[Eeglablist] ERSP common baseline - relation to precomputed baseline

[Maureen Ritchey]

I am trying to understand what happens when the "common baseline" option is
selected for plotting and analyzing STUDY ERSPs-- specifically, how it
interacts with any existing baseline corrections.

For instance, I have used the default baseline options when precomputing my
ERSPs, which I understand to include a condition-specific baseline
subtraction. Then when I plot the ERSPs through the GUI interface, I select
the common baseline option for the ERSP parameters (equivalent to
'subbaseline' set to 'on').

My question is: for the common baseline option, is the mean baseline
subtracted from the already baseline-corrected ERSPs? Or does the
condition-specific baseline get added back in prior to subtracting the mean
baseline?

I would have expected the latter. However, when I inspect the
removeerspbaseline subfunction in std_readersp (EEGLAB version 14.1.1b), it
appears to be doing the reverse: subtracting the original baseline and
adding the mean baseline. See below (starting with line 539 in
std_readersp):

if convert2log
       ersp{c,g} = ersp{c,g} - repmat(10*log10(erspbasetmp), [1
size(ersp{c,g},2) 1 1]) + 10*log10(tmpmeanpowbase);
else
       ersp{c,g} = ersp{c,g} - repmat(erspbasetmp, [1 size(ersp{c,g},2) 1
1]) + tmpmeanpowbase;
end;

I am not sure if this is a bug, or if I am missing some other key step in
the script that would explain how this works out.

Thanks for any advice.

Best,
Maureen
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20170729/9e119a8f/attachment.html>