Hi...
i'm using matlab to calculate de PSD using Welch method from a EEG record.
I know that the frequencies goes from 0.5 to 15 or 30 Hz (max), as seen in the
following image:
http://images.elektroda.net/99_1279418678.jpg
When i calculate in matlab the psd, i'm getting this:
http://images.elektroda.net/99_1279418795.jpg
My signal is sampled at 256 samples/second, 16 bit resolution.
What am i doing wrong?? The "shape" of the spectrum seems right but the
Frequency and the power frequency doesn't..
In matlab i'm typing:
h=spectrum.welch;
Hpsd=psd(h,s(:,1),'NormalizedFrequency',false); % s contains the
signal
plot(Hpsd) %With this i get the last plot
It is really urgent... i need to deliver this tomorrow morning....
Thanks a lot!!
PSD, correct shape, wrong frequency and power frequency axes
Started by ●July 19, 2010
Reply by ●July 21, 20102010-07-21
Actually both plots are similar estimates of the same signal.
As for the last plot it looks like the horizontal axis is the sample number
instead of the frequency, and you have just labeled it "Frequency (mHz)"
right? mHz? ??
If the above assumption is correct you have to create your own frequency
axis.
f=(0:N-1)*(fs/N);
where fs is the sampling frequency and N the signal length, then throw away
half of this vector to adjust to the PSD length.
Arnaldo
From: m... [mailto:m...] On Behalf Of
g...@hotmail.com
Sent: 18 July 2010 03:31
To: m...
Subject: [matlab] PSD, correct shape, wrong frequency and power frequency
axes
Hi...
i'm using matlab to calculate de PSD using Welch method from a EEG record. I
know that the frequencies goes from 0.5 to 15 or 30 Hz (max), as seen in the
following image:
http://images.elektroda.net/99_1279418678.jpg
When i calculate in matlab the psd, i'm getting this:
http://images.elektroda.net/99_1279418795.jpg
My signal is sampled at 256 samples/second, 16 bit resolution.
What am i doing wrong?? The "shape" of the spectrum seems right but the
Frequency and the power frequency doesn't..
In matlab i'm typing:
h=spectrum.welch;
Hpsd=psd(h,s(:,1),'NormalizedFrequency',false); % s contains the signal
plot(Hpsd) %With this i get the last plot
It is really urgent... i need to deliver this tomorrow morning....
Thanks a lot!!
As for the last plot it looks like the horizontal axis is the sample number
instead of the frequency, and you have just labeled it "Frequency (mHz)"
right? mHz? ??
If the above assumption is correct you have to create your own frequency
axis.
f=(0:N-1)*(fs/N);
where fs is the sampling frequency and N the signal length, then throw away
half of this vector to adjust to the PSD length.
Arnaldo
From: m... [mailto:m...] On Behalf Of
g...@hotmail.com
Sent: 18 July 2010 03:31
To: m...
Subject: [matlab] PSD, correct shape, wrong frequency and power frequency
axes
Hi...
i'm using matlab to calculate de PSD using Welch method from a EEG record. I
know that the frequencies goes from 0.5 to 15 or 30 Hz (max), as seen in the
following image:
http://images.elektroda.net/99_1279418678.jpg
When i calculate in matlab the psd, i'm getting this:
http://images.elektroda.net/99_1279418795.jpg
My signal is sampled at 256 samples/second, 16 bit resolution.
What am i doing wrong?? The "shape" of the spectrum seems right but the
Frequency and the power frequency doesn't..
In matlab i'm typing:
h=spectrum.welch;
Hpsd=psd(h,s(:,1),'NormalizedFrequency',false); % s contains the signal
plot(Hpsd) %With this i get the last plot
It is really urgent... i need to deliver this tomorrow morning....
Thanks a lot!!