Acoustic channel modes
Two user-visible sub-modes are exposed in the app: a telephony-grade Basic profile and an inaudible Ultrasonic profile.
| Sub-mode | Carrier | Throughput | Effective range | Audible? | Best for |
|---|---|---|---|---|---|
| Basic (DTMF-class) | Eight tones, 697 / 770 / 852 / 941 Hz rows crossed with 1209 / 1336 / 1477 / 1633 Hz columns. Sixteen symbols, twenty-five symbols per second, Reed-Solomon RS(15,9) parity. | ~5 to 8 bytes per second after error correction. 10 to 15 seconds for a 5 to 50 character message. | Up to a few metres in a quiet room. Also survives a normal phone call or a Telegram / WhatsApp / Bale / Soroush voice note. | Yes (telephony tone burst) | Same-room transfers, harsh acoustic environments, transmission over a voice call, situations where the parties cannot see each other. |
| Ultrasonic | Multi-tone carrier in the inaudible 17 to 19 kilohertz band. Twenty symbols per second. | ~10 to 15 bytes per second. About 3 seconds for a 5-character message, about 6 seconds for 50 characters. | 30 cm to roughly 1 m. Does not carry beyond the room. | No (above adult hearing range) | Same-room transfers where modern audio hardware is present on both phones and an inaudible transmission is preferred. |
Values measured with the standard speaker and microphone of a current-generation mobile device at one metre under indoor conditions. Real-world performance varies with room acoustics, speaker quality and microphone response.
Choose a channel that works with your messaging platform.
If you need to send through a messaging app, this table shows what each platform does to the attached file. Pick the row, column and mode accordingly.
OKTransmitted without modification. Use Silent mode.
LossyUse Robust mode. Capacity will be reduced.
NoSelect a different channel for this platform.
| Platform | Photo (default) | File / document | Voice note |
|---|---|---|---|
| Signal | OK | OK | Lossy |
| Telegram | Lossy | OK | Lossy |
| Lossy | OK | No | |
| Instagram DM | Lossy | n/a | No |
| iMessage | OK | OK | Lossy |
| OK | OK | OK |
Most messaging platforms re-encode voice notes, so the acoustic channel rarely survives transmission in voice note form. For those scenarios, send the audio as a file attachment or use the steganographic channel instead.
Two devices. One shared password. No intermediary.
Install Taftan on both devices, agree on the password through a separate channel beforehand, and run a quick test transfer using whichever channel fits your situation.