[Amps] Audio/RF compression

[Jim Brown]

On Thu,1/12/2017 10:51 AM, Joe Subich, W4TV wrote:
> Delays of over 100 ms begin to become distracting.  It is a matter of
> the user's ("talent" for broadcasters) ability to handle distraction. 

YES! One thing I did a LOT of was providing sound for large events both 
indoors and outdoors. In such systems, the acoustic travel time between 
loudspeaker and listener typically ranges from as little as 20-30 msec 
to as long as several hundred msec. This is all due to the speed of 
sound, which is approximately 0.9 msec/ft.

Latency becomes an issue when main loudspeakers are 30+ feet over a 
stage and further displaced side to side, and spill sound to the stage, 
often in the range of 30-40 msec. This drives musicians nuts, and is 
part of the reason large and complex stage monitor systems are used. 
Performers hear themselves undelayed (or minimally delayed) at levels 
higher than the leaked sound from the main system. Originally this was 
done with big speakers on the stage floor; now it is far more commonly 
done with miniature in-ear headphones connected to stereo UHF receivers.

And as Joe has noted, latencies greater than about 80 msec can turn a 
highly trained announcer into a babbling fool; we've all witnessed guys 
like this slow down and stop. I've experienced it myself talking into a 
mic at the mix position (100 ft or so back in the audience) of a large 
outdoor system.

So far, what I've described is entirely due to acoustic travel time 
between loudspeaker and listener. Roughly 25 years ago, we began using 
DSP extensively in large sound systems. One of those early systems had 
so much latency that it was unusable; the first successful systems had 
latencies in the range of 7-15 msec, depending on the size of the DSP 
system, number of channels, amount of processing being done. These 
latencies ADDED to the acoustic travel time, so keeping latency low was 
always important.

73, Jim K9YC