Spatial Filter Bug

Known Issues and Problems with BCI2000
Locked
dsteines
Posts: 16
Joined: 23 Jan 2010, 17:19

Spatial Filter Bug

Post by dsteines » 06 Feb 2010, 00:40

Hello,

I was doing some basic DC testing of each module in my system to find computational errors. I am using the Signal Generator at 0Hz with a 1 mV amplitude signal on each of 16 channels. I monitor the voltages at the output of each module using the visualization windows.

I instantiated a 4 x 16 spatial filter, the first row of which was all 1's, the second row all 2's, the third row all 3's, and the fourth row all 4's. I am transmitting data in blocks of 16 elements @ 1mV DC. So, at the output of the spatial filter I expected to see 4 channels with ch1 @ 16, ch2 @ 32, ch3 @ 48, and ch4 @ 64 mV.

However, all of the outputs were identical at something around 40-50mV.

After playing around with it, I realized that the rows are not the filters, but there are concatenated matrices which are the filters. So in my case, the first four columns map to channel 1, the next four to channel 2, and so on.

Is this how that component was designed?

I should note that I am using V3.0 and compiling on VS 2008.

Thanks,

David

jawilson
Posts: 109
Joined: 28 Feb 2005, 16:31

Post by jawilson » 09 Feb 2010, 10:57

Thank you for reporting this. The spatial filter matrix was essentially transposed in software, which resulted in the behavior you saw. Update to the latest SVN, and it is fixed there.
Sorry for the bug!
Adam

Locked

Who is online

Users browsing this forum: No registered users and 1 guest