[BC] digital delays revisited

[Robert Orban]

At 09:47 PM 12/8/2005, you wrote:
>From: "Richard Fry" <rfry at adams.net>
>Subject: [BC] digital delays revisited
>To: "Broadcast List" <broadcast at radiolists.net>
>Message-ID: <001f01c5fc32$e4860580$5765e242 at Insp1000>
>Content-Type: text/plain; format=flowed; charset="iso-8859-1";
>         reply-type=original
>
>Rich Wood:
> >I was ready to hear 8 second differences between
> >analog and digital. I didn't.
>_________________
>
>Digital is not delayed with respect to analog.  That would not work when
>"blending" from one to the other in receivers, as a function of digital
>coverage.  So both analog and digital are delayed ~8 seconds with respect
>to GMT (plus the time zone offset).

I would expand on this explanation by saying that the analog is delayed 8 
seconds at the transmitter and the digital is delayed 8 seconds (including 
codec delays) in the radio. So the two signals are both synchronous by the 
time the listener hears them. However, they are asynchronous when they are 
propagating, so an RF dropout will affect a different time chunk of the 
digital than it does the analog and there is a better likelihood that at 
least one channel (FM or digital) will be usable at all times in the radio. 
Technically, this is called "time diversity" and can be combined with 
"spatial diversity" (spaced antennas with automatic switching between them 
depending on which antenna is receiving a better signal) to improve 
robustness even further.

Bob Orban