Install

Three minutes, one folder

Every pack is a SOUNDS folder laid out exactly like the one on your radio's SD card. Copying it over replaces only the sounds in the pack.

  1. Download a pack. Pick one from the themes, or build a mix.
  2. Connect the SD card. Plug the radio into your computer and choose USB Storage (SD), or take the card out.
  3. Back up your sounds. Copy the card's existing SOUNDS/en folder somewhere safe, so you can go back.
  4. Copy the pack. Unzip it, then drag its SOUNDS folder onto the root of the card and choose to replace the existing files.
  5. Eject and listen. Unplug and power-cycle the radio. The boot sound is the first thing you'll hear.

What's in a pack

SD card └── SOUNDS └── en ├── armed.wav disarm.wav lowbat.wav clobat.wav flip.wav … ├── fm-1.wav … fm-8.wav flight mode 1–8 (1–8 rising notes) ├── rates1.wav … rates6.wav └── SYSTEM ├── hello.wav radio boot ├── telemko.wav telemetry lost ├── telemok.wav telemetry recovered ├── rssi_org.wav RF signal low ├── rssi_red.wav RF signal critical ├── thralert.wav throttle warning ├── swalert.wav switch warning └── timovr1.wav … timers, trims, trainer, errors

70 files per pack, all 16 kHz, mono, 16-bit PCM WAV: the format EdgeTX plays.

What stays the same

Packs only replace event sounds. Spoken numbers, units and telemetry readouts (SYSTEM/0000.wav to 0112.wav, volt0.wav and friends) are not touched, so "battery 3.7 volts" still sounds like your voice pack.

Files in the root of SOUNDS/en, such as armed.wav or lowbat.wav, only play when a Special Function or a Lua script asks for them. The SYSTEM sounds play automatically.

Using them in your model

In Model → Special Functions, add Play Track on a switch and pick the file: for example armed on SA↓ and disarm on SA↑. Flight-mode switches work well with fm-1 to fm-3: you hear one, two or three rising notes.

Installing a voice pack

A voice pack is the same idea, one level up: it is the whole SOUNDS/<language> folder, some 750 files, with every number, unit, alarm and callout EdgeTX can say.

  1. Download and copy. Unzip, drag SOUNDS onto the card root, replace existing files. The Dutch packs land in SOUNDS/nl, English ones in SOUNDS/en.
  2. Select the language. On the radio: Radio setup → Voice language. EdgeTX 2.10 or newer lists Dutch; if yours does not, update the radio firmware first, or keep the files in en.
  3. One voice per language folder. Two voices of the same language overwrite each other. Installing Petra over Hans simply replaces the files.

Hosted voices are 16 kHz mono, the same as the official EdgeTX packs, stored as lossless FLAC and converted to WAV in your browser when you download. The voices made by Stickbeats were generated with ElevenLabs, trimmed, peak-normalised, and then transcribed back with speech-to-text to catch mispronunciations. The phrase lists are in the voice repository: fix a translation there and the next build regenerates only that file.

Voice + theme

Voice packs and sound themes stack. The theme replaces the 70 alert files, the voice does everything else. On any hosted voice's page, pick a theme and download both as one zip. The mix builder does the same for an event-by-event mix.

Other languages

If your radio uses a voice language that has no theme folder yet, rename the theme pack's en folder to your language code, such as de, fr or nl, before copying.

Safety

Every theme has to pass the same checks before it's published:

Still, listen to your pack on the ground first, so you know what trouble sounds like before you're in the air.

How it's made

Every sound is written as MIDI, then rendered by a small custom synth: NES-style pulse and triangle channels, Game Boy wave tables, FM operators, Karplus-Strong strings, drum machines and a box of sound-effect generators. Orchestral and acoustic instruments come from the free GeneralUser GS soundfont via FluidSynth. The MIDI files come with each theme, so you can remix them. The code is on GitHub.