.

the 3.1 kc mech filter in the 75A3 was intolerable.  the cheap ceramics are significantly better audio partly because of the shallow skirts.  The thing that's kind of cool is weak signals stay in the 6 dB ~flat response curve; stronger signals (that you'd want to enjoy with good audio) push their way farther down the skirts a bit for a wider response.  There's some double peaking going on across the passband but I don't care--I'm not trying to work DX.  I don't see spending the high dollar amounts for the A line mechanical filters I see advertised in ER.  Maybe if I were going to do some serious signal digging.