Tonedrop: Data Over Sound

Zelicra

About this app

Tonedrop sends messages, files and raw bytes between devices using sound alone. No Wi-Fi, no Bluetooth, no pairing, no accounts. One device plays tones; any device with a microphone writes them down.

SEVEN MODES
• Wave: the fastest. Hundreds of carriers at once (OFDM); on a link it adapts its rate and confirms delivery.
• Wave Far: for echoey rooms and distance.
• Wave Ultra: inaudible, 17.5–20.5 kHz.
• Robust: for people and noisy rooms, across a room.
• Fast: for agents and short bytes.
• Turbo: fast tones at close range.
• Stealth: near-ultrasonic, hard to notice by ear, and it can ride underneath music.
Every frame carries Reed-Solomon error correction and a checksum, so a message arrives intact or not at all, never garbled.

TONEDROP LINK: AI AGENTS TALKING OVER SOUND
A link opens in gibberlink's language (ggwave), so any gibberlink agent understands it. When the other side runs Tonedrop too, both switch to Tonedrop's own modes: faster, any length, files, and stealth.

UNIVERSAL TRANSLATOR
One microphone, many machine languages at once: Tonedrop in every mode, ggwave and gibberlink (with sender ids), and DTMF keypad tones, with Morse and tap code available as options. Each message appears once, labelled with its protocol. Type text and speak it in any of them, or save it as a WAV.

MORE
• Send files, compressed before sending.
• Broadcast loop: keep a message on the air so late listeners still collect it.
• Music carrier: hide a message inside a song.
• Morse, tap code and DTMF: type text to see and hear its code, or decode one live.
• Diagnostics: a self-test that proves the modem works on your hardware.

PRIVATE BY DESIGN
Tonedrop collects nothing: no accounts, no analytics, no ads, no servers. Audio is decoded on your device and never recorded or uploaded.

Sound is a broadcast: anyone nearby can hear and decode what you send, and messages are not encrypted. Do not send secrets over sound.

Open source under the MIT licence.

What’s new

Wave Ultra now reliably hears messages sent from the same phone. Listening uses a steadier audio path, so one damaged frame no longer spoils the next one.