Andres Ramos

Marie, if your interest in using microphones is persisting and it helps I can offer to send you a microphone like the one I gained my results from. I would do this for free because I would benefit from this in the way that I could see if it produces good results for you too or not.

I use a microphone for recording spirit voices that works very well. It contains a standard electret condenser microphone capsule and an adjustable preamplifier. I have the idea of making two of them each feeding one channel of a stereo signal. The question is, if we would operate both microphones with some background noise close together, will we get the same voices on both channels?

Yes you are right Jeff. As I read my reply to your comments I became aware that I oversaw the importance of spectral material to form for the spirits. You now have a very valuable theory at hand.

Thank you Jeff! So great to read your technical comments and evaluation here. That makes me think a lot too and you opened a new perspective to me. If I got you right then the main source of the pk- modulation could be the SSM2167. I supposed the same but as I wrote I was unsure if the pk-modulation comes from itself or if it just amplifies the pk from the phototransistor. What you are suggesting is that the non-linear transfer characteristic of the SSM2167 is the motor for the pk-effect but the varying amplitude level of the noise signal from the phototransistor is the fuel. This is a new perspective because I always saw the phototransistor as an entropy source, which it actually is, but the point could rather be the varying amplitude and not the entropy in first line. I observed what you described with strong pk-effects coming from the scratching sounds of turned potentiometers. Last year this was the reason why I made experiments with a rotating graphite drum and electrodes touching it. This morning I analyzed the signal from the SkySound receiver again. I elevated the inclination a bit and wanted to investigate the change in the signal. I heard, like in the recordings before, that there is a layer of strong impulses above the noise with varying distances, but not much varying in total, and I could hear that the voices were forming along with these impulses. It seems like every time an impulse raises, a small fragment of an allophone comes through much like in a sampling process. This is supporting your hypothesis of varying amplitudes, what impulses actually are, and the non-linearity of the receiver. If we push your thoughts further it could be possible to gain spirit voices from trains of impulses (the good old NE555) and an SSM2167 without using entropy at all? This could be a game changer, Jeff!

Thank you Karyn. I really appreciate your support. I know the article became very long. Originally I wrote it in Google docs as a technical research paper to be published externally (e.g. at Anabelas ITC journal or somewhere else). However, as I described, during the work on the experiments I observed the focus shifting away from skylight to any form of light and I became unsure if this unclear focus would affect the paper's acceptance potential. So I decided to publish it here at first as an experiment with ongoing research.

Abstract I am an electronic engineer by profession and involved in instrumental transcommunication for three years. In the past 10 years I did some experiments with light. Specifically I was investigating different light sources for modulations that could be made audible by a simple setup made of a phototransistor and an amplifier. By the end of the year 2021 I decided to repeat those investigations upon the light coming from the sky, especially at dawn and at night (looking for possibly very low light intensities). This report describes a series of experiments I did with different sources of light and my quest to identify the potential for psycho-kinetic modulation in it. Introduction Modulated light was always fascinating to me. The first experiments I started as a young boy when I discovered that ambient light contains hidden information, by just using a light dependent resistor and a simple transistor. My first encounter was the power line hum imprinted in every light source we use. However, very quickly I found out that TV-screens, monitors, control lamps, remote control LED’s and much more were emitting all sorts of different and often complex modulation patterns. In 2010 I designed the best modulated light receiver I had made so far. It found a place behind the windshield of my car and I drove around to sample as many audio recordings as I could. This way I recorded the sounds of head lights, street lamps, neon lamps, signs a.s.o. You can find some examples in the following. Sound samples with modulated light from around 2010 Sound Link Sound from an external display on a bus Bus_Anzeige_2.mp3 Sound from a LED car headlight LED-Scheinwerfer_4.mp3 Beat of two interfering car headlights Schwebung.mp3 Table 1: Examples for modulated light It was at that time already when I first encountered a phenomenon I could not explain. It was an audible anomaly catched at a moment when no visible source of artificial light was around while I was driving down a country road at dawn. Hear this: Was_ist_das_1.mp3 This sound appeared to me like a bypassing U.F.O and I couldn’t make any sense from it. As I said, back in that time I was not involved yet in the investigation of paranormal phenomena. In the following chapter I will describe the circuit I used for my latest experiments in detail. The SkySound light receiver I named the device “SkySound” because I deliberately designed it for the investigation of light emissions coming from the sky. Paranormal phenomena and spirit voices gained in instrumental transcommunication are merely observed at night, so it was logical to create a device that could pick up those faint modulations also, if present at all. The standard approach for receiving light with electronic components is the phototransistor. Basically this is a standard transistor with a transparent case. That way not only small electrical currents going into the base (control electrode) of the transistor are amplified but also light shining on the transistor substrate. These components are cheap, sensitive and easy to handle. Fig.1: VT9112 phototransistor In my experiments I used the VTT9112 phototransistor because I dragged a bunch of them from the trash can in the company I worked for, 20 years ago. The following image shows the electronic circuit that will be explained in detail. If you are not interested in the electronic functions of this device you can skip this chapter safely. You don’t need to know the inner details to follow my descriptions of the anomaly itself in the later chapters. Fig.2: SkySound electronic schematic The transistor marked as “T1” is the VTT9112 I already described. Instead of the classical load resistor in the emitter path I provided a special circuit made of T2/R1/R2/C9. This is a self-adjusting load that always holds the dc bias in the middle between the power supply voltage and ground to provide always the best output swing for T1. Moreover, it works as a high impedance load for T1 and thus ensures maximum output. Transistors T3 and T4 are used as standard emitter amplifier stages. They are pushing the amplitude level high enough for the circuit IC1 to work with. This circuit is an integrated preamplifier originally designed to pre-process microphone signals. It provides a compressor to limit signal overdrive and a noise gate. The noise gate is a circuit that suppresses signals below a certain threshold and amplifies everything above the threshold. In practice this is not a hard threshold. Moreover, the transmission characteristic is non-linear and, if carefully adjusted, can boost small fluctuations in the noise signal, exactly where the spirit voices are hiding. So in easy words, the noise gate is utilized to increase the so-called “psycho-kinesis modulation” (pk-modulation). However, I found out that this function requires very careful adjusted input levels. The pk-modulations in the signal output is enhanced slightly but the signal quality is not. Obviously the spectral compositions of the signal deteriorated slightly by using noise gating. The use of this feature is up to deeper scrutinization. The other components are used for the power supply of the circuits. As I already said, my intention was to pick-up rather low intensities of light at night. Nevertheless, after my first tests I found out that the receiver should also be capable of handling bright light. the result was the above-described self-adjusting load. The phototransistor has a very small area that is susceptible to light. To make it more susceptible without just raising the background noise the transistor generates, the optically susceptible area must be enlarged. Classical approaches are lenses or mirrors. I decided to build a small reflector funnel that directs every beam of light, entering the opening of the funnel, to the transistor. Fig. 3: The reflector funnel I made the funnel from cardboard that is quite rigid and easy to cut. In total I made 4 triangular shaped planes, covered with reflecting foil and glued them together. At the bottom of the funnel I placed the electronic circuit with the phototransistor “looking” directly into the funnel. The opening of the funnel covers an area of 0.01 m2. Thus the optical area of the transistor of 1.96 10-5 m2 theoretically is enlarged by the factor 509. In practice this factor will be lower due to inevitable reflection losses. The casing for the electronic circuit was realized by pieces of double sided copper plated PCB material, soldered together. This way I obtained an optimal double shielding against electromagnetic interference. Fig. 4: The completed SkySound receiver I made a simple stand from wood with a joint to adjust the inclination of the receiver that looks much like the feeding horn of a satellite LNB receiver. The foot of the stand contained a magnet, so I could attach the whole receiver easily to a metal plate on the window sill. I placed the receiver in my attic where it overlooks the surrounding area from a height of 12m. The inclination angle against the horizon is almost zero, making the receiver looking at the lower sky roughly at horizon level. I connected the receiver to my Windows-10 Notebook with WIFI connection to my home router and used the application “butt” to stream the signal to the varanormal shoutcast server where it was relayed back as “Stream 4” which is reserved for my experiments. Investigation results I started my investigations by the end of 2021 and they are still ongoing. In the previous chapter I already described some of the experiment parameters that I want to list here. List of experiment parameters Parameter Value Height of installation 12m Inclination angle 0° (facing horizon) Viewing angle of funnel approx. 40° horiz. and vert. Facing sky direction North (0°) Stream encoding MP3 44.1 KHz Coordinates of location (Google maps) 53.682491442051194, 9.668064790784461 Table 2: list of experiment parameters General observations I quickly observed some general characteristics of the received signals and the properties of the receiver. Basically the signal was very low at night and very strong at day. This is no surprise because more light causes more current and more current causes more noise in the transistor. This is a rule of thumb in the world of physics. The question is if there would be an additional amount of noise coming from the quality of the light from the sky and not its quantity. We keep this consideration aside for the moment. In any case, the light caused huge signal level changes during sunset and sunrise that were remarkable. Fig.5: Signal level increase at sunrise from february 06-02-2022 (dd-mm-yyyy) At the time I was doing my investigations the sun was rising around 07:45 CET (06:45 UTC). You can see very clearly the non-linear “trumpet” shaped signal rise taking place over a time period of 6 minutes. Fig.5: Signal level decrease at sunset from february 06-02-2022 (dd-mm-yyyy) Sunset these days is around 17:45 CET (16:45 UTC). More or less you can see the sunrise characteristic going backwards in the image above. The shape of the signal is not as beautiful as the “trumpet” at sunrise. You must take into account that clouds passing by and all kinds of precipitation are overmodulating the light from the sun. I recorded and post processed the signals. What I found out was that especially at those transitional periods around dawn the signal contained a lot of spirit voices. At broad daylight there was heavy noise but low pk modulation and at night the signal level was too low for a reasonable signal gain. For the post processing I used Wavepad (NCH). My standard procedures for signal processing were Adding reverb (500ms decay time, 75ms predelay, 50% diffusion, 0dB wet signal gain, -6dB dry signal gain) -3 dB multiband noise gating 20% auto spectral subtraction signal normalization I exported some audio samples, processed with the methods listed above. Of course interpretation of the content of spirit messages is always subjective. Moreover every ITC researcher develops a kind of clairaudience based on the methods she or he is using. The result is that others often can not hear what she or he hears. If the evaluated samples do not have Class A quality they are often interpreted as pareidolia. Thus I won’t blame anyone for evaluating the following examples this way. In order not to influence the reader's perception I provided my interpretation of the samples in a table to be found in the annex of this document. I just want to say that I identified the language as German. Link to audio sample SkySound-A1 SkySound -A1.mp3 SkySound-A2 SkySound-A2.mp3 SkySound-A3 SkySound-A3.mp3 SkySound-A4 SkySound-A4.mp3 SkySound-A5 SkySound-A5.mp3 Table 3: List of exported audio samples from initial SkySound sessions The Stream 4_13 event At February 12, 2022 I was still lying in bed in the morning. It was Saturday and the sunrise had started around 07:45. At 08:17 I decided to switch on the streaming audio app “MR” on my smartphone and recorded the SkySound signal while listening to it with closed eyes. The recording app “MR” named it “Stream4_13”. After 11 seconds I suddenly heard strong bursts breaking through the noise. First it just appeared to me like a periodic interfering signal but then I recognized voice fragments stronger than I ever heard them directly in noise. Fig.6: Envelope of raw signal at the Stream 4_13 event After 8 minutes I stopped the recording and listened to the signal I had picked up. It was full of weird sounding voice breakthroughs. Post-processing of this signal was easily done because of the superior pk-modulation. I applied spectral subtraction based on a noise sample and what came out is visible on the next image. Fig.6: Envelope of processed signal at the Stream 4_13 event This processed audio contained surprisingly loud and partially clear messages overlayed by humming sounds. Stream 4_13 audio samples Raw signal Stream4_13.mp3 Processed signal Stream4_13-processed.mp3 Table 4: Stream 4_13 recordings I leave it up to the reader to make up his own mind upon the results. I evaluated them as most amazing. The interpretation of the voice contents gave me the impression that the spirits had tried something to surprise me or they did something that even surprised themselves. I am not sure about this. Investigation of other light sources At the beginning of the previous chapter I already pointed out the question if the received noise is more a function of the light quantity coming from the sky or, at least partially, also a result of the quality, namely the spectral composition, of the light itself. It is always good to have something to compare your results against to draw further conclusions. So I decided to stimulate the phototransistor with other light sources and to see what comes out. I replicated the SkySound circuit on a breadboard and attached different LED’s to the phototransistor that could be regulated in terms of their light intensity. My first goal was to find out if different LED’s are producing different light intensities and noise characteristics under the same conditions. So I concluded I must operate the LED’s in a way that they are supplied with the same amount of power to make their outputs comparable. Making recordings and evaluating and comparing the results in terms of pk-modulation was also part of the experiments. For the post processing of the recordings I used two different methods in NCH Wavepad. Description of applied post processing methods in NCH Wavepad Id Method Comments #1 Adding reverb (500ms decay time, 75ms predelay, 50% diffusion, 0dB wet signal gain, -6dB dry signal gain) -3 dB multiband noise gating 20% auto spectral subtraction signal normalization This method is a compromise between denoising and the generation of artifacts as they are typical for software-based denoising algorithms. As a result the signal integrity of the outcome is still acceptable but the resulting pk-modulation is not at optimum. #2 Spectral subtraction based on a grabbed noise sample signal normalization This method optimizes the pk-modulation. The downside is that the signal loses much of its original integrity. Nevertheless, voices can be identified easily in general by repeatedly listening to short sequences. Table 5: Description of post processing methods Experiments with a yellow LED A yellow LED is a very rough approach to sunlight since it’s spectrum is limited and does not cover ultraviolet frequencies. I ran the LED with a current of 4 mA. This is a very practical value that gives a reasonable amount of light while keeping the heat loss in the LED low. Because the SSM2167 circuit imposes variable amplification and audio compression on the received signal, I tapped the signal right before this stage (Pin 5 of SSM2167) to ensure the amplification of the signal is stable. The noise signal was measured with the RMS function on my oscilloscope. Here are the values I measured. Results gained from yellow LED Parameter Value IF: LED current 4 mA UF: Forward voltage on LED 1.9 V Calculated input power (UF x IF) 7.6 mW Noise level 36 mV Table 6: Measurements with yellow LED Then I made a recording of a 1 minute length of the noise signal. I could already hear the pk modulation as a faint voice burst on the noisy ground floor. Recordings made with yellow LED Description Link to audio file Raw recording (MP3) Yellow_LED_4mA.mp3 Recording post-processed by #1 Yellow_LED_4mA_1_processed.mp3 Recording post-processed by #2 Yellow_LED_4mA_spectralSubstract.mp3 Table 7: Recordings made with yellow LED The results are showing a fairly strong pk-modulation but rather poor voice quality. Experiments with a ultraviolet LED A UV LED shows a much higher spectrum of frequencies compared to yellow LED. Basically ultraviolet light is higher in energy too. To make the results comparable to the ones I gained from the yellow LED I measured the forward voltage across the LED and recalculated the LED current so that the power consumption of the UV-LED would be identically compared to the yellow LED. Here are the values I measured. Results gained from yellow LED Parameter Value IF: LED current 2.5 mA UF: Forward voltage on LED 2.96 V Calculated input power (UF x IF) 7.6 mW Noise level 40 mV Table 8: Measurements with UV-LED You see that within some tolerance considerations the emitted noise signal is the same compared to the yellow LED Then I made a recording of a 1 minute length of the noise signal with the UV-LED. My first impression was that it was not different from the yellow LED. Recordings made with UV LED Description Link to audio file Raw recording (MP3) UV_Diode_2p9mA.mp3 Recording post-processed by #1 UV_Diode_2p9mA_processed.mp3 Recording post-processed by #2 UV_Diode_2p9mA_spectralSubstract.mp3 Table 9: Recordings made with UV-LED Again the results are showing a fairly strong pk-modulation but rather poor voice quality. My evaluation was that there is no substantial difference between the results obtained from a yellow and a ultraviolet LED. The logical conclusion is that the pk-modulation is not bound to the spectral composition of the light whether it is yellow, ultraviolet or sunlight from the sky. Narrowing the focus Another question that came up during my investigations was if the noise with this substantial amount of pk-modulation was a product of light at all. Phototransistors are producing noise not only from light shining on their substrate but also from any current that flows into the base lead of the transistor causing it to draw collector current. That was easy to test. I disconnected the LED from the power supply and wired different resistors from the phototransistor base to the power supply starting from 680 K ohm up to 15 M ohm. There was almost no noise perceivable, especially not that quality of agile noise interspersed with cracks and bursts. My conclusion was that for the pk-modulation to take place, the transistor needs light but of no special spectral composition. However I was aware that I did not take into account another crucial component, the SSM2167 circuit. Due to its non-linear behavior it imposes a substantial amount of transformation on the signal. I could not exclude that it could contribute to the observed effects. In order to find an answer to this question I changed my experimental setup. According to my measurements the SSM2176 roughly added an amplification of 20 to the system. So, I forked the signal behind the preamplifier and routed it to the SSM2167 or another amplifier stage with an amplification of 20. Thus the resulting signal would always show the same level but one time processed by the SSM2167 and another time just amplified linearly without compression or noise gating. Recordings made with UV LED and with/without SSM2167 Description Link to audio file Recording with SSM2167 UV_with_SSM.mp3 Recording without SSM2167 UV_without SSM.mp3 Table 10: Recordings made with UV LED and with/without SSM2167 Fig.6: Signal without (left) and with (right) SSM1267 I think you can see that there is more modulation visible in the right image that shows the signal processed by the SSM2167. The sound characteristics of the recordings support this observation. The recording of the sample with SSM1267 involved contains some vast and abrupt level changes. These are due to adjustments I did on the potentiometers connected to the circuit and are not related to pk-modulation. Obviously the non-linear behavior of the SSM1267 adds a substantial amount of pk-modulation to the signal while the question, if the processing this circuit does is amplifying the already present modulation or if the circuit itself is a door for pk-modulation itself, remains unanswered. Due to my experience every kind of signal transformation that affects the signal structure is a gate for pk-modulation. Alternative post processing methods So far the post processing methods I applied are standard methods from the signal enhancing software toolbox. One of the researchers in our Varanormal team developed a software that is based on AI algorithms and is able to outperform the classical denoising and filtering techniques under certain conditions. Because I already have the last version of this fantastic application installed I wanted to give it a try. In the following you can see the parameter settings I used. Fig.7: Parameter settings for the machine learning algorithm The parameters were carefully optimized by controlling the output by ear while always only changing one parameter. After reaching the optimum for every parameter the output level of the SkySound device was altered up and down to see if any more improvements could be made. Basically I got a decoded signal that had all properties of the human voice especially in terms of rhythm. I really obtained just one continuously speaking voice instead of a bunch of parallel talking spirits like I got from earlier experiments. Nevertheless the weird metrics of spirit voices remained . Often the pronunciation of syllables is unusual compared to human speech. Some phonemes are enlarged and other shorted unnaturally. I think this is due to the spectral material the spirits find in the signal entropy we give them to work with. They seem to try to put the right words in where these match the actual spectral composition and signal transitions they are observing. However, very often the created vowels and consonants do not match the needed length to form normal speech and that is why the speech sounds so weird. I extracted lots of samples I want to show here. Again you can find my interpretation of the messages in the annex. Audio samples obtained from machine learning Sample Language Sample-B1 Sample-B1.mp3 German Sample-B3 Sample-B3.mp3 German Sample-B6_E Sample-B6_E.mp3 English Sample-B7_E Sample-B7_E.mp3 English Sample-B9_E Sample-B8_E.mp3 English Sample-B8_E Sample-B9_E.mp3 English Sample-B9_E Sample-B10.mp3 German Sample-B10 Sample-B12.mp3 German Sample-B12 Sample-B14.mp3 German Sample-B14 Sample-B21_E.mp3 English Sample-B21 Sample-B21.mp3 German Table 11: Audio samples extracted with machine learning algorithm Final tests and new prototype I did some more tests to support my theory about the origin of the pk-modulation. First I replaced the phototransistor with a light dependent resistor (LDR), a PTC (positive temperature coefficient resistor) and a CCR (carbon composition resistor). I used all these components in earlier ITC-experiments. The CCR showed a bit of pk-modulation but not as good as the phototransistor. Another important part of the circuit I identified to participate on the pk-effect is the self-adjusting transistor stage T2. It's original purpose is to adjust the receiver automatically with changing light conditions. Since working with an LED provides more or less constant light intensity there is theoretically no need for such a circuit. But for some reason it turned out that it adds-up substantially to the pk-effect. So I kept it. At the end I optimized the usability of the SSM 2167 module. It worked very well by replacing the potentiometers for signal compression and noise gating by fix resistor values and do the adjustments only with the potentiometer to control the light intensity of the LED. My findings were the base for a new prototype I made working with a phototransistor and a UV LED. Fig.8: Breadboard layout of prototype Fig.9: Working prototype After making the breadboard layout and the final prototype I recognized they don’t behave identically. The finalized prototype showed large spikes with pk-modulation characteristics breaking through with high energy. They were literally bombshelling the noise gate of the SSM 2167 causing it to full open the gate at every spike that appeared like a small explosion. I could exclude statics and radio interference because the effect grew with the light intensity and at least appeared to me to have the same rhythm as the voices. Sound example of the Prototype: Photon Bridge Prototype.mp3 I processed this recording a bit with reverb and voice pitch. I think at least you should be able to identify a voice although it is so overdriven that you cannot understand it: Noise Explosion Processed.mp3 Conclusions I think this paper is interesting because it does not show a linear and straightforward approach from an idea towards a realization. Instead you can see something that is very typical for the research work in ITC. You are starting with a certain idea or phenomenon to be investigated. While following this path by using scientific methods (experiments, documentation and conclusion) you finally see that the original question is not in your focus anymore. New questions are showing up and very quickly you are following a quest of a yet unknown goal. Someone else now is leading the path you are walking on and in the end you are given something that is not what you were looking for but maybe something even more precious. I was looking for hidden modulations in the light from the sky but what I found was a combination of a noise source with an audio circuit that was designed for a totally different application. Working together they are creating the best quality of continuously speaking voices I ever was able to obtain. Nevertheless there still is a chance that the light from the sky contains more magic than an ordinary LED may cause. At least I don't yet got a Stream4_13 event with a UV-diode. So maybe there is more to come. For now, I just can say “Thank you” to the hereafter. Annex Audio identifier Interpretation Translation SkySound-A1 “Hat Bedeutung gewaltig” “Has tremendous impact” SkySound-A2 “Wir sind schon alle” “We are all here already” SkySound-A3 “Du bereits unsere Achtung” “You have our respect already” SkySound-A4 “Versinken im Feuer” “Drowning in fire” SkySound-A5 “Fenster nun ist fertig” “Window is ready now” Table A1: Interpretation of audio samples in chapter “General observations” Audio identifier Interpretation Translation Sample-B1 “Mich hören jetzt noch andere reden” “Now others can hear me too” Sample-B3 “Nimmt die Freiheit” “Takes freedom” Sample-B6_E “Harm merit” N/A Sample-B7_E “More guess” N/A Sample-8B_E “What a funny kid” N/A Sample-B9_E “Manhatten” Sample-B10 “Es sprechen Erwachten” “The Awakened are talking” Sample-B12 “Und ich war Magie” “I was magic” Sample-B14 “Und wir brauchen Netflix” “And we need Netflix” Sample-B21 “Kann er auch. Verhalten sich schwach” “He can do too. They behave weak” Sample-B21_E “Merryful Crocket” Table A2: Interpretation of audio samples in chapter “Alternative post processing methods” Sample-B18.mp3 Sample-B19.mp3 Sample-B20.mp3 Sample-B22.mp3 Export-1.mp3 Sample-B2.mp3 Sample-B4_E.mp3 Sample-B5.mp3 Sample-B11.mp3 Sample-B13.mp3 Sample-B15.mp3 Sample-B16.mp3 Sample-B17.mp3

Interesting Alain. I have the same problem the other eay around and can't hear what you heard. There is a theory we sometimes discuss in the research team that our perception is the last stage where the information becomes finalized. Only after our perception the information transfer is completed. This means that the same recording can manifest into different messages in the ears of different people. Sometimes I think this is true because the perceptions are so different.

Here some audios from my experiments from today. I completed the first version of the phototransistor noise circuit. I chose all voices in english. Fast blinken Fast blinken.mp3 My woman My woman.mp3 One hole One hole.mp3 Too much is harmful Too much is harmful.mp3 Too much Too much.mp3

Wow that sounds like a very sophisticated improvement. I'm curious! In the meantime I will try to squeeze out what is possible from your code after I finished the latest phototransistor design before I will continue with Sonias Lightbridge device manufacturing. If I should find an easy to use solution for a one click start I will let you know.

Today I reinstalled python. I trashed it while trying to start your app by creating a direct link from my desktop. Reinstalled also the tf25_nongpu environment. It worked smoothly so the installation process is stable due to my experience. I'm actually testing your app with phototransistor noise sources. One setup I'm working on seems to be fairly promising. In my tests I found out the importance of the low threshold and the tone threshold parameter. I really could tweak your sw by carefully adjusting them. By the way, the automatic storing and retrieving of parameter settings is really nice! Moreover I am curious if I could create a link to start your app inside the anaconda prompt by using windows PowerShell.

So true!

Yes. Of course this sheds a completely new light on the objectivity of messages and how to separate them from pareidolia. Maybe the red line is not running between reality and pareidolia, maybe it just separates useful pareidolia from useless?

Exactly! Apart from some very good direct microphone voices we always need software to process the raw signals to even get the chance to distill the original information from it. However, sometimes I think that the the sw processing is not just about extracting hidden information from the recordings but even more that sw is part of the pk effect and thus one factor in the signal transformation chain that starts in the hereafter and ends in our world. Moreover I frequently come to the point that even our perception is part of the transformation chain and that the 'objective' messages is shaped just finally during our perception. But I'm not really sure about this...

I like this clean conceptual approach you are presenting here, Michael. Very well done! I'd like to mention two other possible ITC-components. The first one is fragility. IAt first sight I know fragility seems similar to sensitivity however I think fragility offers something that sensitivity does not show. A sensitive detector detects the desired signal but usually adds noise to the amplification process. In a fragile system an energy quantum coming the hereafter can be multiplied by avalanche effects when the detector system is flipping from one state into another. In this case the signal can outstand the noise ground floor sometimes. I observed this behavior very much with coherers and also your experiments with the whistler you used an operational amplifier in a weak state of feedback, if I remember correctly. Another thing is utilizing chaotic behavior of systems. We see this in white or pink noise, all kind of uncontrolled feedback, the complex impulse patterns in the VISPRE and many more. I suppose the usability of those effects is lying in the fact that everything is already "moving". Spoken literally, it certainly takes more energy to set something in motion as affecting something that already is in motion. I also would like to add something to your description of driving energy. Spoken very generally this energy already should appear in the shape you want the resulting signal to be. We observe this very much with microphone voices. I often got messages that had the shape or characteristic of the background sound I was providing. As I was typing on my keyboard in the background I got voices with clicking sounds while pink noise makes spirit voices sound croaky.

Great!! Thanks for this link. I love those movies. Gonna watch it next weekend.

Avast sent me an advertisement for their online privacy campaign. They are very concerned about my privacy! In their message my email client blocked a spy tracker! Now you know why Avast is so concerned about my privacy.

Today I found a way zo run your program. First I tried to start python from a windows console with your program as a parameter but Windows didn't know the path yo python. Then I started Anaconda, entered the tf25_nongpu environment, opened a console in Anaconda and started python with your program. Et voilà, it works!!!!

Ok, so it's better to start your app directly from a command console with phython? When I do this is the respective folder where the application resides, am I in the correct environment automatically?

Yes and so I did. However I ran the application in Spyder and I cannot rule out that I started Spyder in another environment. Will check this tomorrow.

I activated the tf25_nogpu. Will check the tensorflow version but isn't this defined in the environment.py?

Hi Michael! I installed everything according to your very good description. Now I got the following erro while starting "itc_translator.py" runfile('C:/Users/User/Documents/Code/Python/Michaels ML/itc_translator.py', wdir='C:/Users/User/Documents/Code/Python/Michaels ML') Traceback (most recent call last): File "C:\Users\User\AppData\Local\Temp/ipykernel_15308/3073042103.py", line 1, in <module> runfile('C:/Users/User/Documents/Code/Python/Michaels ML/itc_translator.py', wdir='C:/Users/User/Documents/Code/Python/Michaels ML') File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\debugpy\_vendored\pydevd\_pydev_bundle\pydev_umd.py", line 167, in runfile execfile(filename, namespace) File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\debugpy\_vendored\pydevd\_pydev_imps\_pydev_execfile.py", line 25, in execfile exec(compile(contents + "\n", file, 'exec'), glob, loc) File "C:/Users/User/Documents/Code/Python/Michaels ML/itc_translator.py", line 355, in <module> custom_objects={'alpha':alpha}) File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\tensorflow_core\python\keras\saving\save.py", line 146, in load_model return hdf5_format.load_model_from_hdf5(filepath, custom_objects, compile) File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\tensorflow_core\python\keras\saving\hdf5_format.py", line 166, in load_model_from_hdf5 model_config = json.loads(model_config.decode('utf-8')) AttributeError: 'str' object has no attribute 'decode' Error in callback <bound method AutoreloadMagics.post_execute_hook of <autoreload.AutoreloadMagics object at 0x000002255E30D448>> (for post_execute): Traceback (most recent call last): File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\IPython\extensions\autoreload.py", line 538, in post_execute_hook _, pymtime = self._reloader.filename_and_mtime(sys.modules[modname]) File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\IPython\extensions\autoreload.py", line 184, in filename_and_mtime if not hasattr(module, '__file__') or module.__file__ is None: File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\tensorflow\__init__.py", line 50, in __getattr__ module = self._load() File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\site-packages\tensorflow\__init__.py", line 44, in _load module = _importlib.import_module(self.__name__) File "C:\Users\User\anaconda3\envs\tf21_nogpu\lib\importlib\__init__.py", line 127, in import_module return _bootstrap._gcd_import(name[level:], package, level) File "<frozen importlib._bootstrap>", line 1006, in _gcd_import File "<frozen importlib._bootstrap>", line 983, in _find_and_load File "<frozen importlib._bootstrap>", line 965, in _find_and_load_unlocked ModuleNotFoundError: No module named 'tensorflow_core.estimator'

Yes you are but you made the world a little better now!

Hello Sharon. Could you post your recording here? I have several methods to process it including slowing down.

I have my ectoplasm thrower ready. Since I already have Anaconda running I can directly jump into testing. I bet I will have some troubles with the phython libs as usual. Will check it out tomorrow.