Open Ground: voice lines
Open Ground is a small tactical war game: soldiers fight on real-looking terrain and shout short calls to each other. We are looking for real voices to replace our first, synthetic recordings.
Where the voices stand
14 languages and 3,342 takes in the game today (2026-10-27). Every recording below is synthetic. Real voices replace them one line at a time.
| Language | Takes | Voices | Soldier calls covered (groups / lines) |
|---|---|---|---|
| English (default) | 1,919 | 5 male soldier voices, civilians, radio | reference |
| Spanish (Spain) | 258 | 1 | 93% / 63% |
| Russian | 243 | 1 | 86% / 50% |
| Ukrainian | 188 | 1 | 69% / 36% |
| French | 164 | 1 | 62% / 41% |
| Spanish (Argentina) | 164 | 1 | 62% / 41% |
| Vietnamese | 87 | 1 | 35% / 21% |
| German (1944) | 87 | 1 | 35% / 21% |
| German (East Germany, 1983) | 73 | 1 | 27% / 18% |
| British English | 64 | 1 | 34% / 15% |
| Chinese (Mandarin) | 44 | 1 | 29% / 10% |
| Arabic (Levantine) | 26 | 1 | 12% / 5% |
| Arabic (Iraqi) | 19 | 1 | 8% / 4% |
| Nepali | 5 | 1 | 4% / 1% |
| Australian English | 0 | new | planned: starts with 10 takes |
English has five male soldier voices. Every other language has one, so a second voice per language is open work. A "strict" language is silent on a call it has no recording for, so thin languages sound sparse.
What we record
- Calls: "Contact!", "Medic!", "Reloading!", "Cover me!"... in two registers, shouted and whispered (a hidden squad whispers).
- Orders, answers and relays: "Moving!", "Roger".
- Murmurs when it is quiet, and remarks after a kill.
- Callouts: "RPG!", "Sniper!", "Armour!", "Incoming!", "Clear!".
- Hurt, prisoner and radio lines, then pain and death cries.
- Mission briefings and radio voiceovers (English only today).
Lines are short (one to five words, under 1.5 seconds), natural for a soldier of that country, and ideally come in three variants.
How to record
- Quiet room, no echo (clothes around you help). Microphone 15 to 20 cm from your mouth, a little off to the side. A phone recorder is fine if it saves WAV.
- WAV, mono, 48 kHz (44.1 kHz is fine), 24-bit or 16-bit. No effects, no noise reduction, no normalising: we process every voice the same way.
- Half a second of silence before and after each line. One line per file. Peaks between -12 and -3 dB, never clipping.
- Shouted calls: really shout, with your head turned slightly away from the microphone. Whispers: a real breathy whisper, close to the microphone.
- Two takes of each line: the main one and an alternative with another delivery.
- Name the files
NN_key.wav(for example03_reloading.wav) andNN_key_alt.wav.
How to proceed
- We send you a sample pack for your language: 10 lines with a first synthetic recording of each, as a reference for length and mood.
- You record the same 10 lines in your own voice, changing the wording if a real soldier would say it differently.
- You send the files back (below). We listen, process and put them in the game. Then we plan the next lines with you.
What you get
Alpha demo. Every contribution we use in the game earns you access to the Open Ground alpha. We email you the alpha demo a few weeks after your first contribution, so you can hear your voice in a real battle before anyone else.
Credit in the game is yours if you want it. Anything beyond this, and the licence for your recordings, is agreed with you in writing before your recordings are used.
Send your recordings
- Put your WAV files in one folder, with the filled
lines.csv(the column "text you said"), and zip it. - Upload the zip to WeTransfer or a similar service (please avoid large e-mail attachments).
- E-mail the download link to machingames@proton.me, with the name you want in the credits (or "no credit").
We agree the licence and the terms with you in writing before any recording is used in the game.