Signal-to-Noise Ratio

The SNR is the green gap: how far your signal towers above the constant noise floor, before it clips at 0 dBFS.

What it is

How far your wanted sound sits above the constant background hiss, measured in decibels.

Key facts

• SNR = signal level (dB) minus noise floor (dB). It is the gap between your audio and the constant background hiss.
• Decibel formula: SNR(dB) = 10 x log10(P_signal / P_noise), where P_signal = power of your audio, P_noise = power of the background noise.
• Bigger number = cleaner. 60 dB SNR is good kit, 80-90 dB is studio-grade, 110-120 dB is top-end converters.
• +6 dB = double the voltage (level). +3 dB = double the POWER (watts). +10 dB sounds about twice as LOUD to ears.
• Threshold of hearing = 0 dB SPL. Whisper ~30 dB SPL, conversation ~60 dB SPL, gig PA ~100-110 dB SPL.
• Digital SNR limit = about 6.02 x bits + 1.76 dB. So 16-bit ~= 96 dB, 24-bit ~= 144 dB of dynamic range.
• Noise floor = the lowest constant signal (hiss + hum) your rig makes with NO input. Every gain stage adds its own.
• Headroom = the gap from your signal up to clipping at 0 dBFS. Aim peaks around -18 to -12 dBFS to keep it.
• Australian mains hum sits at 50 Hz plus harmonics (100, 150 Hz). The US is 60 Hz.
• Gain-staging rule: set the FIRST stage (mic preamp/trim) hot and clean, because all later noise rides on top of it.

How it works

1. Plug in the source, turn the channel up to working level: that loud bit is your signal.
2. Mute or pause the source and watch the meter: that residual hiss/hum is your noise floor.
3. Subtract: signal level minus noise floor = your SNR in dB.
4. Push gain at the EARLIEST stage (mic preamp) so the signal towers over noise before it hits the desk.
5. Never crank later faders/amps to fix a weak signal: that just amplifies the hiss too.

Real examples

• Quiet lapel mic on low gain, then amp cranked = audible hiss behind the speaker.
• A long unbalanced guitar lead near a dimmer pack = 50 Hz hum riding under the music.
• Cheap radio mic at the edge of range = static mush, a collapsing SNR.
• Recording at -30 dBFS then normalising later = you drag the noise floor up with the signal.
• Laptop headphone-out into a PA instead of a balanced DI = noticeably worse SNR.

How it helps in live sound

• Gain-stage from the front: set preamp/trim so peaks hit -18 to -12 dBFS, faders near unity (0 dB).
• Use balanced XLR/TRS and DI boxes for any run over ~3 m to kill hum and noise pickup.
• Lift a hum problem with the DI's ground-lift switch, not by turning gain down.
• Keep audio leads away from mains and lighting power; cross them at 90 degrees if they must meet.
• High-pass filter at 80-100 Hz on vocals to dump low-end rumble and 50 Hz hum.
• Get the mic closer: doubling distance to source drops wanted signal ~6 dB while room noise stays put.