Paul's Brainwaves

[RogerB]

I'll be recording lots of stuff. No problem with what you mention. The big problem is keeping motion artifacts out of it. Even with fresh electrodes with mucho "glue" to keep them still the readout is still very sensitive to facial motion, despite years of practice with TRs at keeping the face still. I'll mainly be working on that for a bit.

That video in the OP was more of a curiosity than anything else. There is no obvious correlation visible, but that doesn't mean it isn't there. (Or maybe there is no correlation at all. I don't know either way.)

Paul

Well, mate, I think that is a wonderful experiment.

While you would be aware of any bodily movements or frownings and such that we wouldn't pick up out here; in my view there was correlation between the brain wave reader and the meter.

BUT . . . what throws it out is that the meter appeared to register it before the brainwave visualiser software got the image out into the visual part of the screen . . . . in other words, what was registered in the brainwave visualiser software was being formulated just off the bottom of the screen and yet still to come up into being viewable when the meter needle registered.

SO, what would be more accurate, is if the brainwave visualiser software had its "change point" registering in the middle of the screen rather than coming up from "off screen" . . . . this because, by the time you can see the pattern of the mind that is the "read," it has already shown on the meter and is past.

What I'm saying here, in case I've not expressed it clearly, is that what is visible on the brainwave visualiser is a delayed visual of the actual event in the mind; whereas the meter registered more "instantly" or coincident to the event.

Rog

 

[Dulloldfart]

What I'm saying here, in case I've not expressed it clearly, is that what is visible on the brainwave visualiser is a delayed visual of the actual event in the mind; whereas the meter registered more "instantly" or coincident to the event.

Rog

Yes, there does seem to be a delay of perhaps 1/4 second. This IBVA unit is very CPU-intensive. I'm using a new Mac Mini with a 2.3GHz dual-core Intel Core i5 chip and 8 Gigs of RAM and the processor load is literally 105% per the OS X "Activity Meter."

I'm guessing that it takes that long to do the FFT calculations and put them onto the screen, rather than that the brainwaves are slow.

Now, the brainwaves are recorded anyway. The current meter software doesn't record, but future software versions will. So it will be possible to video a version with the meter readings moved back 1/4 second or however long the actual delay is. Trick stuff this!

Paul

 

[Emma]

Well, perhaps if you put just a little more effort into creating some charge ... (j/k)

In the name of research, should we all band together and try really hard to really piss Paul off?

 

[RogerB]

Yes, there does seem to be a delay of perhaps 1/4 second. This IBVA unit is very CPU-intensive. I'm using a new Mac Mini with a 2.3GHz dual-core Intel Core i5 chip and 8 Gigs of RAM and the processor load is literally 105% per the OS X "Activity Meter."

I'm guessing that it takes that long to do the FFT calculations and put them onto the screen, rather than that the brainwaves are slow.

Now, the brainwaves are recorded anyway. The current meter software doesn't record, but future software versions will. So it will be possible to video a version with the meter readings moved back 1/4 second or however long the actual delay is. Trick stuff this!

Paul

Yes. I was thinking maybe, even up to a 1/2 second delay.

Though you lost me on the chip Ram specs thing Though I do understand the 108% of CPU load . . . my XP system routinely jams up at 100% and the world has to wait

I think you are onto something rather good with this experiment . . . it's a pity the physical universe equipment isn't up to the job just yet.

Rog

 

[Dulloldfart]

Well, perhaps if you put just a little more effort into creating some charge ... (j/k)

In the name of research, should we all band together and try really hard to really piss Paul off?

It's not too hard to dig up some charge. It's a matter of REALLY stirring stuff up instead of just glancing at it. See my write-up here. A short excerpt:

As of April 12, I have run about 18 hours solo on these R2-60 items. I am up to "Inhibited Thinking", which ran for about 25 minutes a couple of hours ago as of this writing. Still going strong. About 90-95% of these items are charged. The secret is in thoroughly grooving them in, to make them real to the pc. I spend about five minutes seeing how the item would apply to me. As an example, the last one, inhibited thinking. I'll discuss this one simply because it is fresh in my memory, otherwise it isn't special at all. Inhibited thinking? Mustn't think about critical thoughts in the CofS; mustn't think about sex with underage girls; mustn't think critical remarks about friends as it indicates overts; mustn't think of restimulative things as it, er, restimulates one; other flows on this; inhibiting the relay of information from remote viewpoints one has put out; inhibiting info flow to remote viewpoints others have put out; all the above as it relates to other real or imaginary terminals (includes entities and such possibly mythical individuals); and so on.​

So it's just a question of working at it a bit.

The biggest problem is you can't just grab someone off the street and see what it looks like with them, because of the motion artifacts. Maybe if one drilled through the skull and put the electrodes under the skin. . . .

Paul

 

[Panda Termint]

In the name of research, should we all band together and try really hard to really piss Paul off?

I seem to be able to do that sometimes without even trying.

I'm intrigued by this experiment though.

 

[programmer_guy]

I'll be recording lots of stuff. No problem with what you mention. The big problem is keeping motion artifacts out of it. Even with fresh electrodes with mucho "glue" to keep them still the readout is still very sensitive to facial motion, despite years of practice with TRs at keeping the face still. I'll mainly be working on that for a bit.

Paul

Damn. So, the body motion problem crops up yet again.

 

[programmer_guy]

Re: Floating Needles and Fast Fourier Transforms

http://www.random-science-tools.com/maths/FFT.htm

It occurs to me that an F/N is a composite of more than one sine-type wave. Any engineers around who could use something like the above FFT calculator to (try to) break one down into its components?

Having done so, maybe we could then get closer to figuring out what it is that an F/N is showing. "Fluctuating resistance" is a silly answer to the question, and might not even be true. What if the voltage is changing instead?

Paul

Interesting that you mention FFT.

IF (and that is a BIG if) facial movement is expressed in only certain frequency areas then they can be filtered out using FFT by filtering those certain frequency "buckets" and then performing a reverse FFT on those filtered results to possibly get some or most of the actual brain activity.

Unfortunately, a simple FFT is typically used on a 2 dimensional input (time and single measured value like voltage or sound pressure level).

The output of what you are using looks to be 3 dimensional. So, FFT would have to be performed on each row and/or column of data... then filtered (if that is possible) then perform a reverse FFT on the results to go back to the time domain and see the resulting graph. (OR stay in the frequency domain to see if that is good enough for your purposes).

 

[programmer_guy]

I'll be recording lots of stuff. No problem with what you mention. The big problem is keeping motion artifacts out of it. Even with fresh electrodes with mucho "glue" to keep them still the readout is still very sensitive to facial motion, despite years of practice with TRs at keeping the face still. I'll mainly be working on that for a bit.

Paul

Have you tried asking the manufacturer about the motion artifacts? Just curious.

 

[Dulloldfart]

Re: Floating Needles and Fast Fourier Transforms

Interesting that you mention FFT.

The IBVA software uses FFT bigtime. I'll write more on this later.

-----

From http://en.wikipedia.org/wiki/Delta_wave:

A delta wave is a high amplitude brain wave with a frequency of oscillation between 0–4 hertz . . . They are the slowest, but highest amplitude brainwaves.​

Very roughly, Delta is 0-4 Hz, Theta is 4-8 Hz, and Alpha waves are 8-12 Hz. The classifications seems a bit wobbly and vary from source to source. The "waterfall" display in the video shows from 0-30 Hz. Beta is 14-30 Hz, and represents normal awake consciousness (supposedly). If you can't read the labels going from left to right underneath the waterfall display they say "Eye Delta Theta Alpha BetaL [low] BetaH [high] Gamma. The "Eye" means eye-blinks mainly show up in the 0-0.5Hz range, although they also spread across the whole spectrum if the body motion is marked.

The books say Delta is associated with deep sleep; Theta with deep relaxation and meditation; Alpha with relaxed, calm, lucid, not thinking.

Except for the fake representations in the CofS films, and Hubbard's definition, F/Ns are not regular fluctuations, although they are kinda rhythmic. The wider (deeper) the F/N, the slower its "frequency." A floating TA is *really* slow (maybe 1/6 Hz?) and relatively long-lasting, a normal F/N might be 0.5 or 1 Hz, and a titchy F/N might be 4 Hz although it won't last long at all.

I'm guessing right now that an F/N reflects a composite of slow brainwaves in the delta and to a lesser extent theta bands. The wider the F/N, the lower the brainwave frequency.

As for the "deep sleep" bit, pffft! There are lots of delta waves in my brainwave recordings, even when I am talking.

Paul

 

[Dulloldfart]

Have you tried asking the manufacturer about the motion artifacts? Just curious.

Nope. Still exploring.

The programmer is a Japanese guy in New York (it seems). The person I bought it from in the UK (and whom I have a comm line with) is more of an importer than a manufacturer or programmer, and has somewhat limited knowledge.

Paul

 

[Dulloldfart]

From http://brainathletesports.com/pdf/WP-lee-neurosky-eeg.pdf:

The amplitude of an EEG signal typically ranges from about 1 μV to 100 μV in a normal adult, and it is approximately 10 to 20 mV when measured with subdural electrodes such as needle electrodes.​

Since I'm not about to drill holes in my skull to implant electrodes, I'll work with the IBVA above-skin electrodes. There are three horizontal white bands showing at the front of the waterfall display, which represent the "power" of the brainwaves. Most of the IBVA videos you will see online have these levels set at 0, 5, 10 μv (microvolts). This corresponds to a gain in the preferences set-up of 2x.

The first video in this thread has the gain set at 5x, with corresponding levels shown by those white bands of 0, 2, 4 μv. I turned up the gain to make the display look prettier. . Most of the time it looks kinda dull and mostly shows as dark blue/green except when blinking or introducing other artifacts. The light-coloured "doughnuts" show the peaks. At any one moment, each of the left- and right-brain displays shows three peaks, three doughnuts, to indicate the three points across the spectrum that have the highest amplitude (voltage).

Anyway, I digress. Let's assume the brainwaves at source are in the 10-20 mv range, although only a fraction of this gets through the skull to the above-skin electrodes. I'm going to assume (guess) that an "F/N" corresponds to a varying voltage of 0-20 mv, roughly speaking. At a TA of 2.0 or 3, how much variation in "resistance" (hee-hee) does an F/N show? 100 ohms? 500 ohms? 1000 ohms? Very roughly this would correspond to a change of 100/5000 or 100/12000 or 1000/5000 or 1000/12000, maybe 1% - 20% of the "resistance" or 1-20% of the current or voltage. The voltage of an e-meter is about 2 volts, the USB-powered one maybe 5 volts. 20mv is 1% of 2 volts.

Brainwaves per se are not going to be measurable in the fingers, but neuronal impulses obviously travel between the brain and the fingertips. Neuron voltages are in the 50 mv range, it seems.

Anyway, the figures seem to be in the right ballpark in terms of my very sloppy calculations.

So does the e-meter measure fluctuations in VOLTAGE caused by brainwave output with its effects mirrored all over the body? I'm talking about short-term fluctuations as seen in reads and F/Ns, rather than long-term fluctuations as reflected in gross TA changes. TA changes must be a different phenomenon. An LFBD would be a combination of the two.

Of course, there are all the subtle-energy mechanics as well in there. I am not assuming it is only the brain and purely physical channels being involved, but it is easier to concentrate on that which can be measured with real-world instrumentation.

Paul

 

[beltway]

This topic is fascinating... I hope FFT analysis can be worked out.

Regarding the problem of speech artifacts. what about experimenting with electrode placement? Maybe further back on the scalp? Smaller (pinpoint?) electrode extensions or 'caps'?

I did notice similar spikes mostly in the far-left (white) band twice when you said the word, "theta", but since you mentioned that area also responds to blinking also it might just be registering speech-related facial movement...

It seems to me that the parts of the brain you're monitoring with the IBVA are higher function type areas whereas if the physiological GSR explanation (apocrine/sweat glands, fight-or-flight) is right, the meter is reflecting much 'deeper' autonomic brain activity. Again, this is really neat stuff, thanks for sharing it!

 

[Dulloldfart]

This topic is fascinating... I hope FFT analysis can be worked out.

Regarding the problem of speech artifacts. what about experimenting with electrode placement? Maybe further back on the scalp? Smaller (pinpoint?) electrode extensions or 'caps'?

Yes, thank you. I was considering that earlier today. I already had them fairly high up on the forehead. Maybe I'll try the front of the scalp next. I'll have to shave my head a bit first, but there is very little hair left so it's not drastic at all.

I can't mess with the electrodes as there is a special (small) 4-pin jack plug that connects to the electrode "holder" and I am not going to take that apart.

The sweat glands have nothing to do with it. See my underwater metering videos at the ESMBDOF YouTube link in my sig, where the meter still reads despite the cans being several inches away from the hands underwater (two buckets!).

Paul

 

[Helena Handbasket]

I had the idea years ago of doing an FFT (spectrum analysis) on the signal from the cans. Unfourtunately, I didn't move on it and others beat me to it.

I understand Hank Levin has come up with a "meter" that does just that -- it has 6 coloured LEDs that show the different reactions -- but no needle! You just look at the LEDs -- no more e-meter drills! (Unfortunately, this is a new product that is not yet shown on his web site.)

I think they've also done something like that in Russia.

Filtering out body motion, I believe, is easy, because it occurs on different frequencies than thoughts -- you just ignore that part of the spectrum.

Tommy Thompson once showed me his e-meter read simulator. He said this is a floating needle -- it was a perfect sine wave, at just one frequency. I said that's not a floating needle. He got mad at me.

It shouldn't be too difficult to analyze the spectrum without overloading the processor. The kinds of frequencies we're talking about only require samples every few milliseconds (rather than microseconds as in audio spectrum analysis) and the core of the routine can be written in assembler language for speed. At the worst, you can have a digital signal processor (DSP) pre-process the input and give only the analyzed results to the main processor.

I am, however, working on a project to flowchart the various processes. If we can link those up to a computerized FFT meter, it just might take most of the burden off of the auditor, making auditing something anyone could do with just a few hours instruction.

Helena

 

[Dulloldfart]

http://www.eserc.stonybrook.edu/ProjectJava/WaveInteractionApplet/

This Java applet allows you to see the cumulative effect of the superposition of three sine waves. You can adjust their frequency (wavelength), amplitude, and phase. For an example, I set these parameters to:
Wave 1: W197 A40 P0
Wave 2: W151 A30 P36
Wave 3: W100 A20 P176

I would have used rounder figures but there is a slider rather than a numerical input for the changes.

You can also set the velocity in the final animation box. I set these to 3, 2, 3. These are all somewhat arbitrary numbers, but the intention is to show the addition of three sine waves in the delta brainwave band.

To put this into more familiar terms, look at the right hand end of the blue line, the vertical line where the wave is created. Turn your head 90 degrees (or the monitor ) and imagine the point where the wave starts as the end of a meter needle. It's easier if you cover up all of the blue line except the end 1/8 inch or so.

Juggle the figures around a bit. Looks close enough to me. It's not exactly experimental confirmation, but it doesn't hurt my hypothesis that an F/N reflects the addition of more than one sine wave.

Paul

 

[Dulloldfart]

I had the idea years ago of doing an FFT (spectrum analysis) on the signal from the cans. Unfourtunately, I didn't move on it and others beat me to it.

I understand Hank Levin has come up with a "meter" that does just that -- it has 6 coloured LEDs that show the different reactions -- but no needle! You just look at the LEDs -- no more e-meter drills! (Unfortunately, this is a new product that is not yet shown on his web site.)

I saw that. Think he sent out an email.Without trying one out, it looks like more of a toy than a serious instrument. It also adds an impenetrable layer between what is happening and the "reads," so you have to take it on faith that the programming is spot-on. I wouldn't mind trying one out, but I'm not going to buy one.

I think they've also done something like that in Russia.

Filtering out body motion, I believe, is easy, because it occurs on different frequencies than thoughts -- you just ignore that part of the spectrum.

Well, no. On the IBVA graph the 0-0.5HZ part of the spectrum is labelled "eye," which is all very nice in theory. If you blink quickly and gently, it does restrict to this range. If you blink heavily, scrunching your brow even just a little, it slams across the whole 0-30Hz range, and I imagine it goes higher too but I haven't figured out how to set it to read over 30 Hz yet.

Tommy Thompson once showed me his e-meter read simulator. He said this is a floating needle -- it was a perfect sine wave, at just one frequency. I said that's not a floating needle. He got mad at me.

Yeah, that sounds like Tommy.

It shouldn't be too difficult to analyze the spectrum without overloading the processor. The kinds of frequencies we're talking about only require samples every few milliseconds (rather than microseconds as in audio spectrum analysis) and the core of the routine can be written in assembler language for speed. At the worst, you can have a digital signal processor (DSP) pre-process the input and give only the analyzed results to the main processor.

Feel free to write your own software. The OpenEEG Project has a lot of stuff already done: http://openeeg.sourceforge.net/doc/.

I am, however, working on a project to flowchart the various processes.

What exactly do you mean by this? You mentioned this before. The idea is familiar to me as I need to do it every time I write a new PaulsRobot module as I need to categorise every possible response to an auditing command into a few options. I was just curious what you had come up with.

If we can link those up to a computerized FFT meter, it just might take most of the burden off of the auditor, making auditing something anyone could do with just a few hours instruction.
Helena

There's always my Robot approach.

Paul

 

[RogerB]

Boy! What a neat thread.

Real talented people pitching in here.

Great to see . . .

Rog

 

[Dulloldfart]

I saw that. Think he sent out an email. Without trying one out, it looks like more of a toy than a serious instrument. It also adds an impenetrable layer between what is happening and the "reads," so you have to take it on faith that the programming is spot-on. I wouldn't mind trying one out, but I'm not going to buy one.

I think they've also done something like that in Russia.

Paul

Damn! That Russia sentence is spurious, a quote from Helena that shouldn't be in my response field there.

Paul

 

[MattD]

My "impression" is that the brainwaves would be analogous to "significance" whereas the meter movements would be "force". Apples and Oranges to a certain extent. (both fruit though)

The meter is used to hunt for the force; processes are used to align significance to the point of cognition, where the significance that holds terminals apart and generates force, collapses. So you may not find a correlation in the two visualizations as they are coming at the problem from different vectors.

It would seem that the visualization of significance (if my idea is correct) would be a more complex method of resolving the problems being addressed. 2wc moves significance around nicely, but is less useful to find force.

You may have stumbled into a route to a new paradigm of approach to "mental" stuff if you could direct someone via brainwave feedback to align them with a more optimum state. (whatever that may be! and that is probably what it would hinge on)

Have fun!



(it made sense to me when I wrote it)