Uniqueness of the Mind Mirrors

The Mind Mirror is an electroencephalograph (EEG) which was originally created in 1976. It performs frequency analysis of signals generated by the brain and displays the results on horizontal barographs. The display panel has two columns of bar graphs which represent the left and right hemispheres (LH, RH) of the brain. Zero signal is displayed by all the indications showing at the middle of the display. Each bargraph represents one filter; if a LH signal is being displayed, the indication moves away from the center of the panel to the left and similarly for the right.

The filters which analyze the brain signals are centered on frequencies chosen to give optimum analysis. These are 0.7Hz, 1.5Hz, 3Hz, 4.5Hz, 6Hz, 7.5Hz, 9Hz, 10.5Hz, 12.5Hz, 15Hz, 19Hz, 24Hz, 30Hz, and 38Hz. These frequencies were chosen by previous experience and needed to be modified only slightly after early work with the Mind Mirror. The bandwidth of each filter is adjusted so that the 3db loss points coincide with the equivalent points in the adjacent filters. This means that a brain rhythm whose frequency lies mid-way between two filters will appear at
a reduced level in both. This disadvantage is balanced by the fact that a signal lying between two filters will not be lost.

The performance of analogue filters is excellent, the rejection of unwanted signals is 50db per octave. Translated, this means that the response of the 9Hz filter to a 4.5Hz signal is reduced by a factor of more than 100 times.

These machines and filter frequencies were the basis of the early work in "consciousness" research by C Maxwell (Max) Cade and its more recent development by Anna Wise. The analogue filters proved to be uniquely suited to this application because the optimum bandwidth of each one can be easily chosen.

Brain rhythms respond in both amplitude and frequency to changing thought patterns. Indeed if any frequency is very stable, it seems to indicate a rigid thought pattern which can manifest as a seizure. Our work suggests, for example, that the alpha frequency of a well-developed subject (someone who is very good at what they do) may average 9Hz but may be varying between 7 and 10Hz. This seemed to be true of all the subjects who excelled: what they excelled at did not seem to be important. They could be yoga teachers, television presenters or healers.

The original analogue filters in Mind Mirrors 1 and 2 were precise but difficult to construct with individual components and became too expensive to manufacture. An unusual design was adopted for the filters which allowed the center frequency and bandwidth to be set independently. I have never seen this in a filter textbook even though it is very easy to implement. Fortunately digital techniques can precisely imitate analogue filters and even to some extent improve on them so that the Mind Mirror technology has been translated into the digital domain. The cost is of the instrument is reduced and the precision of the resulting filters guaranteed.

FFT Analysis of Brainwaves

We are often asked why we do not use Fast Fourier Transforms (FFT), which are easy to implement in computer software. Due to the way it functions, this method gives a different pattern to that seen on the Mind Mirror and is not very responsive to the grouping of bands of frequencies known as the beta, alpha, theta and delta responses for a variety of reasons.

FFT works by taking a sample of the signal for a precise interval of time and then by calculation delivers an analysis of the signals present during that interval. It is excellent when the signal:

• is a constant frequency
• is repetitive
• is a much higher frequency than the sampling window interval
  

An example of its use is tone recognition used for telephone dialing.

The signal to be analyzed must be of a higher frequency than the sampling window time, i.e., a delta signal of 1Hz cannot be captured (a change of the signal in 1 second) by a sampling window which lasts one second without gross errors occurring. Any system with 1Hz window will be inaccurate below about 3 or 4 Hz.

Why is this? The sampling window chops into the signal, generating false answers, an effect called leakage. If the window was precisely synchronized with the signal being analyzed, there would be no problem. But we are analyzing the signal because we want to know the frequencies present. This means that at the beginning and end of the sampling interval there must be distortion. Consider a 1Hz wave and a 1 second sampling window. If the beginning of the window was aligned with a peak of the 1Hz wave, then the sample would contain a fast excursion from zero to maximum and this would imply a whole range of frequencies not present on the original. The problem can be minimized by shaping (tapering) the sampling window-switching voltage.

The leakage shows itself in every channel as a false reading. The effect is less when the frequencies are higher, e.g., 10 cycles of alpha during a 1-second window because only the first and last cycles are distorted by the sampling but the effect is still present. A reasonable guess might be that the error has been reduced to 5%. In most cases this does not matter (i.e., in telephone dialing). It can be eliminated by cutting out the low-level spurious signals, a process called application of a Hanning window. Suppose though that the wanted signal is at a low level; it too would be eliminated.

This use of the Hanning window partly explains why it is difficult to recreate an accurate beta response. A typical beta filter bandwidth is from 17 to 22Hz. The beta frequency is responding in both amplitude and frequency to the subject’s thoughts. Any beta signal lying within this range is automatically included in the response of the analogue filter and displayed even though it is varying in frequency. With FFT it would seem possible to obtain the same result by adding together the output of the windows 17, 18, 19, 20, 21 and 22Hz. In practice, the continuously changing beta does not rest within one window for any length of time and is therefore often low
enough in amplitude to be chopped out by the Hanning window.

If the beta amplitude is large enough, FFT can generate beta readings which are approximately accurate, but the method becomes more and more inaccurate as the beta amplitude reduces. This makes it impossible to use to display Mind Mirror patterns.

The FFT can only generate outputs which are distributed linearly in frequency. If for EEG analysis the window is chosen as 1 second, the frequency "bins" are 1Hz wide from 1 to 40Hz. The relative width of each bin falls with frequency, i.e. from 1 to 2 Hz is a whole octave wide whereas from 39 to 40Hz is only a fraction of an octave. This does not matter if the signal consists of a single tone but when the source is an EEG which is varying continuously in frequency, then this unavoidable change in the relative bandwidth provides another explanation why the beta response is very different with FFT. FFT does not respond accurately to rapidly changing or transient signals. This is important when reading alpha frequencies which are rarely very steady in amplitude.

To understand this limitation we have to consider how the FFT analyzer works. Typically in EEG analysis, the sampling window is 1 second long but this does not mean that one has to wait for 1
second before a response is available. Overlapping windows are used; one choice might be 8 per second. After the first second, an eighth of a second of fresh data is added, the oldest eighth second is discarded and the latest 1 second of data is calculated. This speeds up the response after the first second so that low-frequency signals can be displayed within about 0.3 of a second of their occurrence.

Paradoxically this does not improve the transient response. If a 10Hz wave is being sampled by a 1-second window, it cannot display the full amplitude of alpha until the window is full and this takes a second to complete.

Consider what happens when the alpha is fluctuating. Assume the window is 1-second long and that bursts of five alpha waves are arriving every second. The rise time of the analogue filter would allow it to follow this easily, even though the fall time would be extended slightly by the detector time constant. But the FFT would see a half-full bin each time and so display the average - an unvarying output of half the relative amplitude. The relative time position of the sliding window wouldn’t make much difference because the window will still be only half full wherever the burst of 5 cycles is sitting within the window. The representation will look completely different on the two systems, one would be correctly following the peaks whereas the FFT would be showing an unvarying level. These bursts represent useful information in Mind
Mirror technology.

The transient response of the FFT cannot be improved by shortening the window time because it will then become comparable to the frequency and thus leakage will introduce large leakage
errors, especially into the adjacent channels. For example, an FFT with a quarter-second window could respond quickly but as the window will only contain 2.5 cycles of 10Hz alpha, one can see that considerable distortion (leakage) will be generated.

This means that FFT limitations cannot be overcome by using two or three FFT analyzers running
at the same time. To summarize:

• The Delta response is not good with tolerable response times.
• The process is not very responsive to fast-changing signals such as alpha.
• The FFT analysis itself produces spurious signals which are not present in the original. A
  correction for this problem causes beta inaccuracy
• The pattern shown by an FFT analyzer is sufficiently different to make it unusable for Mind
  Mirror work.
  

Uniqueness of the Mind Mirror Pattern.

Anna Wise has been contacted by many frustrated users of FFT systems wondering why they could not see the patterns which she described in her book The High Performance Mind. She says:
"Twenty years’ work with the Mind Mirror has led me to understand the uniqueness of its pattern display and the complexity of the information I gain from it.

People need to know that they do not gain the same information from FFT systems and not realizing its limitations, could lead them to dismiss their own abilities and/or dismiss my years of research as not being replicable." "Accessing alpha allows the flow of information from the unconscious (delta), through the subconscious (theta), to the conscious (beta) mind. It is this openness or availability of awareness on all levels that constitutes the state Max Cade called an
Awakened Mind. Applying and using and manifesting with this open flow of conscious awareness gives us a High Performance Mind. We have to focus on, find, and thoroughly develop the alpha bridge to allow this to happen." All the original studies of healers, swamis and anyone who excelled at their job by Max Cade and myself described in his book The Awakened Mind were performed with the Mind Mirror and will not be replicable with an FFT system.

See Mind Mirror Contact Placement