DFT in Practice - Frequency Resolution Issues

Поделиться
HTML-код
  • Опубликовано: 16 дек 2013
  • An example of how to analyse the fundamental frequency of two notes being played on a bass guitar.
    Documentation on the DFT is available at dx.doi.org/10.13140/RG.2.2.330...
    Other notes available at at pzdsp.com/docs.
    Matlab code available at dadorran.wordpress.com/2013/12...
    A pdf version of the tutorial available at dadorran.wordpress.com/2013/12...
    Download the bass guitar recording from eleceng.dit.ie/dorran/matlab/b...

Комментарии • 17

  • @MinhTran-wn1ri
    @MinhTran-wn1ri 4 года назад +1

    Great demo. I wish I'd seen this years ago in my college intro to signals class.

  • @martinfultot5756
    @martinfultot5756 4 года назад +3

    I was struggling with the exact same issue. Very clarifying and great accent.

  • @MinhTran-wn1ri
    @MinhTran-wn1ri 8 лет назад +2

    Thank you. I'm taking a DSP course now and this demo is very helpful.

  • @samioannidis2455
    @samioannidis2455 6 лет назад

    great work

  • @husseinalgusab4129
    @husseinalgusab4129 10 лет назад +1

    Very good totorial you are manifiqes David thank you very much and please continue because your method to illustrate the concepts is vrey clear and useful

  • @jackhansenguitar
    @jackhansenguitar 6 лет назад

    Fascinating! what criteria does MATLAB use to distinguish fundamental from harmonic??

  • @marksteven6824
    @marksteven6824 6 лет назад +1

    Sir will it recognize a note from vocal inputs?

  • @mmuuuuhh
    @mmuuuuhh 9 лет назад

    Could I interpolate between Samples, to increase the number of values, and therefore increase the resolution of the DFT?
    Would this work as an alternative to the zero-padding methode?

    • @ddorran
      @ddorran  9 лет назад +2

      I'm afraid this wouldn't work since when you resample you are changing the sampling interval between samples. For example say you had 1000 samples obtained using a sampling rate of 500 Hz (sampling interval =.002) then the spacing between DFT bins would be fs/N = 500/1000 = 0.5 Hz. If you interpolated by a factor of 2 then you'd have 2000 samples with a sampling interval of 0.001 (or an effective sampling rate of 1000Hz). Taking the DFT of these interpolated samples would result in a separation between bins of fs_interp/N_interp = 1000/2000 = 0.5Hz.

  • @olamideajala687
    @olamideajala687 8 лет назад

    hello, please how do I zero pad column data. I used this (y7F_zpad = [y7F zeros(1, 99000)]; ) but I keep getting error (Error using horzcat). the variable y7F is a column data with 1002 samples.

    • @ddorran
      @ddorran  8 лет назад

      +olamide ajala y7F_zpad = [y7F ; zeros( 99000,1)]; should work

  • @c33g
    @c33g 7 лет назад

    I am simulating a mass spring damper system in simulink and i am trying to calculate the frequency response of the system. I can work out all parameters other than Fs.
    Can you suggest anything?
    thanks

  • @upendrasingh9390
    @upendrasingh9390 5 лет назад

    Good explanation, Please provide us simulation of HSQC Pulse sequence in Matlab ASAP.

  • @muntu2008
    @muntu2008 10 лет назад

    I didn't understand why you took the frame size 1 sample less than the number of samples you specified. For eg. at 2:40 you took N=2000, but the 'frame' has 1 up to N-1 (1999) samples of the 'seg'. So, 'ft_mags' has 1999 bins. But then you used N=2000 instead of 1999 to calculate the frequency resolution using bin*fs/N (4:35). I think, since the size of the 'frame' is 1999, N should be =1999 instead of 2000 for calculating the frequency resolution.

    • @ddorran
      @ddorran  10 лет назад

      You are correct Pragoti Pran Bora. I made a mistake in my code. Thanks for identifying that! Apologies for any confusion caused.

    • @muntu2008
      @muntu2008 10 лет назад

      Thanks David Dorran for the reply. And no Problem. This was a minor mistake and I could identify it only because I thoroughly followed your other videos on the topic. Your tutorials are indeed excellent and give a practical insight into digital signal processing. Looking forward to more such amazing tutorials from you in the future.