FTM and spectral analysis help

    wannabeafrog
    Jul 25 2013 | 2:31 pm
    Hello,
    In the attached picture you will see the bins data outcome of an FFT process made with FTM objects (it only displays 11, but there are 512 rows), I would now like to take that Fmat and convert those values into measures of frequency domain in order to work with them, can anyone help me?.
    I would like to know if this can be done with the max FFT~ object? How can I obtain simultaneous data for each sinusoidal component, I need to keep track of several peaks at the same time.
    Please help, I've been feeling really dumb lately working with this.
    Thanks in advance!
    • Floating Point
      Jul 26 2013 | 5:25 am
      there's no attached picture, but i think the conversion is (bin number)*(1/2)*(sample rate)/(total number of bins)
      the first bin is bin number 0 (last is 511)
      to keep track of multiple peaks have a look at zsa decriptors package by emmanuel jourdan. i think there is something there that does what you want
    • wannabeafrog
      Jul 26 2013 | 2:24 pm
      First of all, thank you for your answer, I will check the descriptors out.
      And about the picture, I swear I attached it, trying again
    • Floating Point
      Jul 26 2013 | 11:47 pm
      That sort of thing (basic fft transform) can be done in max 'native' using pfft and related objects. Have a look at the tutes/helpfiles. I think the max native way of going about it is easier than ftm for starting out. If then you find you need more flexibilty then ftm may be an option. as for the picture, it might be mag and phase (but expressed in cartesian coords?), but the units are not evident to me. that info should be in the ftm help files.