[ot] What I am doing for Biosignals Logging atm

[Undescribed Horrific Abuse, One Victim & Survivor of Many]

First of all, be aware there are communities of lifeloggers who love
to log their biosignals. I’m not up to speed on that. There are also
groups you can join where they collect medical data from you for the
public good. I’m also not up to speed on that either.

Second of all: I am still _practicing_ this. I haven’t verified that
anything is actually working, so it should be assumed that it doesn’t
yet.

The current biggest hiccup for this email is that I don’t have a
public entrypoint to find my data. With some digging, a lot can likely
be found.

Karl’s current EEG situation:

I am using a Muse S with the Mind Monitor app to collect occasional brainwaves.
I practice doing this by wearing the headband even when the app is not
running. Over time, it gets more comfortable until it is pleasant like
other clothing.

I have backup headbands for swapping out, and I get warrantees on
them, because my experience is that with extended use their lifespan
is reduced.

The default setup with the Muse and Mind Monitor yields a lot of data
loss for me. I am always getting bluetooth connection blips. Mind
Monitor also has a bug where it squeezes dropped packets in the data,
resulting in timestamp corruption. Still, I have a ton of data. I’ve
spent a lot of casual daydreams thinking of ways to repair the
timestamps, I don’t think it would be hard.

I have also used OpenBCI equipment and recommend it mostly to support
efforts that have both community and mainstreamness. I believe it
provides much more data than the Muse; however I believe the Muse also
makes accelerometer and gyroscope recordings which are likely good for
tagging physical events.

I presently store my recordings zstd-compressed in a git-annex
repository, and from the git-annex repository I believe many of my
braineave recordings have reached the arweave blockchain using an
arweave git annex remote I kluged together.

I would like to upgrade to an open hashing system more streamlined for
this than git-annex and my kluge.

git-annex:

It’s cumbersome to use git-annex for this, although it does work, one
can run into space exhaustion problems and file count problems that
are easier to stimulate than recover from. Nowadays git-annex does
support an adb remote that can pull directly from android phones; I
haven’t tried it yet, but it would ease use.

If git-annex seems appealing, I recommend learning datalad rather than
git-annex first, as datalad has more provisions for handling things
like large filecounts or deterministic transforms such as zstd
compression.

Voice Recording:

I have both a voice recorder and a mobile recording app, that I use to
take voice recordings. Since I haven’t streamlined storage yet, I
usually don’t run these 24/7 but rather turn something on when I am
engaging a situation that can be confusing for me, or saying something
I value. Sometimes I’ll run more than one at once as a backup, since I
can easily run into unexpected issues.

I manually copy the recordings into a git-annex repository. This would
be streamlined by setting up my devices as a git-annex remote or such
in some way, which I haven’t done yet. It can take a long time to copy
them. They then eventually can end up on arweave etc.

I’m excited when both voice and EEG are running at the same time, as
this means the conditions and workings of my mind can be profiled and
checked more accurately via the second channel.

Additionally, having more than one audio channel recording at once
provides data to triangulate where things are physically located and
moving in increased detail.

Keystrokes:

I am also recording all my terminal sessions using asciinema with its
—stdin flag. Asciinema includes millisecond timestamps, so this
records the timing of my keypresses. These are also added to a
git-annex repository where they end up on arweave.

Since keypresses are made by my muscles and mind, they can also be
used to profile my mental state as it changes from day to day or event
to event, especially given the potential accuracy of asciinema’s
timestamps.

There is a more normative terminal recording program than asciinema,
but I’ve lost track of what it is.

Video:

I occasionally record both desktop video and self-surveillance. Some
of this may have also reached the arweave blockchain. This takes much
more resources than the other things due to the data size, although it
is possible to configure qualities low enough that things become more
reasonable.

On linux, there is a video recording device that treats X sessions as
a video input.

I would really like to get video going better, as most of my weirdness
has been gross human behaviors I engage in, such as hiding objects
from myself. Video also provides the peace of mind of physical
security.

I let doctors know I suspect I might have occasional seizures and want
investigation of this, and from doing this I hope to have medical
video and eeg at some point.  I’m thinking it could give me a leg up
on videorecording to try to mimic the setup a doctor provides, letting
my issues more easily remember it is medically important.