The quality of meeting minutes depends directly on the quality of the captured audio and the context you provide. A few simple habits make all the difference — here's what matters most. (For more detailed questions, see the help page.)
The built-in microphone on your phone works fine for 2 to 4 people around a table. For larger groups, a 360° conference speakerphone significantly improves transcription accuracy: all voices are captured evenly, even those furthest from the device.
| Room | Participants | Recommended model | Indicative price |
|---|---|---|---|
| Small (office) | 2–4 | Jabra Speak2 55 | ~130 € |
| Medium (standard meeting room) | 4–8 | Jabra Speak2 75 or EPOS Expand 40 | ~250–280 € |
| Large (boardroom) | 8–12 | Anker PowerConf S500 | ~200 € |
| Extra large (assembly hall) | 12–20 | Anker PowerConf S500 ×2 (paired) | ~400 € |
For long meetings (3 hours or more), connect the speakerphone via USB rather than Bluetooth, to avoid any disconnections.
If some participants are in the room and others are joining remotely (Teams, Meet, Zoom), your phone's microphone will only pick up the voices in the room. Instead, on a computer connected to the video call, use the "Mic + browser tab" mode in the Web version: Geremy captures both the remote voices (from the video call tab) and those in the room (the computer's microphone, or a connected speakerphone).
The more context Geremy has, the better the result. A few minutes of preparation — before or even after recording — makes a real difference.
At the start of the meeting, ask everyone for a quick round of introductions: «Hello, I'm Marie, project manager.» This 3-second phrase per person lets Geremy link each voice to a name — otherwise you'll end up with «Participant 1», «Participant 2» (which you can correct afterwards, but it's better to start on the right foot).
All this information is optional, and it is taken into account even if you add it after the recording (during generation or regeneration).
Geremy offers over 50 categories that automatically adapt the structure and level of detail to your context (team meeting, summary, decision log, annual review, municipal council or general assembly minutes, consultation, project debrief…). You choose the category when you launch or regenerate a report.
If you are recording with a dictaphone or an external app before uploading the file to Geremy, here are the settings that ensure the best transcription.
| Priority | Format | Comment |
|---|---|---|
| Ideal | WAV 16-bit PCM | Lossless, maximum quality (~110 MB/h) |
| Very good | FLAC | Lossless, 50–60% lighter than WAV (~55 MB/h) |
| Acceptable | MP3 ≥ 192 kbps | Good size/quality trade-off (~17 MB/h) |
| Minimum | MP3 128 kbps | Sufficient for speech (~11 MB/h) |
| Avoid | MP3 < 64 kbps, WMA | Consonant loss, degraded transcription |
Geremy accepts MP3, MP4, M4A, and WAV files, up to 4 hours or 400 MB per file. In any case: a clear recording in a quiet environment will always matter more than the format — an MP3 at 128 kbps captured in silence is worth more than a WAV recorded in a noisy room.