Amplitudes values multiplied by unknown constant (Y axis skewed)

Bugs
3

Hi Francois and Raymundo,

Thank you for helping Shady.

A bug with the filtering does appear to be involved.
The ERP pipeline was as follows:
ICA to remove blinks
Apply SSP projectors
Bandpass 1-50 Hz
Import into database
remove bad epochs
Average all epochs

Here is the file history of the resulting ERP (yes, I know that the number of epochs is few)

Path: D:\BrainstormDB_D\Existenz\data
Name: P01/Stim1/data_Stim1_average_210825_0929.mat

  • 25-Aug-2021 09:29:24 | average | Average: By subject
  • 25-Aug-2021 09:29:24 | average | History of the first file:
  • 25-Aug-2021 09:27:52 | import | - Import from: D:\BrainstormDB_D\Existenz\data\P01@rawP01_Combined_Filtered_clean_band_02\P01_Combined_Filtered_clean_band_02.bst (Event)
  • 25-Aug-2021 09:27:52 | import | - Apply SSP projectors
  • 25-Aug-2021 09:29:24 | average | List of averaged files:
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial001_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial002_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial003_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial004_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial005_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial006_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial007_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial008_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial009_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial010_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial011_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial012_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial013_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial014_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial015_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial016_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial017_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial018_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial019_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial020_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial021_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial022_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial023_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial024_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial025_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial027_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial028_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial029_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial030_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial031_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial032_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial033_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial034_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial035_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial036_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial037_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial038_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial039_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial040_02.mat
  • 25-Aug-2021 09:29:24 | average | - P01/Stim1/data_Stim1_trial041_02.mat
    (For info on the SSP/ICA projectors: read the history of the channel file)

However, it appears that the band pass filter was not hard encoded in the file. Because, as you suggested, the resulting ERP appears to not be high-pass filtered.

Here is what the resulting ERP looks like without the visualization high pass:

ERP_noViz_highpass
ERP_noViz_highpass1276×568 80.1 KB

Here is what the resulting ERP looks like with the visualization high pass at 1Hz:

ERP_withViz_highpass_1Hz
ERP_withViz_highpass_1Hz1276×568 85.8 KB

It seems quite clear that filtration of the cleaned data did not work as it was supposed to. I don't think this should happen. However, I am going to invert the steps of the pipeline to bandpass filter first, and then clean. In the meantime, I would appreciate your further thoughts on this issue when you have a moment.

Best,

Jared