Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Fourier Transforms is Inflicting Pain!

  1. Jun 25, 2012 #1
    So I'm doing an internship this summer and one of the things I have to be acquainted with is Fourier Transforms.

    My adviser gave me a simple example with that of sin(x). He said that what the Fourier Transform does is transform a time domain signal into one of a frequency signal. Essentially, he added, the peak of a Fourier Transform is that of the greatest match of the frequency to the function sin(x)--where points things that are closest to x-axis are the worst match.

    But since I've started this assignment earlier today, I've noticed that Mathematica does not plot Fourier Transform of functions such as sin(x) that have constant frequencies and periodicity for all x. Mathematica plots things such as sin(x)/x where the frequency isn't constant throughout the x-axis. Upon doing some research on FT, I suspect this is because Fourier Transform works for things that vary, not for functions that have constant periodicity. But I don't know any better so I'm asking for your input.

    It would also be helpful if someone can clarify the whole concept of Fourier Transforms and how it relates to statistics and data analysis.
     
  2. jcsd
  3. Jun 25, 2012 #2
    It's probably not plotting the Fourier transform of [itex]\sin x[/itex] because only two frequencies contribute (1 and -1), and everywhere else the transform is zero.

    You can imagine the Fourier transform as decomposing a signal into a sum of many sine and cosine functions. The Fourier amplitude [itex]F(\omega)[/itex] tells you what the amplitude of that frequency sine or cosine wave contributes to the signal.
     
  4. Jun 25, 2012 #3
    I had trouble running other Fourier Transforms as well, but I'm not sure why.

    I've attached a file with some of the Fourier Transforms that I attempted to plot. It would be greatly appreciated if you can take a look at it!
     
  5. Jun 26, 2012 #4
    sin(x)/x is called sinc x. Its different than sine. Sinc(x) is most important function in communication systems and filters. because theoretical impulse response for "best" filter is a sinc.
     
  6. Jun 26, 2012 #5
    Hm interesting. So does sin(x) give itself to fourier transform methods or am I doing something wrong?
     
  7. Jun 26, 2012 #6

    Mute

    User Avatar
    Homework Helper

    How are you trying to determine the fourier transforms in Mathematica?

    sin(x) does indeed have a fourier transform, but it is a special kind of fourier transform. For example, if you try to calculate the fourier transform in mathematica via the command

    Code (Text):

    Integrate[Exp[-I*2*Pi*k*x]*Sin[x],{x,-Infinity,Infinity}]
     
    you will not get a result. Mathematica will tell you the integral does not converge. This is because in the regular integral sense the integral indeed does not converge. This is because both Sin(x) and Exp(ikx) oscillate as x tends to infinity, but neither decays, so the integral does not converge. However, if you try to transform via

    Code (Text):

    FourierTransform[Sin[x],x,k]
     
    I believe you should get a result that involves the Dirac Delta function. Did you try this way?

    Now, if you tried the same two commands with Sin[x]/x, both should give you the same result, I think. Why is this? It's because Sin[x]/x decays as x -> Infinity, so the integral does converge in the usual sense.
     
  8. Jun 27, 2012 #7
    Even my internship is on Fourier Transforms.

    A part of the work is to find the frequency component of the signal which has the maximum amplitude (power density). Right now, I'm performing an fft in Matlab and searching for the maximum peak.

    The fft is done to do just this and nothing else. An fft of about 1024 samples would consume a lot of time and would drain battery power.

    Is there any simpler way to do this?

    Thanks,
    Chaitanya.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Fourier Transforms is Inflicting Pain!
  1. Fourier transforms (Replies: 2)

  2. Fourier transformation (Replies: 0)

  3. Fourier transform (Replies: 2)

  4. Fourier transforms (Replies: 1)

  5. Fourier transform (Replies: 1)

Loading...