I'm inlcined to agree with Martin in general. With traditional analog, I could get a 12 db sinad with as little as .3 microvolts on most mainline systems. When we went to digital conversion, I figure I lost as much as 5 db of usable signal as the digital systems would struggle to decode much below .5 or even .6 microvolts. Even at that level, the cut out was usually more than most users would tolerate. With analog, you'd get static at 12 db sinad, but you always had the signal, unless you set your squelch too tight.

Even with FleetSatCom in the Navy, signal degradation and propogation loss was a common bugaboo. It was more than frustrating when we'd be sitting at our good old R-390s chasing russian code and the damned df equipment wouldn't discriminate the signal out of the grass. I could hear the signal just good enough to make copy, but the stupid million dollar hardware couldn't even tell it was there. But that's besides the point, the sat come guys were constantly fussing with their gear trying to maintain a signal lock on a TDM from the Pac Sat and half their traffic would be lost as the ship rolled which usually caused a 30 db signal swing as they lost peak azimuth. You'd never have an issue like that working analog.

I do agree, distinguishing between signal quality and the quality of the programming is important. In my opinion, it doesn't matter what signal format they use these days, the programming stinks, and I'm not even sure free is worth it anymore.

As for audio signal level during commercials, if it is just compression, why is my db meter reading twice the signal level during commercials? Sound power is sound power, and it doesn't matter if you are feeding audio fidelity at 10 Khz if the majority of the audio is voice because the power band for voice is less than 5 Khz, so the only way the voice audio level goes up is if someone is jamming the gain over 0db. However you want to put it on the front end, the result is a matter of fact that audio for commercials on some channels is twice the power level at the tv output as what it is during the program. I would bet good money if I put an O'scope on the audio signal in my tv set that I'd see the peak voltage significantly higher during the commercial than during the show. Also having worked on commercial broadcast equipment, I've seen plenty of examples where the rf signal has been overmodded and in fact has bled into adjacent channel space. How does that happen if the audio input isn't being overdriven?