To understand how your offline (digital) filters might be distorting your data, it is very useful to try filtering known, artificial waveforms.  Three files are provided here to help you with this. For more information, see Chapter 7 in An Introduction to the Event-Related Potential Technique, 2nd Edition.

fake_data.xlsx is an Excel spreadsheet that shows how you can make both artificial waveforms (e.g., impulses and square waves) and realistic ERP-like waveforms.

fake_data.txt contains the same waveforms, saved in a text file.  Each column contains the data from a different waveform, and each row contains the data from a single time point.  The leftmost column is the time of each sample in milliseconds (with a sampling rate of 200 Hz and a sample period of 5 ms).  Many data analysis packages will allow you to import this file, creating a set of waveforms.  Each waveform will appear as a different channel in the file.

fake_data.erp contains the same waveforms, saved in an ERPLAB file.  This file was created by importing the “fake_data.txt” file into ERPLAB Toolbox (which you can download at http://erpinfo.org/erplab).

10 different waveforms are included:

  • c1: This is intended to be analogous to an isolated P1 wave.
  • c2: This is intended to be analogous to an isolated N1 wave.
  • c3: This is intended to be analogous to an isolated P3 wave.
  • c1+c2+c3: This is the sum of c1, c2, and c3.
  • impulse: An impulse function has a value of 1 at time zero and a value of 0 at every other time point.
  • boxcar: This boxcar function has a value of 1 from 0 to 100 ms and a value of 0 at every other time point.
  • 60 Hz noise: This is a 60 Hz oscillation with a peak-to-peak amplitude of 0.2 µV. It is designed to simulate line frequency noise in the U.S., Canada, and many other countries (it would be 50 Hz in several other countries).
  • c1+c2+c3+60Hz: This is the sum of c1, c2, c3, and the 60 Hz noise.
  • white noise: These are just random values between ±0.1 µV.  This is designed to approximate EMG noise.
  • c1+c2+c3+white noise: This is the sum of c1, c2, c3, and the white noise.

 

Filtering Data All.png