[FieldTrip] Automatic artifact rejection

[Porada Danja]

Hi,

I want to use automatic artifact detection to clean my MEG-data.
Does it make sense to clean the data visually first, remove very odd trials (like very huge slow waveforms which were caused by a passing car) and then apply the ft_artifact_zvalue function? 

I went step by step through the ft_artifact_zvalue function using the same parameters as proposed in the "automatic artifact rejection" tutorial for jump artifacts. It seems to me that the function doesn't use the parameter "absdiff" which is specified in cfg.artfctdef.zvalue.absdiff = 'yes'. Is this right? I am using the fieldtrip-20111115-version.

Best,
Danja