I tried a Nihon Kohden to EDF converter and I got the same error in trying to open the .EEG file as I did in the process of trying to start a subject in a fresh protocol with the data from a disk. Am I going to have to brute force my way into the actual EEG data by defusing the newer encryption or is there a more direct route for handling situations like this? My understanding is that there has been a new proprietary signature added to Nihon Kohden format EEG files with the new year and the default variation of Brainstorm doesn’t have a work around.
Brainstorm output
Converter error message:
error, deviceblock has unknown signature: "�a ʕ���2"
Hi @EternalExpiry, can you share one of those .eeg files to test in our side?
Spolitakiewicz_2c9b7412-5d67-4264-aaf0-8b12bb6635f8.eeg (246.8 KB) Well this is just the .EEG format portion but the entire portion of the of the files related is a few different formats and I might need a different source to upload. Not sure what the file limits are here.
You can upload the file somewhere and post the download link either in this post of send it by direct message
@EternalExpiry, I removed the links from the public for improved privacy
We could access to the files
1 Like
So am I just using an outdated version of Brainstorm and you were able to open the EEG regularly?
@EternalExpiry, I meant that we were able to download the files.
Any reply to the original thread now that you’ve gotten a chance with the files?
As you have noticed, this NK .eeg file does not match any of the previous versions which were supported by different software: Brainstorm, MNE-Python, BIOSIG-toolbox, etc.
There is not explicit information regarding the new version of the software, nor of the acquisition device in the provided files. Also, this new format does not seem supported by any other toolbox.
As a alternative path, do you have a way to export the data in a different format (e.g, EDF) from the NK software?
Looking over the command prompts processes while Brainstorm tried to open my EEG I noticed the following errors after it recognized the unknown signature related to them:
NK> Warning: Could not find .PNT file.
NK> Warning: Could not find .LOG file.
NK> Warning: Could not find .21E electrodes file.
Was thinking maybe there as a way to circumvent the search for these files in the processing of the .EEG, but it seems the structure of the collective file is too strict so the work around might not be capable. Though this does suggest encryption won’t hold up through modern AI decoders.
A post was merged into an existing topic: Error Reviewing Raw File
