[Eeglablist] FIR FFT

[Andreas Widmann]

Just in case your question referred to the EEGLAB legacy "Basic FIR filter" option "Use (sharper) FFT linear filter instead of FIR filtering" (corresponding to usefft option with pop_eegfilt() on the command line), ignore otherwise:

This particular option does NOT refer to frequency domain filtering as described by Carlos. This option was introduced to allow filtering without MATLAB Signal Processing toolbox. It will convert a rectangular (!) window in the frequency domain to time domain FIR coefficients with an inverse FFT. As a rectangular window is unsuitable for this purpose the resulting filter will introduce considerable filter ringing artifacts into the data. This option should only be used as a method of last resort.

Preferably update to a recent EEGLAB version and use the "Basic FIR filter (new)" which does not depend on Signal Processing toolbox but the firfilt plugin.

Hope this helps,
Andreas

Am 06.03.2013 um 16:54 schrieb Carlos Andrés Guerrero M. <cguerrero76 at yahoo.com>:

> Dear Carlos
> FFT filters are basically FIR filters with the difference that filtering is not done in the time domain, is done in frequency domain. This filter response is much faster if the FIR filter has a high order (large number of coefficients).
> Here, more explanations about that:
> http://www.qsl.net/d/dl4yhf///////speclab/filters.htm
> Best
> Carlos
> ULB-Erasme 
> 
> De: Carlos CEVALLOS <ccevallo at ulb.ac.be>
> Para: eeglablist at sccn.ucsd.edu 
> Enviado: Martes, 5 de marzo, 2013 9:21 A.M.
> Asunto: [Eeglablist] FIR FFT
> 
> Good Day
>  
> I am going crazy about this, and I can not figure out what would be the main difference between using the FIR filtering or the FFT alternative, is it just computation resources or is there anything else?
>  
> The question may be dumb but I have been trying to answer this by myself and can not come across a conclusive nor coherent answer. Thanks in advance, and looking forward for your replies
>  
> Best
>  
> Carlos CEVALLOS
> Laboratory of Neurophysiology and Movement Biomechanics Faculty of Movement Science (FSM, N.5.109) Université Libre de Bruxelles Campus ERASME - CP 640 808, Route de Lennik
> 1070 ANDERLECHT
> Belgium
> Phone : +32/2/555.33.93
> Fax : +32/2/555.69.63
> Website : http://www.cheron.be
>  
> 
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu
> 
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu