[Eeglablist] causal bandpassfilter with very low cutoff

Fri Aug 10 01:13:23 PDT 2012

Dear all,

I would like to ask you, which are the parameters you take under
consideration to decide whether filter your data with a IIR or FIR filter.

summarizing the information regarding these filters:

IIR filter: The phase response of a IIR filter is not linear. This would
always distorts the signal, unless the data are filtered back
and forward via filtfilt Matlab function (this is of course possible only
via off-line processing)

FIR filter: The phase response of a FIR filter is always linear, thus the
group delay of a FIR filter is constant number. That means that all the
frequencies of the filtered data have been delayed by the same amount of
time (might be several seconds). Since the delayed introduced by the FIR
filter is exactly half of its length, it can be exactly compensated without
the need of use filtfilt (again this is possible on via off-line processing)

>From my point of view both these solutions work (I mean it s possible to
get to the same output both via a IIR filter using filtfilt and via a FIR
filter using filter and compensating the delay introduced by the filter).

I would like to know if and why do you prefer one of these two types of
filters.

Thanks a lot!

Davide.

On Fri, Aug 10, 2012 at 8:26 AM, Arnaud Delorme <arno at ucsd.edu> wrote:

> Hi Sophie and David,
>
> Butterworth is available in EEGLAB from the command line (eegfiltsig
> function) and from the ERPLAB plugin of EEGLAB (which does not require
> ERPLAB itself and is installed by default when you download EEGLAB).
>
> In the most recent version of EEGLAB (to be released this month), we have
> replaced the default FIR filter by Andreas Widmann optimized filter
> functions (in the firfilt plugin). This fixes several minor issues such as
> sometimes non-optimal widths for the transition bandwidth. It is also more
> flexible and has many more options for plotting the filter properties. Note
> that all scripts will remain backward compatible.
> Thanks,
>
> Arno
>
> On Aug 9, 2012, at 2:20 PM, David Acunzo wrote:
>
> Hi Sophie,
>
> To design your IIR filter, you can use the buttord function (or ellipord,
> etc. for filters that are not Butterworth).
> To avoid roundoff errors and therefore instability or NaNs, you may want
> to use the zero-pole-gain filter design (see the examples in
> http://www.mathworks.co.uk/help/toolbox/signal/ref/butter.html for
> Butterworth filters), which is, I think, not used in the IIR filters
> available in EEGlab.
>
> Hope that helps.
>
> Best,
> David
>
>
> On 8 Aug 2012, at 14:58, Sophie Herbst wrote:
>
> Hi again,
> thank you for your very helpful comments. Andreas, I was trying to
> implement your suggestion but can't quite get it to work.
> After low-pass filtering (30Hz), downsampling (128Hz), the butterworth
> filter you suggested transforms large parts of the data into NaNs.
> I think it might already go wrong with the low-pass filter. Do you have a
> suggestion what to use here? It should also be causal, so I tried both
> iirfilt and butter but apparently I was not able to put together the right
> combination of parameters.
>
> Sorry for being so ignorant but my understanding of filtering is at a very
> early stage.
> Sophie
>
>
>
>
> On Wed, Aug 8, 2012 at 12:13 PM, Andreas Widmann <widmann at uni-leipzig.de>wrote:
>
>> Hi,
>>
>> Sophie asked for a *causal* filter. IIR is an appropriate filter type for
>> this purpose. Using a linear phase FIR filter, as suggested, one will
>> introduce a long filter delay. Using a minimum-phase FIR filter one will
>> loose linear phase anyway. Thus, it is reasonable to use IIR as the filter
>> delay is usually considerably smaller than for minimum-phase FIR in this
>> type of application. (I have a slight preference for Butterworth due to its
>> good (flat) frequency response in the passband.)
>>
>> See also recent contribution on causal filtering by Rousselet:
>>
>> http://www.frontiersin.org/Perception_Science/10.3389/fpsyg.2012.00131/full
>> and my short comment (mainly focused on why one should not use the EEGLAB
>> Basic FIR filter)
>>
>> http://www.frontiersin.org/Perception_Science/10.3389/fpsyg.2012.00233/full
>>
>> Best,
>> Andreas
>>
>> Am 08.08.2012 um 10:17 schrieb Davide Baldo <davidebaldo84 at gmail.com>:
>>
>> > Hi!
>> >
>> > I don't understand why to use a IIR filter. They always cause some
>> (phase) distorsion.
>> > I suggest you to use a FIR filter. If you use Matlab, just type
>> "fdatool" (Filter design & analysis tool). Than select: High pass and FIR
>> (Equiripple). Set the Srate to 512 and the Fstop to 0.01. You did not
>> specfied which frequencies you do not want to distort. I guess you could
>> set Fpass to 0.5 Hz (all frequencies higher than 0.5 Hz won't be modified).
>> Then set Dstop to 0.0005 and Dpass to 0.01.
>> > Now you can click on Design Filter. When done...click on File -> Export
>> (It export the filter on Matlab workspace). Set Numerator to "HP_Filter"
>> (it s just a name for the filter).
>> >
>> > Now you are ready to filter your data:
>> >
>> >     HP_Delay = round( mean(grpdelay(HP_filter)) ); % a FIR filter
>> introduces a delay in the signal. you need to compensate it.
>> >
>> >     HP_Data = filter( HP_filter, 1, your_data ) ;  % Filter the data
>> >     HP_Data  = circshift(  HP_Data , [1 -HP_Delay] ); %compensate the
>> delay introduced by the HP filter
>> >
>> > I hope it helps you.
>> >
>> > Ciao!
>> >
>> > Davide.
>> >
>> > On Mon, Aug 6, 2012 at 6:38 PM, Andreas Widmann <widmann at uni-leipzig.de>
>> wrote:
>> > Hi Sophie,
>> >
>> > a 0.01 Hz highpass filter with 512 Hz sampling frequency is very
>> extreme. My personal rule of thumb is that the srate / cutoff ratio for IIR
>> highpass filtering should not be much higher than ~1000 (acknowledgement to
>> BM). The problem is increased by the EEGLAB default estimation of required
>> filter order by a very narrow transition band (defined as cutoff/3 in your
>> case; the help text in pop_iirfilt is wrong!). The extreme filter cutoff
>> with narrow transition band requires a high filter order (here 6), but the
>> resulting 6th order filter is instable.
>> >
>> > I would suggest first downsampling the data to an as low as possible
>> sampling frequency (after lowpass filtering the data to 1/4-1/5 of new
>> sampling frequency!).
>> >
>> > Then, I would suggest filtering the data with a butterworth filter at
>> 0.1 Hz cutoff frequency. Roll-off is a function of filter order (approx.
>> order times -6dB/octave).
>> > E.g.:
>> > >> [b, a] = butter(4, 0.1 / (srate / 2), 'high')
>> > for a forth order filter.
>> >
>> > Check frequency response with
>> > >> freqz(b, a, 2^14, srate)
>> > At 256 Hertz this gives a reasonable frequency response and very good
>> DC attenuation. Downsampling your data to 128 Hz you can use the same
>> filter for a 0.05 Hz highpass.
>> >
>> > Check filter stability with
>> > >> zplane(b, a)
>> > All poles should be inside the unit circle. If you test the 6th order
>> butterworth filter you will see that also this filter is instable.
>> >
>> > Causal filtering can be done easily on the command line using the
>> MATLAB built in filter function. Take care not to filter across
>> boundaries/DC offsets! Filter each segment separately.
>> >
>> > Hope this helps, best,
>> > Andreas
>> >
>> > Am 06.08.2012 um 17:44 schrieb Sophie Herbst <ksherbst at googlemail.com>:
>> >
>> > > Hi EEGlablist,
>> > >
>> > > I am trying to apply a causal bandpass filter with a very low cutoff
>> (0.01Hz) to my EEG data (continuos data with ~2,700,000 points, srate =
>> 512Hz)
>> > > by using iirfilt.m:
>> > >
>> > > iirfilt(EEG.data, EEG.srate, 0.01, 40, 0, 0, 0, [], [], 'on')
>> > >
>> > > I seem to run into similar problems as described in EEGLAB Bug #1011:
>> the default values for transition bandwidth and passband/ stopband ripple
>> do not seem to work as
>> > > the filter runs but leaves a matrix of NaNs.
>> > >
>> > > I have been playing around with values for the transition bandwidth
>> etc, but I could not get a satisfying frequency response.
>> > > Any idea why this is and which filter would be better to use?
>> > >
>> > > Thanks a lot,
>> > > Sophie
>> >
>> > _______________________________________________
>> > 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
>
>
> The University of Edinburgh is a charitable body, registered in
> Scotland, with registration number SC005336.
>
> _______________________________________________
> 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
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20120810/0750799b/attachment.html>