I am currently working on a project that has been relatively simple so far. The main project is the transfer of data / messages through lasers using sound conversion. In short, the process is currently similar to this
- User enters message
- Message turns into binary
- For each 1 and 0 in the binary message, it reproduces the corresponding signal for the signal, which in my case is 250 Hz for 1 and 450 Hz for 0.
- The outgoing tone is transmitted through a stereo cable to an audio transformer equipped with a laser.
- The solar panel acts as a microphone and records the incoming “sound” as a file
- It plays the file back and reads the tones and tries to match each 250 and 450 Hz with 1 or 0 (where is my problem).
Until the actual sound processing is perfect, my current problem is the following.
I play tones every time for x time, at the receiving end it is recorded for y time, y time is cut out by sampling many times, and then the sample is analyzed by the sample, which then records each frequency. It is inefficient and inaccurate. I had a lot of problems, regardless of the time when I play tones, because he often hears a tone twice or does not hear it at all, which completely discards entire messages. I tried to compare the speed with which it is counted, each time each tone is played, but if you have not aligned, this will not work. I only had a few successful tests for messages such as "test" and "hello." I already studied bpsk and fsk, but it seems to me that I am already doing something similar, but I have a bad reception to decrypt all of this.
Python, , , . pyaudiere, pyaudio.
!
-Steve