It's a pastebin, except text/files are encoded into .wav files using real FSK modem audio. Image sharing is supported via Slow-Scan Television (SSTV), a method of transmitting images as FM audio originally used by ham radio operators.
Everything runs in the browser with zero audio libraries and the encoding is vanilla TypeScript sine wave math: phase-continuous FSK with proper 8-N-1 framing, fractional bit accumulation for non-integer sample rates, and a quadrature FM discriminator on the decode side (no FFT windowing or Goertzel), The only dependency is lz-string for URL sharing compression.
It supports Bell 103 (300 baud), Bell 202 (1200 baud), V.21, RTTY/Baudot, Caller ID (Bellcore MDMF), DTMF, Blue Box MF tones, and SSTV image encoding. There's also a chat mode where messages are transmitted as actual Bell 103 audio over WebSocket... or use the acoustic mode for speaker-to-mic coupling for in-room local chat.
thebuilderjr•6h ago
A feature that would make this even more useful as a debugging or teaching tool would be a decoder inspector: recovered bytes, framing errors, and maybe a small waterfall or discriminator view while decoding. For acoustic transfers especially, that would make it much easier to see whether a failure came from timing, levels, or just picking the wrong protocol.
a13x57•5h ago