[FieldTrip] 'powandcsd' on short time window

[Eelke Spaak]

Dear Nenad,

Unfortunately the limit to the frequency resolution is fundamental and not
practical. It is impossible to get a better frequency resolution than 20 Hz
with only 50ms of data. (It might help to think of this as analogous to
Heisenberg's uncertainty principle in physics.)

Why do you need a better frequency resolution? Perhaps there are
alternatives to what you are trying.

Best,
Eelke
On Nov 8, 2013 7:44 AM, "Nenad Polomac" <polomacnenad at gmail.com> wrote:

> Dear Fieldtrip users,
>
> I have one problem and I hope that somebody has idea. For my source and
> connectivity analysis I need cross spectrum matrix at 40Hz and for time
> window of 50 ms(the effect is evoked gamma and it last less then 50 ms).
> The problem is that for this very narrow window the best I can get is the
> frequency resolution of 20Hz and that is far to low. I am trying now
> different padding options but it seems it doesn't help. If I band pass
> filter(35-45Hz) data before fft then result is as expected, but without
> filtering I pickup wide range of frequencies(probably from 20-60Hz) even if
> I specify cfg.foilim=[40 40]. Furthermore, all this stuff is included in my
> cross spectrum and brought into the further calculations.
> my code:
>         cfg=[];
>         cfg.toilim    = [0.03 0.08];
>         data_redef= ft_redefinetrial(cfg, data);
>
>         cfg = [];
>         cfg.trials = 'all';
>         cfg.keeptrials = 'no';
>         cfg.channel    = 'MEG';
>         cfg.method    = 'mtmfft';
>         cfg.output    = 'powandcsd';
>         cfg.tapsmofrq = 1;
>         cfg.foilim    = [40 40];
>         cfg.taper      =  'hanning';
>         cfg.pad        =2;
>         cfg.padtype='zero';
>         gamma = ft_freqanalysis(cfg,  data_redef);
>
> Thank you in advance!
> Cheers,
>
> Nenad
>
> _______________________________________________
> fieldtrip mailing list
> fieldtrip at donders.ru.nl
> http://mailman.science.ru.nl/mailman/listinfo/fieldtrip
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20131110/6d94d66c/attachment-0002.html>