Comments on: DFT Magnitude/Power Spectra http://hans.fugal.net/blog/2009/04/02/dft-magnitudepower-spectra/ Counterpoint by Hans Fugal Fri, 12 Mar 2010 20:21:40 -0700 http://wordpress.org/?v=2.8.5 hourly 1 By: Josef http://hans.fugal.net/blog/2009/04/02/dft-magnitudepower-spectra/comment-page-1/#comment-2156 Josef Wed, 16 Dec 2009 10:53:28 +0000 http://hans.fugal.net/blog/?p=1129#comment-2156 my friend i have signals stored in text file (each file contains 2 signals represents the left and right hands) which method is better for me to calculate and plot the PSD for each one? your comments is highly appreciated best wishes, Josef my friend

i have signals stored in text file (each file contains 2 signals represents the left and right hands)

which method is better for me to calculate and plot the PSD for each one?

your comments is highly appreciated
best wishes,
Josef

]]> By: Hans http://hans.fugal.net/blog/2009/04/02/dft-magnitudepower-spectra/comment-page-1/#comment-1944 Hans Thu, 16 Apr 2009 14:26:56 +0000 http://hans.fugal.net/blog/?p=1129#comment-1944 That's a good point, Ely. So you're saying that the standard 20*log10 of the magnitude gives you dB power anyway - is that right? So then I should be comparing 10*log10(X) to 10*log10(X.^2)? Of course that's just a linear difference from what I had before. If we compare 20*log10(X) to 10*log10(X.^2) then of course they're the same - they're the same mathematically. That’s a good point, Ely. So you’re saying that the standard 20*log10 of the magnitude gives you dB power anyway – is that right? So then I should be comparing 10*log10(X) to 10*log10(X.^2)? Of course that’s just a linear difference from what I had before.

If we compare 20*log10(X) to 10*log10(X.^2) then of course they’re the same – they’re the same mathematically.

]]> By: Ely http://hans.fugal.net/blog/2009/04/02/dft-magnitudepower-spectra/comment-page-1/#comment-1943 Ely Thu, 16 Apr 2009 13:33:16 +0000 http://hans.fugal.net/blog/?p=1129#comment-1943 In the last bit plot(20*log10(abs(fft(x)))) plot(20*log10(abs(fft(x)).^2)) I believe these are wrong. The first one will give the the Power in dB. Second one is squared twice. See the 20 log rule in the following. http://en.wikipedia.org/wiki/Decibel In the last bit
plot(20*log10(abs(fft(x))))
plot(20*log10(abs(fft(x)).^2))
I believe these are wrong. The first one will give the the Power in dB.
Second one is squared twice.

See the 20 log rule in the following.
http://en.wikipedia.org/wiki/Decibel

]]>