Noise Floor Calculator

Knowledge Base

How to Use the Noise Floor Calculator

The Noise Floor Calculator helps you measure and understand ambient noise levels in your recording space. Input microphone sensitivity and ambient noise dB level to see your signal-to-noise ratio. Critical for assessing whether your space is suitable for professional recording.

Understanding Noise Floor

Every recording space has ambient noise: air conditioning, traffic, computer fans, electrical hum. The noise floor is the lowest level of sound your mic picks up. A low noise floor (quiet space) is essential for professional recordings.

Noise Floor References (dB SPL)

• 0 dB SPL: Threshold of hearing (theoretical).
• 30 dB SPL: Quiet library or bedroom at night. Excellent for recording.
• 40-50 dB SPL: Normal home interior, office. Acceptable for voice recording.
• 60 dB SPL: Normal conversation, typical office noise. Noticeable in quiet passages.
• 70+ dB SPL: Loud traffic, construction. Not suitable for professional recording without treatment.

Improving Your Noise Floor

Use acoustic treatment: blankets on walls, bass traps in corners, soft furnishings. Close windows and doors. Record during quiet hours. Use high-pass filters in post to remove rumble below 80 Hz (traffic, HVAC). Dynamic range compression can help mask low-level noise if subtle.

FAQ

Q: How much noise is too much?
A: Anything below -40 dB (relative to your voice peak) is acceptable. Below -50 dB is professional. Above -30 dB is problematic.

Q: Can I fix noisy recordings in post?
A: Partially. Noise reduction removes some noise but also removes voice detail. Prevention (quiet space) is always better than cure.

Where the noise in a recording comes from

Three sources add together. The microphone contributes self-noise, the electrical hiss its own circuitry generates. The preamplifier contributes its own, usually quoted as equivalent input noise. The room contributes whatever is happening in it. Estimating the first two tells you the best a chain can do, and the third is almost always the one that decides the actual result.

Self-noise figures are quoted in dB SPL A-weighted, and the number describes how loud a sound would have to be to match the microphone's own hiss. A microphone rated at 7 dB is very quiet, one at 25 dB is ordinary for a dynamic model, and the difference matters most when recording quiet sources at a distance.

Distance dominates everything. Halving the distance between mouth and microphone raises the signal by about six decibels while leaving the room and the electronics where they were, so it improves the ratio by the same amount. No microphone upgrade delivers that improvement as reliably as moving closer does, and closer costs nothing.

Gain structure matters more than gain quantity. Setting the preamp low and raising the file afterwards amplifies whatever noise the preamp contributed, because the signal was small relative to that noise when it passed through. Setting a healthy level at the preamp and leaving headroom for peaks gives a cleaner result than fixing level later.

Match the microphone to the source rather than to a specification sheet. A very quiet condenser in an untreated room records the room beautifully, which is not what anybody wanted. A dynamic microphone with higher self-noise, used close, frequently produces a cleaner-sounding result in a real space, because it hears less of what surrounds it.

Vocabulary

Self-noise. The hiss a microphone generates on its own, quoted in dB SPL A-weighted. Effectively the quietest sound the microphone can meaningfully capture, since anything below it is buried.

Equivalent input noise. A preamplifier's own noise expressed as the input signal that would produce the same output, so it can be compared against microphone self-noise. Written as EIN and quoted as a negative decibel figure.

Signal to noise ratio. The gap in decibels between your wanted sound and the noise underneath it. The single figure describing how clean a recording is, and the one improved most cheaply by moving the microphone closer.

A-weighting. A filter applied to noise measurements approximating how human hearing responds to level, discounting very low and very high frequencies. Used because unweighted figures overstate noise nobody actually hears.