How loud is a singing drone? We measured it
Everyone asks, so we borrowed an SPL meter and measured motor melodies across builds, distances, and octaves. Real numbers, including the disappointing ones.
"How loud does it get?" is the most common question we get after someone hears a quad sing for the first time. Until now our answer was "eh, conversation-ish?" — so we borrowed a calibrated SPL meter (a class-2 unit, A-weighted) and spent an afternoon measuring properly.
Short version: a 5-inch quad playing a melody peaks around 78 dB(A) at one metre. That's louder than we expected, quieter than you hope, and there's a lot of interesting texture in the details.
The setup
- SPL meter at fixed distances (0.5 m, 1 m, 5 m, 10 m), tripod-mounted at frame height
- Quiet suburban garden, ~38 dB(A) ambient, no wind
- Test melody: our Mountain King arrangement (steady eighth notes, octave 5–6 — a fair "typical song")
- Measurements are peak values during playback, three runs averaged
Results by build
| Build | 0.5 m | 1 m | 5 m | 10 m |
|---|---|---|---|---|
| 5" freestyle (2207/1950KV, carbon) | 84 dB | 78 dB | 64 dB | 57 dB |
| 7" long-range (2806.5, carbon) | 87 dB | 81 dB | 68 dB | 61 dB |
| 3.5" cinewhoop (ducted, plastic) | 71 dB | 64 dB | 49 dB | below ambient-ish |
| 75mm tinywhoop (0802) | 58 dB | 51 dB | lost in ambient | — |
For calibration: normal conversation is ~60 dB at 1 m, a vacuum cleaner ~70 dB, city traffic ~80 dB.
So a 5-inch quad at arm's length genuinely competes with a vacuum cleaner. At 10 metres it's down to "someone talking nearby". The tinywhoop numbers confirm what our ears already told us: charming at desk distance, gone at field distance.
Results by octave
Same 5" build, same melody transposed across octaves:
| Melody centred on | 1 m peak |
|---|---|
| Octave 4 | 71 dB |
| Octave 5 | 78 dB |
| Octave 6 | 79 dB |
| Octave 7 | 74 dB |
Octaves 5–6 win, which matches the advice we've been giving on feel alone. Octave 4 loses ~7 dB — that's half the perceived loudness. The high octave-7 falloff surprised us less after watching the meter: the frame radiates those frequencies more directionally, so unless the meter sat exactly in the beam, energy missed it. Wind eats these frequencies first outdoors too, as covered in field-testing your tune.
Doubling motors ≈ +3 dB
One motor playing the melody: 75 dB at 1 m. All four in unison: 78 dB. That's the textbook +3 dB per doubling of sources, and it's why "put the same melody on every motor" is a legitimate loudness strategy when you don't need harmony. A four-voice arrangement spends that loudness on richness instead — the Canon in D measures quieter per-note than unison Mountain King, but nobody who heard both preferred the louder one.
What didn't matter
- Battery voltage: 4.2 V vs 3.6 V per cell measured within 1 dB. (Pitch stability is a different story — see the firmware comparison.)
- Props on vs off: within measurement error. Feels like it should matter; doesn't.
- Fresh vs worn bearings: no SPL difference, though worn bearings add an audible rattle the meter doesn't capture but your ears definitely do.
The honest takeaway
A singing drone is a close-range instrument. The physics caps you: no cone, no resonant cavity, just windings shaking a frame. 78 dB at a metre is genuinely impressive for hardware that was never meant to make music — but manage expectations for the far side of the field, and if you want to share the song beyond the flight line, record it properly instead of hoping volume will carry.