A3 5(b) info

[Scott MacLean]

Hi,

There seems to be a bit of confusion about what question 5(b) is asking
in regards to finding the time at which the reference signal was
received. The idea is as follows:

-using the correlation function, you can find the sample number at which
it seems most likely the reference tone started.
-using the sampling rate (8192 Hz) you can determine the actual time at
which the tone was received. The physical units work out as follows:

(time) = (sample) / (samples/time)

Note that the length of the synchronization tone (0.125s) doesn't come
into play here since all you are concerned with is the time of the start
of the tone.

Also be sure you are measuring the shift of the signal in the correct
direction. If you use the correlation function from the course notes, a
positive shift offset means the second signal is shifted to the right
with respect to the first signal.

Scott

[Angus Sinclair]

I think that is wrong.
The p-value or shift is the amount that the reference signal (signal 1) is
shifted to the right, to match with signal 2. I am pretty sure that is what
it says in the course notes. Did i misread that? P is always non-negative,
how would you get a negative p value when it is based on an index of the
corr vector (non-negative).

Post by Scott MacLean
Hi,
There seems to be a bit of confusion about what question 5(b) is asking
in regards to finding the time at which the reference signal was
-using the correlation function, you can find the sample number at which
it seems most likely the reference tone started.
-using the sampling rate (8192 Hz) you can determine the actual time at
(time) = (sample) / (samples/time)
Note that the length of the synchronization tone (0.125s) doesn't come
into play here since all you are concerned with is the time of the start
of the tone.
Also be sure you are measuring the shift of the signal in the correct
direction. If you use the correlation function from the course notes, a
positive shift offset means the second signal is shifted to the right
with respect to the first signal.
Scott

[Scott MacLean]

Sorry! You are right. A positive shift offset means the second signal
is shifted to the left with respect to the first signal. I had written
down the formula earlier and had the subscripts backwards. My apologies.

You could have a negative index by wrapping around the array since it is
periodic.

Scott

Post by Angus Sinclair
I think that is wrong.
The p-value or shift is the amount that the reference signal (signal 1) is
shifted to the right, to match with signal 2. I am pretty sure that is what
it says in the course notes. Did i misread that? P is always non-negative,
how would you get a negative p value when it is based on an index of the
corr vector (non-negative).

Post by Scott MacLean
Hi,
There seems to be a bit of confusion about what question 5(b) is asking
in regards to finding the time at which the reference signal was
-using the correlation function, you can find the sample number at which
it seems most likely the reference tone started.
-using the sampling rate (8192 Hz) you can determine the actual time at
(time) = (sample) / (samples/time)
Note that the length of the synchronization tone (0.125s) doesn't come
into play here since all you are concerned with is the time of the start
of the tone.
Also be sure you are measuring the shift of the signal in the correct
direction. If you use the correlation function from the course notes, a
positive shift offset means the second signal is shifted to the right
with respect to the first signal.
Scott