OpenViBE forum

I was running the mi_csp scenario when i got the error

".... CSP Spatial Filter> Bad matrix size _ Filter has 11 channels but data needs 16 channels"

I 'am using emotiv epoch+ and I think I know where the problem arises.

in the scenario mi-csp-1-acquisition . using signal display I figured out the "Acquisition client" box gives all the 16 channels to the "generic stream reader" box but only 11 is read in mi-csp-2 .

can sb help ?


edit

ok i think the problem is that in the defaul example in "mi-csp-1-acquisition" the address for the writer is :

${Player_ScenarioDirectory}/signals/motor-imagery-csp-1-acquisition-[$core{date}-$core{time}].ov

BUT in the "mi-csp-2" the address for the reader is :

"${Path_Samples}/signals/bci-motor-imagery.ov"

so I changed both of them to :

${Player_ScenarioDirectory}/signals/motor-imagery-csp-1-acquisition-[$core{date}-$core{time}].ov

but then scenario 2 wouldn't end even after fast forwarding . so I thought that it might be becuse all the last recordings were concatenated to the same file (I tried it with the other address same result!) so I decided to make a new .ov file of my own but how can I do that?

Hi,

$core{date} and $core{time} variables are computed at runtime.

It means that the value in "mi-csp-2" will never be the same as in "mi-csp-1".

You should manually replace the name of the file in "mi-csp-2" with the actual name of the file created by "mi-csp-1".

It should look like "${Player_ScenarioDirectory}/signals/motor-imagery-csp-1-acquisition-[20170524-1704].ov" or "c:/path/to/file/motor-imagery-csp-1-acquisition-[20170524-1704].ov"

Hope that helps !

thanks for the reply.

1. as you said I should manually change the address to the new file created each time.

2. I was not paying attention that in OV addresses are separated with '/' instead of the Windows default '\'!