We use
a short time Fourier transform, which means that we
essentially break up the sample into
a number of smaller samples which can be analyzed and
broken into a sum of sinusoidal's. However it is not
enough
to simply break down the sample into a non-overlapping
set of smaller samples, there is some framing that
must
be done, in our processing the sample windowing is overlapped
by 75%. This provides for a smoother scaled output
signal
without the large number of signal artifacts, which would
otherwise be present at the boundaries of our processing
sample size. The processing sample size is set at 20ms
which provides for a small enough sample so that we
can
use a Short Time Fourier Transform to generate our scaling
data since over 20ms of time the signal will not significantly
change in the time domain.
Using the STFT (Short Time Fourier Transform)
we generate a Frequency Domain analysis of the signal
by generating an array of bin frequency analyses. Our
bin frequencies are separated by 48hz to provide maximum
resolution in the Frequency Domain. We probe our 20ms
sample for each of our bin frequencies resulting in a
Magnitude, Frequency and Phase result.
We then do some additional processing
to manage phase shifts, which occur due to the fact that
our input sample frequencies are not spaced exactly 48Hz
apart. When a sample frequency participates in more than
one bin frequency probe the phase of the resulting output
will shift. We take this into account in our processing
by the use of an algorithm designed to take the phase
difference in our bin processing output and apply it to
the Magnitude of the frequency and shifting the phase
of the output to be coherent with the expected phase.
Then
it is a simple matter to take the median frequency
domain analysis of the input sample and
our target frequency and arrive at a scaling factor.
This scaling factor is applied to the frequency result
of our
processing. We then process the results of our processing
using an Inverse Fourier Transform which basically
takes
our processed set of sinusoidal frequencies and regenerates
a complex wave form that has been frequency shifted.
We
use this Alpha - Theta - Delta information to imprint
the Structured Water and create the cellular message
CD or MicroSD Chip.
Our application takes audio samples at a rate of 44100
samples per second with an amplitude resolution of 16
bits giving 65536 discreet amplitude steps per sample.
This full CD quality sampling rate ensures that all available
frequency and amplitude information in the voice is collected
and analyzed. Sampling at this rate results in a data
set that is able to represent frequency information where
the Nyquist frequency is 22050 kHz, well above the range
of human speech.
Our application applies a standard Fast Fourier Transform
to the mathematical representation of the voice sample
data to convert the information in the time domain as
it is represented by the sample data collected from the
user to a data structure representing the same information
in the frequency domain. This is an industry standard
analysis function used by all the spectrum analysis tools
available today.
We supplement the utility and resolution of the FFT (Fast
Fourier Transform) by the use of a specialized and custom
arithmetical mathematics library that allows for a far
greater degree of resolution than currently available
in commercial math libraries. Our application also applies
a variant of the FFT algorithm to the input data called
the Goertzel Transform. The Goertzel Transform is mathematically
related to the FFT but acts on only a single frequency,
allowing us to apply a different algorithm to the same
data and increasing again the accuracy of our analysis.
The combination of these two algorithms is unique to our
approach and to this writer's knowledge is not used commercially
in any other product.
Both the FFT algorithm and Goertzel algorithm we have
developed are modified to work against an intermediate
data representation that expands and extrapolates the
data contained within the voice sample. This is required
due to the way that these algorithms work. Both algorithms
result in a series of bins each bin contains two complex
numbers that can be further manipulated mathematically
to produce a frequency/intensity value. It is this value
that is used subsequently in our analysis algorithm.
Due to mathematical constraints the size and thus resolution
of this set of bins is one half of the sample size. An
analysis set size of 1024 samples will result in the entire
frequency domain map spanning only 512 bins; each of these
bins therefore will contain information regarding 43.06
Hz of the frequency spectrum - obviously very low resolution.
This is the type of frequency domain analysis used by
media player visualizations and by some other spectrum
analyzers on the market.
Our application uses a technique whereby the output range
is vastly increased resulting in an output structure that
contains over 1,099,511,627,776 bins. These bins are mathematically
represented with a proprietary format and method that
requires virtually no storage on the sample processing
computer. This representation allows us to analyze voice
data at a resolution which would otherwise require more
storage per sample window than is present on any modern
day computer. Our sample resolution results in each bin
containing frequency information about .00000002005 (2.005E-8)
Hz of the frequency spectrum - as you can see this allows
us to more accurately gain information about the frequency
spectrum of a sample since each bin represents such a
small section of the entire spectrum.
It is difficult to compare our mathematical approach to
a hardware based approach simply because of the limitations
of the hardware based method. Hardware methods have a
resolution that depends on the cost and complexity of
the circuitry used to generate the frequency domain data.
Hardware based approaches use a resonant filter circuit
for each bin that filters out intensity information not
configured for that filter. For each individual frequency
the hardware system analyses there must be a single corresponding
circuit. Due to the physical nature of these circuits
there is a small upper limit on the number of bins that
a hardware based system is able to provide whereas our
software based system is virtual in nature and relies
on mathematical concepts for it's representation and analysis
allowing us practically unlimited resolution.
Our synthesis engine is also mathematically based on
trigonomic functions that output waveform data directly
and allow
us to modify and control the phasing of individual components
of the synthesized audio. Other applications rely
on wave table
synthesis whereby the output waveform is stored in small
chunks (the wave table) and simply copied out to
the output
data. Wave table synthesis is faster but results in aliasing
of output data as a result of scaling which must
take
place to generate waveforms of a different frequency
than what is stored in the wave table. Our method
generates
a smoother, more natural sounding output. Being able
to modify the phasing of component waveforms also
allows
us to generate with great precision beating of the signals.
It is this beat frequency generation that results in
the great impact our system has on the user. By the application
of a proprietary algorithm we are able to tune the standing
wave generated inside the user's brain. A standing wave
is an interference pattern generated when two or more
waveforms interact. The important thing about standing
waves is that they apply energy to a single spot continuously
whereas a regular waveform applies energy only for a brief
period during each cycle. Manipulation of the phasing
of the component signals allows us to generate standing
waves inside the neural circuitry of the user's brain
to initiate and sustain immensely powerful change.
However, our system does not simply beat two frequencies;
the output waveforms are complex and contain more than
simply two waveforms. We generate a complicated interference
pattern comprised of more than 6 waveforms and the interference
pattern thus generated exhibits dynamic shifting in four
dimensional space (the three spatial dimensions and time).
By the use of a phasing equation we are able to manipulate
the Scalarwave energy construct so that it maximizes the
impact on the receiving system - the user.
 |
Adding your personal
frequencies to the IQubes brings more smoothness
and balance into your life while accelerating
personal growth.
We encourage
you to take a closer look at our new Quantum
Sound Therapy software and see how simple it
is to use. We are offering special discounts
to previous owners of our VAHS software, Theta
Love IQube, Awaken IQube and the Tesla IQube. Self
empowerment is really this easy.
"Our system
is by far the most accurate and reliable system
available. It melds the science of mathematics
and sound to produce a system that mediates
change with a precision unprecedented by any
other system. Other systems rely on simple
monotone frequency generation, low resolution
analysis, basic tonal analysis and generally
do not offer the complexity required to mediate
change within the user. When coupled with proprietary
Scalar Vortex Technology, this system is unbeatable." Leslie
J. Marshall (M. Sc.) July 24, 2006
|
|
Our
award winning VAHS software has been renamed “Quantum
Sound Therapy”
Compatible – on
Windows 7 and 8
