streaming encoder/decoder for binary frequency-shift keying
https://en.wikipedia.org/wiki/Frequency-shift_keying
Work-in-progress. Works great if you pipe the encoder to the decoder, but still working on getting it working well in noisy / real-world environments!
Let's encode the string 'Hello!' into sound waves and then back again to a string:
var bfsk = require('binary-fsk')
var speaker = require('audio-speaker')
var microphone = require('mic-stream')
var opts = {
mark: 884,
space: 324,
baud: 5,
sampleRate: 8000,
samplesPerFrame: 320
}
var encode = fsk.createEncodeStream(opts)
var decode = fsk.createDecodeStream(opts)
// pipe to your speaker
encode
.pipe(speaker())
// receive sound from your speaker and decode it back to text to print out
microphone()
.pipe(decode)
.pipe(process.stdout)
// write a message!
e.end('heya!')This will make your computer scream garbage at you (make sure your speakers and microphone are on!) briefly, but it should then output
hello warld!
var bfsk = require('binary-fsk')Returns a Transform stream. Pipe text or other interesting data into this, and it will output audio data. You can pipe this to your speakers.
Valid opts include:
mark(required) - the "mark" frequench, in hertz. This frequency is used to signal a '1' bit.space(required) - the "space" frequench, in hertz. This frequency is used to signal a '0' bit.baud(required) - the number of bits to transmit/expect, per second.sampleRate(optional) - the number of samples per second. Defaults to 8000.
Returns a Transform stream. Pipe audio data into this, and it will output the original data that was passed to the encoder.
This takes the same opts object as the encoder. For proper results, use the
same options as your encoder.
With npm installed, run
$ npm install binary-fsk
ISC