Jump to content

Michael Lee

iDigitalMedium Research Team
  • Content Count

    140
  • Joined

  • Last visited

  • Days Won

    29

Michael Lee last won the day on January 31

Michael Lee had the most liked content!

Community Reputation

86 Excellent

1 Follower

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Jeff, Generally speaking, I haven't spent much time on the phonetic typewriter as I've switched to machine learning assisted "direct voice." However, I agree with you that this work does subtly reveal the abilities and limits of spirit influence for a given hardware system.
  2. Sean, Welcome! I see you have an impressive list of publications. Hopefully, this forum will spur some new collaborations in this important, but sparsely studied field. -michael
  3. Jeff, I should point out that spirits don't have to share their voice directly - they are also capable of activating phonemes or even converting their voice into frequency space. What they are limited to, however, it appears to me, is spikes of energy, at least with the hardware we have given them.
  4. The general method is to figure out how a spirit voice is corrupted when we hear it directly from our noise-generating devices and then train a machine learning model to reverse the effect. Specifically, I have found at least three corruption processes: (1) the spirit signal is often heavily buried in noise (i.e. additive noise). 2) the spirit signal is "quantized" or in other words it sounds like it's (e.g.) 2 to 4-bit audio vs. clean 16-bit audio. 3) the signal is "sparse" or missing a lot in time - instead of hearing a smooth waveform, we are randomly getting 10-20% of the ti
  5. Here's some "greatest hits" from the last year : youre_getting_good_voice_now.wavtotally_fine_with_us.wavsuch_a_great_signal.wavthis_is_so_exciting.wavisnt_portal_so_fun.wavstop_mixing_it_directly.wav
  6. Jeff, Although a little off topic from the original question, I agree with your understanding that spirits utilize the sounds available to them. This can actually be stated mathematically as convolution: roughly speaking they can slightly change the volume of the individual frequency components of environmental sound. I recently experimented with two microphones and several projected sounds (testing one at a time) from a speaker. Then using microphone cancellation and machine learning to disentangle the original voice. I like to think of the played sound as having two function
  7. One technology that the commercial space has been exploring is microphone arrays for smart devices like the Amazon Echo. The idea is that multiple microphones better cancel out environmental noise and reverberation leading to a clearer voice for speech recognition. What would be the benefit for ITC? Localized spirit voices? Improved signal-to-noise? It's not easy to make a microphone array, so the argument for pursuing this would have to be compelling. BTW, I have played with two microphone setups. This is easy to do and does help with sound cancellation if there is a localized audio
  8. No problem. We have a technique for converting Python scripts to executables. The caveat is each one takes up 700 MB on the user's hard drive.
  9. Because the stream of phonemes is fixed (just playing a WAV file) you can compare different runs to see if you're getting different messages. If you're getting the exact same messages each time, then you know the gate isn't set correctly, etc. I generate the scrambled phonemes with my own Python script, which I've shared here. It ensures that each blip is the same magnitude and clipped so that, in theory, none of them should trip the noise gate without help from an extra noise source. Getting Python scripts running on your machine requires Anaconda3 and a few module installations, bu
  10. Yes. I assume it's the point where the current is in between saturation and zero.
  11. Let's see what others think. My feeling is while a forum is fairly small in number of people, it shouldn't have a zillion rooms. With my machine learning - assisted audio ITC, I call it audio ITC, but it really is a form of speech synthesis, too, where the machine learning is replacing the weakly detected speech with something more solid.
  12. Although I don't know if it's that good of an ITC entropy / noise source, I think it is pretty cool that we can actually hear the flow of photons from a laser or flashlight. Simply make a circuit of 48V (phantom power) -> 10K resistor -> PhotoTransistor (PT) -> Gnd (audio interface output). I use a BPW85A (purchaseable from Mouser for ~$1 each) Turn off all the lights Take a continuous light source (either a bright flashlight or pen laser) and point it at the PT face. Then slowly turn it away. You should hear a click sound as the voltage goes from high to near zero.
  13. The corona noise is a decent source for ML input but nothing extraordinary. It could have a quantum component due to reflectance/transmittance of the laser lens, but it could also have a noise component due to the driver, which is commonly considered to be more classical electrical noise (however, that's up for debate!). I had a delicate setup with two PDs at 90 degree angles of a beam splitter, that also yielded voices in ML. Maybe we could develop a "pocket" vibrometer for ITC/non-ITC use? It could be a better microphone in terms of sensitivity and locality. I was even imagining us
  14. Two more things Ive observed today. The laser shot sound is known as laser doppler. It is a type of interferometry where the beam reflects back on itself, yielding the ridge patterns, that with movement of the reflective material, creates a fast washboard effect. It is most easily achieved by shining the laser directly into the phototransistor and lightly wiggling the laser. The other thing I noticed is that my pen laser has an incoherent corona around the beam, that I can hear by pointing the beam just off center from any direction to the phototransistor. Its a mix of pink an
  15. To add, my blog is slowly walking through the noise conversion methods in my software. Simply put, I can turn anything into voice-like sounds. Now we have to listen closely and try to understand what they are saying and see if they are giving us advice on how to proceed.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.