[Eeglablist] pop_newtimef vs newtimef

Arnaud Delorme arno at salk.edu
Mon Oct 1 00:22:37 PDT 2007


Dear Ilana,

this is because the meaning of padratio has changed between the "timef" 
and "newtimef" functions for wavelets (not for FFT). The padratio for 
wavelet for timef was not consistent with the padratio for FFT so this 
was corrected in newtimef. The problem is that for the same padratio 
(zero padding), newtimef computes an image with has four times more 
resolution in frequency (and thus is four times slower). Here is another 
message containing the description of other minor differences between 
these functions.

http://sccn.ucsd.edu/pipermail/eeglabnews/2007/000033.html

Best,

Arno

Ilana Hairston wrote:
> HI there,
> On my computer, running newtimef (e.g., [ersp, itc....]=newtimef(...) 
> is much slower than pop_newtimef , with the same parameters, whether 
> from the GUI or command line and saving the output. 
> Is that normal? 
> and if so, is there a way to fix this.
>
> thx,
> ilana




More information about the eeglablist mailing list