Page 1 of 1

Large <1Hz, no 8-13Hz

Posted: Sun Jun 05, 2011 12:19 pm
by alexp
Hi all,

I have an openEEG Modular EEG device (amplifier stage is modeeg; the digital stage has been adapted for usb etc.).
I use sintered Ag/AgCl-Electrodes. Impedace is checked and was below 15k.

I made a test acquisition with DRL and - of Ch1 connected to neutral positions (checkbone near the ear), + connected to Cz.
I find low 50 Hz noise, but large signal below 1Hz and almost no signal in the 8-13Hz range even when eyes closed.
I standard papers I read that the amplitude in the range 8-13Hz should be around 9µV.

Could anyone help me to make / check a standard recording in order to check if the hardware works as intended?

Further informations:
------------------------------
I attached the raw file in csv-reader format and a simple scenario which brings the data in the µV range and extracts/displays:
raw data
<1Hz
8-13Hz
power spectrum

The power spectrum does not display anything, but I can't find my error...

Re: Large <1Hz, no 8-13Hz

Posted: Mon Jun 06, 2011 7:25 am
by ddvlamin
Hi,

I also do not understand why it does not display the spectrum, while it does display the raw data and no error or warning is displayed in the console. However, if you convert it to gdf with the gdf writer box and change the block size to 100 (as your sampling frequency is 1000hz) it does display the spectrum. Maybe the developers can find out why it does not work with the csv reader?

To see the eyes-closed-effect (sometimes you can even see it in the raw EEG), it is better to measure over occipital regions.

Best regards,
Dieter Devlaminck

Re: Large <1Hz, no 8-13Hz

Posted: Mon Jun 06, 2011 10:01 am
by jlegeny
Hello,

we have looked at the issue and we have found several issues with the CSV reader box (which is why it is still marked as unstable). Namely, the sampling rate sent by the box is not correct which might be the source of the problem with the spectrum not being displayed.

As dieter said converting the file to GDF will make the spectrum display work.
signal display.png
signal display.png (67.85 KiB) Viewed 5017 times
As for the values, in the alpha values I can indeed see the power oscillating between -9 and 9 so I suppose your measures are correct.

We will look into the sampling rate issues shortly to see what exactly happens in here. Thank you for reporting the problem.

Cheers
Jozef

Re: Large <1Hz, no 8-13Hz

Posted: Sat Jun 18, 2011 8:06 am
by alexp
Ok, thank you.