This test setup has been very effective. By isolating primitive functions as a set of selectable tests, it is possible to verify the action of the parts individually and determine where they conflict when combined. There is no mistaking the differences between the specific noises and they do resolve to some smaller vector set. They are unique when identified with SHASUM, but that gives no information about their character and association. I am reasonably certain that I have witnessed extremely complex interspecies communication. It is known that even on a cellular level, bacteria and fungi can communicate using chemical messengers. What I find interesting is that when you can isolate a sound by its character and consider it in context, it represents that unique situation and thus is communication. Simple expressions of dominance roars of animals may be understood, but there is much more going on there than most suspect. There was a movie about the "Beastmaster" and I think it was Arnie in that one, where he communicated or controlled animals.
It is assumed that a bird's song is just some arbitrary randomness. I really doubt that. The courtship of even a creature as small as an ant or spider is complex. Many animals do exhibit mimicry and it would seem reasonable that eventually urban creatures would begin to try what they see and retain that if it is useful. Perhaps the biggest hurdle is understanding those very complex issues of the mind "set".
In this case I am attempting to implement a transom transform. "A window above the data" in vector dimensions.
I did some investigating into planet hunters, but it seems that a simple Fourier of the data would reveal any harmonic patterns like a sun crossing and my time would be wasted looking for simple patterns in complex data.
There are some new sets that I wish to search in star data. Given that much of the universe's atoms have been recycled at least 4 times, there would be an atomic glow that is similar to the COBE. It would seem that if a survey was made of how many supernovas occurred in time t and it was extrapolated along with the planet data, it would seem there would be a very large number of clumpy things moving about in galaxies, as well as between them. It would likely follow something like diffusion as well as brownian motion to some extent. It would seem that that chaos might have some strange attractor. I have been working on a technique I developed years ago to identify the "snake in the grass" data in noisy signals. Much of it has to do with known sets. The data is not any more random than randoms in a computer. It originates from a physical system and not all things are possible. It may be complex, but it isn't without form. There are only so many different types of atoms and though the factorial is huge, the combinations of harmonic absorption, refraction, reflection, and emission are limited. In addition they are limited in proportion by the process of the universe and also limited by electric fields, magnetic fields, illumination, light pressure, charge distribution and gravity. Not a simple pattern, but one which has an underlying form and thus progression.
The city scape on the top left and the spectrogram on the right originate is vastly different contexts, and yet there is a correlation between them. What is really odd is that the one on the right is the Dr Who theme song. I was consider how sound spectrograms might be used to draw pictures or even as SVG. I would think that anything which could be drawn, would have a corresponding sound in gray. Interesting thought, perhaps my next test will be to perform a reverse Fourier on various images and observe what they sound like. It would seem that I could create easily decipherable secret written messages in a really odd melody.
An odd thought is that data sets have some hidden associations and wouldn't it be odd if the FFT of some copyrighted code was in fact another copyrighted image. It would seem that certain transforms would describe the relationship of many different copyrighted entities. In the same way that a sound might recall an emotion, a sound might actually represent an image in transform. I was wondering if exhibiting the FFT spectrogram of a copyrighted sound could be considered publishing that sound. If it were not the case, then if I create a transform of a copyrighted set, then it could be distributed legally. The problem arises from the similarity of sets in transform. If one set can be shown to be a specific logical transform of another and the transform itself is patented in its method, it would seem that there is a conflicting state of ownership involved. I find it bizarre that one could claim ownership of a data set in this complex age, but then it serves to generate tax in the form of commerce and so suits the purpose of government. It allows collecting taxes on something which does not have any value in a physical sense. If the cost of production is virtually zero, that is the highest point of profit and the greatest return on investment. It makes no difference if the concept is sketchy or even paradoxical and impossible. It serves the state and as such it will be enforced for their gain and continuation above the interests of individuals.