From: J=C3=B3n Fr=C3=ADmann 

To: psnlist 
Sent: Mon, 26 Nov 2012 14:13
Subject: Period extensions in WinSDR


Hello all,

How is it best to use the period extensions in WinSDR with a 4.5Hz
geophone? I am currently testing it on my Bov geophone station.


Hi J=C3=B3n,=20

    You set the extension period to x10 maximum. You need to have quite a h=
igh noise amplitude signal.=20
The signal amplitude at 0.5 Hz is a factor of 100 down on the signal at 4.5=
 Hz, so you need say +/-5 counts,=20
minimum at 0.5 Hz to be able to reconstruct the full signal. The full signa=
l noise amplitude will likely be several=20
hundred counts. Check the relative amplitudes of the P and S waves. Can you=
 get an actual data trace of a=20
broad band signal to compare the true relative amplitudes of a few quake si=
gnals of widely differing overall=20
amplitudes ? One problem with digital reconstruction of the signal is noise=
 - the other one is the step size of=20
digitiser. If there is no signal there, or only noise, what you see will no=
t be a true signal. This is the=20
disadvantage of using numerical reconstruction as opposed to a low frequenc=
y 'bass boost' 1/f^2 type of=20
compensating amplifier.=20

    Regards,=20

    Chris Chapman=20

=20
=20
=20
=20

































From: J=C3=B3n Fr=C3=ADmann=
 <jonfr@.........>















To: psnlist <psnlist@..............>








Sent: Mon, 26 Nov 2012 14:13








Subject: Period extensions in WinSDR

































Hello all,

How is it best to use the period extensions in WinSDR with a 4.5Hz
geophone? I am currently testing it on my Bov geophone station.


Hi J=C3=B3n, 


    You set the extension period to x10 maximum. Yo=
u need to have quite a high noise amplitude signal. 
The signal amplitude at 0.=
5 Hz is a factor of 100 down on the signal at 4.5 Hz, so you need say +/-5 =
counts, 

minimum at 0.5 Hz to be able to reconstruct the full signal. The full signa=
l noise amplitude will likely be several 

hundred counts. Check the relative amplitudes of the P and S waves. Can you=
 get an actual data trace of a 

broad band signal to compare the true relative amplitudes of a few quake si=
gnals of widely differing overall 

amplitudes ? One problem with digital reconstruction of the signal is noise=
 - the other one is the step size of 

digitiser. If there is no signal there, or only noise, what you see will no=
t be a true signal. This is the 

disadvantage of using numerical reconstruction as opposed to a low frequenc=
y 'bass boost' 1/f^2 type of 

compensating amplifier. 



    Regards, 



    Chris Chapman