scaling factors

Post Reply
tw3
Posts: 1
Joined: Tue Jun 22, 2010 11:39 am
Location: UK

scaling factors

Post by tw3 »

I compared the data obtained via Matlab/EEGLAB from the BDF files and the data coming from the "Downsample.vi" for the EXi channels and found a difference by factor 10000 (probably 8196 ?).

What is the exact reason for that and where does the scaling happen in the full Actiview version ? I need comparable data for the investigations I have done offline and which I will do online using modified Actiview.

I ve found a scaling by 8196 caused by the variable ElecGain, but that appears to me as if it only affects the displaying. Is there sth. similar affecting the savings as well ?

Coen
Site Admin
Posts: 1124
Joined: Fri Mar 26, 2004 7:00 pm
Location: Amsterdam, Netherlands
Contact:

Post by Coen »

The 32-bit data words in the output array from Downsample.vi have a scaling of LSB = 1/8192 uV (1/32 * 1/256)

The 24-bit data words in the BDF file have an scaling of LSB = 1/32 uV

The Elecgain variable changes both the above LSB values for a few special versions of the ActiveTwo with passive electrodes. With standard systems with active elctrodes, Elecgain = 0 (set by the CFG file), and the above LSB values apply.
The above LSB values are not affected by the selected display gain in ActiView.
The above LSB values are valid for the active electrode channels, some of the sensor channels have different scaling (see BDF header).

Best regards, Coen (BioSemi)

Post Reply