0

So I wanted to fix the resolution to: number of frequencies is 56 and time-localised point is 3

Hoever I cannot understand the arguments of the function scipy.signal.spectrogram:

scipy.signal.spectrogram(x, fs=1.0, window=('tukey', 0.25), nperseg=None, noverlap=None, nfft=None, detrend='constant', return_onesided=True, scaling='density', axis=-1, mode='psd')

It's here in this documentation: https://docs.scipy.org/doc/scipy-0.19.0/reference/generated/scipy.signal.spectrogram.html

What is nseg? nfft?

enter image description here

Becuase it says here that "optimal number of frequencies is 56 and the optimal time-localised point is 3"

I'm new to the concept of spectrogram. How do you adjust it's resolution? Can you give an intuitive explanation on how to generate the spectrogram? And apply it to this function in Scipy?

Chaine
  • 1,368
  • 4
  • 18
  • 37
  • 2
    what do you mean by "time-localised point is 3"? – Nils Werner May 30 '17 at 13:56
  • Please refer to the picture added – Chaine May 30 '17 at 14:14
  • 1
    The figure doesn't explain it. – Nils Werner May 30 '17 at 14:23
  • I was actually going to ask what it means because I'm new to the concept of spectrogram. How do you adjust it's resolution? – Chaine May 30 '17 at 14:27
  • 2
    @ChaineSanBuenaventura Have you tried playing with the parameters to see what effect they have on the outcome? A while ago I explained in https://stackoverflow.com/a/43537494/3005167 how to achieve different spectrogram resolutions in scipy. Maybe this helps you along. – MB-F May 31 '17 at 06:38
  • Yes I did but I don't get the reason behind the changes. Okay I will look into it. Thanks! – Chaine May 31 '17 at 07:04

0 Answers0