Low amplitude in FFT Plot

YoshiYoshi Japan
edited October 2023 in Ganglion

I have just bought Ganglion and checked it.
In FFT Plot, amplitude is quite low. They are around or less than 2 microvolts.

Settings

  • Placement: F4 (1CH)
  • Filter: 1-50Hz
  • Impedance: less than 10k
  • Eyes: open

I use BioExplorer and Neurobit Optima. Usually, amplitudes are between 2 -15 μV.
I use regular electrodes and they work fine with BioExplorer.
I played sample data and their amplitudes are also quite low.
It seems that OpenBCI gets only 1/2 or 1/3 amplitude. Is this normal?
Do you guys get amplitude of 2 -15 μV?

My recording data is much lower than sample data as well.

I attached a console log as well.
Does anyone have a clue?

  • Operating System: Windows 10
  • GUI Version: 5.2.2 (The downloaded app)
  • Type of OpenBCI Board: Ganglion
    (I tried Ganglion firmware 2 and 3, but got the same results)

Comments

  • wjcroftwjcroft Mount Shasta, CA

    Yoshi, hi.

    So you are using the exact same touchproof electrodes / paste / prep on both systems: Ganglion and Neurobit Optima? And getting much weaker results with Ganglion. Have you tried emailing customer support (contact at openbci.com)? It's remotely possible that the touchproof adapter cable(s) (needed for Ganglion) may be malfunctioning. Try switching around the channels / cables.

    Have you tried the alpha eyes closed test, with an electrode at an occipital location?

    William

  • YoshiYoshi Japan
    edited October 2023

    Hi, William @wjcroft.
    Thank you for your comment.

    Yes, I used the exact same touchproof electrodes on both systems: Ganglion and Neurobit Optima.
    I tried diffrent channels and cables, but I still got around only 2 microvolts.
    I also tried alpha eyes closed and got 5 μV with OpenBCI, which is about half of 13 μV with BioExplorer.
    I haven't contacted the customer support yet, but I will.

    I played sample data that comes with OpenBUI GUI, "OpenBCI_GUI-v5-meditation.txt".
    In the sample data below, alpha on 10Hz is around 5 μV, which is about 1/3 of normal amplitude.
    Adults amplitude of Alpha with eyes closed is about 16 μV, according to a research below.
    https://www.isnr-jnt.org/article/view/17234

    This is nothing to do with electrodes, adapters, signal or Ganglion because playing sample data does not requre those. So, I assume that calculation or logic of FFT Plot in OpenBCI codes might be wrong. Or, I thought this might be done intentionally for some reasons.
    The accuracy of amplitude is important in EEG analysis.

  • wjcroftwjcroft Mount Shasta, CA

    Hi, William @wjcroft.
    Yes, I used the exact same touchproof electrodes on both systems: Ganglion and Neurobit Optima.
    I tried diffrent channels and cables, but I still got around only 2 microvolts.
    I also tried alpha eyes closed and got 5 μV with OpenBCI, which is about half of 13 μV with BioExplorer.
    I haven't contacted the customer support yet, but I will.

    I played sample data that comes with OpenBUI GUI, "OpenBCI_GUI-v5-meditation.txt".
    In the sample data below, alpha on 10Hz is around 5 μV, which is about 1/3 of normal amplitude.
    Adults amplitude of Alpha with eyes closed is about 16 μV, according to a research below.
    https://www.isnr-jnt.org/article/view/17234

    This is nothing to do with electrodes, adapters, signal or Ganglion because playing sample data does not requre those. So, I assume that calculation or logic of FFT Plot in OpenBCI codes might be wrong. Or, I thought this might be done intentionally for some reasons.
    ...

    Hi again, Yoshi. The Ganglion voltage amplitude accuracy has been checked with calibrated sine wave signal generators (in mulitple labs), so if your Ganglion appears to be always recording much weaker than your Neurobit Optima, that would be something customer support would address. You have already tried both Ganglion firmwares with the same result. Please continue to use only the v3 firmware.

    Regarding the GUI sample meditation data file: we don't know what circumstances or individual that was recorded with. Each person generally has different alpha amplitude, the figure you cite is an average. And even the center frequency peak of that alpha band is known to differ, not always 10 Hz but can be slightly higher or lower.

    https://www.google.com/search?q=individual+alpha+peak+frequency

    Please do email customer support (contact at openbci.com) as soon as possible to get resolution.

    Mentioning @Shirley and @philip_pitts, Yoshi is seeing markedly different Ganglion EEG amplitude, than his reference Neurobit Optima amp. He is already running v3 Ganglion firmware, and v2 performed similarly.

  • Thank you William @wjcroft
    I sent an email to OpenBCI.

  • wjcroftwjcroft Mount Shasta, CA

    https://github.com/OpenBCI/OpenBCI_GUI/issues/1203
    "Linear FFT amplitude is incorrect"

Sign In or Register to comment.