[BC] Arbitron

[Warren Shulz]

You can connected a PC to the RS-232 port of the PPM decoder and get a line entry that repeats over and over.  Another issue is clock.  Seems the down load will track and correct clock error.  We never did get he external clock to work and it was not important from what I was told.
 
That is an important point - dead air and no PPM.  So if you loose audio no encode action. Ah but then the PPM decoder also calls for help with dead air.,  You may want to have a time delay on that status channel to stop false calls.
 
I found it interesting the decoder ( hands of listener) has a motion detector.  Lack of movement an d it stops logging.  I suggested clipping the PPM unit the dog's collar!
 
Warren Shulz
QLS CGO

________________________________

 on behalf of Gary Blau

Jim:

Some basic info from the horse itself:
http://www.arbitron.com/portable_people_meters/home.htm

The encoding is a narrow band psycho acoustically masked data burst
imbedded into the audio, containing the encoder ID and a time stamp.
The encoder needs some surrounding audio to mask the data, or it won't
generate anything.  So, if you have dead air, you also have no PPM, and
thereby no reported listening.

It's also level dependent to some degree, but finding out much detail
about that is impossible beyond vague generalities.  You probably want
some AGC in front, especially with a less than balls to the wall pop
music format.

The 'monitors' we have only give us an idiot light and contact closure
in the event of a failure.  For that to occur, an entire three minute
window of zero detected encoding has to pass.  The bursts are normally
about 5 seconds apart.  If you have only one good 5 second burst in
three minutes, your monitor tells you you're good:
http://www.w3am.com/PPM.pdf
There is currently no way of telling if you're sending out any of those
individual bursts, granularity-wise.  You just have to 'assume' that
everything is fine as long as your idiot light is green.

It's very different than relying on human recall.  There, if someone
thinks they listened, or wants to admit to listening, you get credit
even if they really didn't.
With PPM if they aren't exposed to your audio you absolutely won't get
credit, but if they are and there's some other factor going on we don't
know all about yet, you may still not get credit.
The upside may be that if they fall asleep in the doc's waiting room the
soft AC station gets huge numbers.

They're not measuring the same things.

g

Jim Tonne wrote:
> Dana, Blaine and all:
>
> I am going to reveal a  *gross*  level of ignorance
> on a subject here but please bear with me!
>
> I have been following the PPM thread for quite a
> while now.
>
> Just how does this PPM "encoding" work?
>
> I cannot believe that little "beeps" or something
> that is easily heard would be transmitted.  Surely
> that would drive listeners bonkers.
>
> I can see perhaps looking at the envelope of the
> audio in the home and comparing it with the
> envelope of the transmitted signal.  Look at a
> transceiver for 30 seconds and compare.
>
> Perhaps I am not alone in my ignorance of the
> subject . . .
>
> - JimT