How good is LavaRnd?
LavaRnd Quality: Cryptographically sound argument
The LavaRnd reference implementation uses the digitized luminance
values from a lens-capped webcam CCD whose gain has been turned up to
The CCD noise comes from dark current thermal electron migration,
imprecise charge measurements, and inaccurate Analog to Digital
The inability to perfectly predict the physical processes within the
CCD rests on the Heisenberg Uncertainty Principle of theory of quantum
The digital luminance output of a CCD frame is a measurement of the
physical state of the device.
The Heisenberg Uncertainty Principle says that not only will this
measurement be imperfect, but the very act of taking the measurement
will disturb the physical conditions of the CCD device itself.
The Heisenberg Uncertainty Principle can be used to show that
you will fail in your ability to perfectly model the physical conditions
that lead to CCD noise.
The chaotic nature of the webcam CCD means that the slightest
mistake predicting its physical state now will quickly render future
predictions completely useless.
Not only does the Heisenberg Uncertainty Principle frustrate your
ability to perfectly predict a given frame, it makes your imprecise
measurement useless as a tool to replay past frames or predict future frames.
Even a small error in predicting a given frame casts doubt upon your
ability to approximate the potential luminance values of the next frame.
The uncertainly of one frame is compounded with the error in taking the
next frame measurement.
Due to the nature of chaos, uncertainly of luminance values of a given
frame is amplified to produce even greater uncertainly about the next
The high gain of the lens-capped webcam CCD produces noise that cannot
be predicted with absolute certainty.
Even the act of taking
measurements disturbs the underlying
physical state of the CCD in an uncertain way.
The error of one frame is compounded with the imprecise measurement of
the next frame.
This chaotic compounding quickly dominates all calculations making
prediction of future frames based on the current frame intractable.
You have no better chance of accurately predicting the state of the
high gain lens-capped webcam CCD than you do in making accurate long
term weather forecasts.
A wrong guess for a single bit of a single pixel of webcam CCD
usually leads a incorrect guess beyond more than 80 bits of LavaRnd output.
While LavaRnd lacks a formal proof that it is cryptographically strong,
one can make the heuristic argument that the Heisenberg Uncertainty
Principle and the nature of chaos makes prediction of luminance frame
destroys any non-chaotic luminance data.
Combined with the uncertainly of the next frame, the LavaRnd output from
one frame cannot be correlated with the LavaRnd output of the next frame.
Even having the LavaRnd output of a large number of frames renders
predicting past or future frame values an intractable problem.
What is next?