[FieldTrip] change in yokogawa trigger and raw data handling

[Robert Oostenveld]

Dear Yokogawa users

Earlier this week I made a change to the handling of Yokogawa data  
that might require some small updates to your analysis scripts. Let me  
summarise it in this mail.

Although most of the meta information and header information is still  
read with the yokogawa p-files, the sqd data can now also be read with  
the sqdproject/sqdread function. As of this week the code is included  
in fieldtrip/external/sqdproject and will automatically be added to  
your path.

The consequence is that much less memory is needed if you only want to  
read a few channels (or the trigger channels). You can now use it on a  
small memory computer.

Another consequence is that data is expressed in fT and mV (if I am  
correct) and not no T and V, so please note the scaling difference.

Furthermore, I have added an option to ft_read_event/read_trigger that  
automatically detects whether you have downgoing or upgoing flanks in  
the data. It used to detect both, and then it was up to you to select  
e.g. '169_up' as the relevant one. It will now (try to) autodetect  
whether your data has downgoing or upgoing flanks and only detect the  
triggers for the approprate one. The resulting trigger code does not  
have the down/up tag any more, so you should now work with '169'  
instead.

And finally the ft_read_event function now has the 'trigindx' option  
documented (see also http://bugzilla.fcdonders.nl/show_bug.cgi?id=494#c2) 
  and has an option 'threshold' for the flank detection in the analog  
trigger channels.

best regards,
Robert

PS if you want to follow up to this mail with potential problems or  
suggestions for improvements, please use http://bugzilla.fcdonders.nl  
and file a "bug" for the yokogawa component.