I understand why there are peaks in integer multiples of the fundamental in the Fourier transform. But why is there often a peak in half the fundamental?

I am testing for the guitar specifically. I am doing an STFT on an guitar audio monophonic signal and then get the 4 strongest frequencies in each note. I always get the fundamental in those 4 and of course some multiples of it but also half of it.

  • $\begingroup$ At the risk of sounding like a total dofus: you have taken into account that guitar is a transposing instrument, sounding one octave lower than written? $\endgroup$ – user28833 May 31 '17 at 15:05

When I play A3 (220Hz) in my guitar, the fifth string which is A2 (110Hz) also vibrates a bit: it is what is called Sympathetic resonance. Besides other non-linear effects, this could be the case.

  • $\begingroup$ If I check not immediately after the onset but 10 or 15 frames later, then I actually stop getting that half frequency. Could that be the case? $\endgroup$ – pavlos163 May 30 '17 at 16:43
  • $\begingroup$ Do you see a continuous (progressive) attenuation of that half-harmonic as time goes by? $\endgroup$ – oxuf May 30 '17 at 16:48
  • $\begingroup$ Yes. At a faster rate than other harmonics. $\endgroup$ – pavlos163 May 30 '17 at 17:02
  • $\begingroup$ Most probably something in the path from string being picked to microphone is in resonance. And that "something" (probably a lower string, but it could be the body of the guitar for example... or both) has a big impedance at that half-harmonic frequency, that is why it gets attenuated faster. $\endgroup$ – oxuf May 30 '17 at 17:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.