Latest Entries »
in every change in direction I had a short period of time of stillness.
the axis cannot retrieve the right orientation due to the position of the sensor, which is calibrated for staying in plane on a table. so looking up/down doesn’t mean an Y movement, for example.
A dot is a gentle blink, so that the NeuroSky EEG channel spikes above 500, but below 900. A dash is more of a frown, or wink or raising of the eyebrows so that the NeuroSky EEG channel spikes above 900. To practice, reference the International Morse Code chart. Start with some E’s (dots) and then a T (dash). Then say SOS and finally CONSCIOUS.
I caught that feeling you get when you are startled awake by an intense falling sensation on file again, this time with EEG. At 0052 I was still awake and making sure the audio tracks were triggering, then at 0101 I twitched awake and checked the time.
The minute above shows the episode – the first spike was the twitch and seems to have flat-lined the channel for a few seconds. The rest of the movements are from me checking the time and settling again.
Download entry as Lucid Scribe Data (LSD) or Comma Separated Values (CSV). The CSV file doesn’t contain the RAW channel – it was running at 256 Hz and the CSV export doesn’t work for the illuminated plugins yet. I will fix that in one of the next updates.
The first printout that I recorded last night, 5 minutes past midnight, shows the strained beats picked up by the halograph FM. The fREM channel can be seen counting up each blink or strained beat that peaks over the top line and then triggers the next track in the playlist after the 8th blink.
And the second printout, recorded at 0017, shows the same thing happening with the OpenREM channel from the OpenEEG electrodes.
This will make it a lot easier to perfect the algorithm by highlighting where it was thinking about triggering. The tracks at 0618 made it into the dream and helped me stave off a hoard of zombies, but didn’t trigger lucidity…
The first minute, 0143, shows a pattern that I have seen many times: a sequence of heartbeats where every third beat is three times as strong as a normal beat, 12 of which can be seen below. I have begun to suspect that this may be a phenomenon that only occurs in male subjects. Note that the EEG channels flatline.
Moving on, two and a half hours later at 0408 there are two such beats right before the onset of a marathon REM session that lasted half an hour.
And peaked with eye movements like this at 0412:
Only to end again with some strained beats at 0450. No idea what to make of that spike on channel 2, it seems to pop up quite regularly.
As evidenced by the collection of Lucid Scribe plugins that connect to third-party hardware, I would have been among the most excited if it had been a legitimate product. But their EEG didn’t complete a circuit, their induction percentage was unheard of, their amp was orders of magnitude out of the range of brainwaves and – what ultimately exposed it as a scam to the masses: the images of their prototype were created in Photoshop.
I reported it to KickStarter right away, but they didn’t do as much as acknowledge my report. I couldn’t sleep during the last weekend of the campaign as they were about to coin half a million dollars while making a mockery of lucid dreaming research. So as a last resort I backed the campaign in order to post my findings as a comment.
I then witnessed what appeared to be a P300 event-related potential in the attitude of the crowd. On a much slower scale, of course, but similar to what happens in the brain when you recognize something. I have been keeping my eyes out for P300 waves, as I hope to one day find that the onset of lucidity, or the recognition of the dream state, produces such a wave.
Here is some interesting research that was done at the Usenix Security conference:
The researchers designed a program that flashes up pictures of maps, banks, and card PINs, and makes a note every time your brain experiences a P300. Afterwards, it’s easy to pore through the data and work out, with fairly good accuracy – where a person banks, where they live, and so on.
And DARPA used the P300 event to recognize threats. They showed users ten images per second (of desert terrain, for example) and noted which images triggered a P300 – 91% of the time it was because there was a threat in the image that the person might not have consciously recognized.
So I tried my hand at measuring and graphing the “brainwaves” of the crowd. I wrote a script to flip the comments so they appear in order and measured and weighed each comment as either positive (excitatory) or negative (inhibitory). If we pretend that each backer is a brain cell and each comment is a neurotransmitter, then we can plot an electroencephalograph of the crowd.
The printout featured above shows the amount of positive (blue) and negative (red) comments per hour over the last 100 hours of the campaign, with the inhibitory comments peaking at -120. Not as close to what an event-related potential looks like as I expected to find, but interesting nonetheless and entertaining at the least.
The campaign was cancelled after enough people reported it and it was covered on pandodaily, in the Wall Street Journal and in Crowdfund Insider. Shout out to the bodhisattvas Majid, Yeti, Sascha and Mr. Brand for staying behind and waking up the others! And special thanks to Highlander over at the DreamViews forum for sleuthing the name of the amp.
The eight waves in the minute featured above seem to have repeated every hour, around the 50 minute mark, starting at 2 am. The relative quite in the accelerometer channel suggest that the waves aren’t artifacts. I might set an alarm for them one day.