Hello.
I have the BMI Emotiv Epoc in Research version. I am starting with the capture of raw EEG signals.
The first version of the design was:
- the acquisition box and
- then the writing box in .csv.
The .csv file contained data, but it caught my attention that they were only positive values.
I contacted the support of emotiv and they replied that I should eliminate the background noise. I was recommended to apply a high pass filter to the data, then there will be positive and negative values centered at zero.
In case anyone is interested, they recommended the following link: https://emotiv.gitbook.io/emotivpro-v2- ... /dc-offset.
After reading that information, I am trying to make the design of the second version, which contains:
- an acquisition box,
- a temporal filter box
- filter method: Butterworth
- filter type: High Pass
- filter order: 1
- low and high cut frequency: when selecting the filter type,
autocompleted with 1 and 40 hz respectively.
- band pass db: 0,5 (default)
- writing box in .csv
I finally got positive and negative values!
But my question arises, if anyone has experience with Emotiv Epoc and OpenVibe, to see if I am really using the correct boxes and the necessary steps to obtain real EEG data (preprocessing). Or if they recommend some extra step to improve data information.
Thank you!.
EEG offset Emotiv Epoc
Re: EEG offset Emotiv Epoc
Hi,
First It's good to save raw signals, next before some signal processing is really common to use a temporal filter. The default option can be stay : butterworth, band pass, order 4, Low Cut 1, High Cut 40, Ripple 0.5. You have a 50Hz (or 60Hz dpeend on your country) noise because it's the frequency of you electric power.
Second, I think you must avoïd to save all in csv it's good to read signal but it take more space than gdf or ov files. gdf is common with other applictions. ov is openvibe format. In csv save the result of your processing (example energy of the signal, classification result...) or something like that to analyse your datas after.
Thibaut
First It's good to save raw signals, next before some signal processing is really common to use a temporal filter. The default option can be stay : butterworth, band pass, order 4, Low Cut 1, High Cut 40, Ripple 0.5. You have a 50Hz (or 60Hz dpeend on your country) noise because it's the frequency of you electric power.
Second, I think you must avoïd to save all in csv it's good to read signal but it take more space than gdf or ov files. gdf is common with other applictions. ov is openvibe format. In csv save the result of your processing (example energy of the signal, classification result...) or something like that to analyse your datas after.
Thibaut