Technical discussion about Matlab and issues related to Digital Signal Processing.
|
Hi all, this is regarding my doubt in frequeny domain filtering. is it means to say that we convert a signal x(n) into its frequency domain and we get the fourier coefficients which represent the amplitude and phase of the frequency components. then we make the fourier coefficients( which are the amplitude and phase of frequency components) of all unrequired frequencies to zero and make a inverse fourier transform. please make it clear whether my concept is right? thanks , --krishna |
|
|
|
I think frequency domain filtering usually means implementing the convolution in frequency domain using fft which is computationally efficient. Navan --- krishnam acharya <> wrote: > Hi all, > this is regarding my doubt in frequeny domain > filtering. > is it means to say that we convert a signal x(n) > into its frequency domain and we get the fourier > coefficients which represent the amplitude and phase > of the frequency components. then we make the > fourier > coefficients( which are the amplitude and phase of > frequency components) of all unrequired frequencies > to > zero and make a inverse fourier transform. > > please make it clear whether my concept is right? > thanks , > --krishna > ------------------------ Yahoo! Groups Sponsor > > _____________________________________ > /groups.php3 __________________________________ |
|
|
|
Not exactly. You must multiply with the fft of the filter, and then find the ifft if you want to impliment a filter in the frequency. All convolution operations in the time domain becomes algebraic operations in frequency domain, and vice versa. And it is not just making the unwanted frequencies zeroes, but involve some smoothing operations for avoiding Gibb's phenomenon. If you have anymore doubt regarding this, pls feel free to write to me. I will try to do my best At 08:37 AM 4/27/03, navaneetha krishnan wrote: >I think frequency domain filtering usually means >implementing the convolution in frequency domain using >fft which is computationally efficient. > >Navan > >--- krishnam acharya <> >wrote: > > Hi all, > > this is regarding my doubt in frequeny domain > > filtering. > > is it means to say that we convert a signal x(n) > > into its frequency domain and we get the fourier > > coefficients which represent the amplitude and phase > > of the frequency components. then we make the > > fourier > > coefficients( which are the amplitude and phase of > > frequency components) of all unrequired frequencies > > to > > zero and make a inverse fourier transform. > > > > please make it clear whether my concept is right? > > thanks , > > --krishna > > > > > > ------------------------ Yahoo! Groups Sponsor > > > > _____________________________________ > > /groups.php3>http://www.dsprelated.com/groups.php3 > > > > >http://docs.yahoo.com/info/terms/ > > > >__________________________________ >Yahoo! Groups Sponsor >_____________________________________ >/groups.php3>http://www.dsprelated.com/groups.php3 > >>Yahoo! Terms of Service. |