[FieldTrip] Can ft_freqanalysis under certain condition ignore sampling rate information?

Ivaylo Iotchev ivaylo.iotchev at gmail.com
Tue Dec 6 10:11:06 CET 2022


Dear all,

I have been told by several more experienced users so far, that in
Fieldtrip the sampling rate need not be specified explicitly and is read
automatically in the course of other operations.

I have come to doubt this. With my colleagues we are looking for a
pronounced response in the lower frequency range, expected to vary somewhat
between 0.5 and 1.5 Hz, what we recently discovered is that this pronounced
peak shifts with the length of the segment that is analyzed, specifically
the peak is always at the frequency which is 1/segment length. Is it
possible to have accidentally made ft_freqanalysis ignore sampling rate?
Because I imagine this happening when the fft is oblivious to the sampling
rate or is there additional explanations/errors at play here? I also
considered spectral leakage, but the shift of the power peak is to
precisely corresponding to the segment-length to be that.

Thank you dearly in advance!

Dr. Ivaylo (Ivo) Iotchev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20221206/41375acd/attachment.htm>


More information about the fieldtrip mailing list