[Eeglablist] Triggers issues

[Alex Lepauvre]

Dear all,

Thanks for you quick and helpful answer. As you mentioned, we did not reset
the triggers to 0 which led to this issue. However, as the response latency
is of no interest to us, the pop_bdfread worked perfectly.

Again thanks a lot for your help.

Best

Alex

On Wed, 17 Oct 2018, 14:27 Andreas Widmann, <widmann at uni-leipzig.de> wrote:

> Hi Alex,
>
> I had a look at the data. Phillip's diagnosis is correct. You indeed have
> to reset the trigger channel to zero (or one; depending on TTL logic) after
> a certain delay. The delay is system specific. For BioSemi amplifiers one
> to two samples should be fine. You actually do reset the trigger channel
> but only some time before the next trial (~750ms; therefore the stimulus
> triggers are unaffected). You should do that directly after stimulus and
> response trigger, respectively.
>
> The problem is complicated due to either your stimulation system being
> faster for rising TTL edges than for falling or the BioSemi amplifier
> detects rising edges TTL edges faster than falling. Therefore you sometimes
> get overlapping stimulus and response triggers (actually also resetting the
> non-active trigger bits to zero) by a single sample introducing the
> spurious 22, 42, etc. triggers.
>
> For new recordings you should correct your experiment logic. For existing
> recordings you may use my BioSemi importer you find here:
> https://github.com/widmann/erptools/blob/master/pop_bdfread.m
> This gives you the same results as with BrainVision Analyzer (no missing
> response triggers as with biosig but also including the spurious 22, 42
> triggers; your description was misleading with respect to the latter). If
> response trigger latency is irrelevant you may just ignore the spurious
> events. If response trigger latency matters you have to additionally
> correct all response trigger latencies *if they were directly preceded by a
> spurious event* by minus one sample.
>
> Hope this helps! Best,
> Andreas
>
> > Am 17.10.2018 um 11:31 schrieb Phillip Alday <phillip.alday at mpi.nl>:
> >
> > What software are you using to send the triggers? I've seen this type of
> > thing happen when the trigger channels aren't reset to zero after being
> > sent. In the VMRK file, this should also be somewhat apparent because
> > the duration field will be a non-trivial number.
> >
> > Phillip
> >
> > On 10/16/2018 05:38 PM, Andreas Widmann wrote:
> >> Hi Alex,
> >>
> >> the raw data are in BrainVision file format? Which function/plugin do
> you use for import? In case bva-io I will have a look into the issue if you
> provide (Dropbox etc.) an example file (raw data) showing the problem.
> >>
> >> Best,
> >> Andreas
> >>
> >>> Am 16.10.2018 um 14:39 schrieb Alex Lepauvre <alex.lepauvre at gmail.com
> >:
> >>>
> >>> Dear All,
> >>>
> >>> We are currently having an issue with the response triggers of our
> experiment. It is design as followed: the participants are presented to a
> stimuli, to which they have to answer, during which we recorded brain
> activity using 64EEG cap system. On answer, a triggers is sent (S1 for
> right, S2 for wrong). We therefore theoretically have 2 triggers per trial:
> a stimulus onset trigger (S10, S20, S30 depending on condition) and an
> answer trigger. The time between stimulus onset and answer varies, as
> participant can answer within five seconds after stimulus onset. The issue
> is that when importing the data in EEGlab, the response triggers get added
> to the onset triggers. We therefore don't have any S1 or S2 triggers, but
> only S12, S22... Moreover, some of the answer triggers disappear. We should
> indeed have an answer trigger following each onset trigger, which is not
> systematically the case. However, we processed the data using brain vision
> analyzer, but the problem did not appear with this software (we still have
> the S1 and S2 triggers, nor any missing triggers inbetween), despite using
> exactly the same raw data. The problem seems to be due to the importation
> of the data in EEGlab, as the problem is already present following
> importation.
> >>> We could not figure out where the problem lies, and we are hoping you
> could help us figure it out, as this could potentially impair our
> statistical analysis.
> >>>
> >>> Thanks in advance for your help,
> >>>
> >>> Best
> >>>
> >>> Alex Lepauvre
> >>> _______________________________________________
> >>> 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/20181022/4fb797d7/attachment.html>