Noise Floor
Quick Definition
The lowest level of background noise present in an audio system or recording, below which no useful signal can be captured. Measured in dBFS. Every recording chain has a noise floor determined by microphones, preamps, interfaces, and the recording format's bit depth.
In-Depth Explanation
Noise floor is the constant background noise present in any audio system or recording, measured in decibels relative to full scale (dBFS). It represents the lowest signal level that can be captured or reproduced before the signal is indistinguishable from the system's inherent noise. Every component in a recording chain (microphones, preamps, cables, audio interfaces, and the digital format itself) contributes to the noise floor, which sets the bottom limit of the usable dynamic range.
How the Noise Floor Works
Every audio device generates a small amount of internal noise. This comes from thermal noise in electronic components, electromagnetic interference from nearby devices, and the quantization noise inherent in digital audio. When you record, this noise is captured alongside the intended signal. The noise floor is the level of that background noise.
The usable dynamic range of a recording is the difference between the noise floor and the clipping point. If your noise floor is at -75 dBFS and your signal clips at 0 dBFS, your dynamic range is 75 dB. Any signal quieter than -75 dBFS is buried in noise.
Sources of Noise Floor
-
Microphone self-noise: Every microphone produces a small amount of internal noise, specified as an equivalent input noise (EIN) rating in dB-A. A quality large-diaphragm condenser microphone has a self-noise of around 5 to 12 dB-A. Budget condensers may reach 20 dB-A or higher. Ribbon and dynamic microphones have very low self-noise but require high-gain preamps, which introduces their own noise.
-
Preamp noise: Microphone preamps add noise when boosting weak microphone signals. Quality preamps from companies like Grace Design, Millennia, and Audient add less than -130 dBV of noise. Budget interface preamps may add -120 dBV or more. The preamp noise becomes part of your recording's noise floor.
-
Audio interface noise: The analog-to-digital converter in your interface has its own noise floor. Professional interfaces (RME, Universal Audio, Antelope) typically achieve noise floors below -110 dBFS. Budget interfaces may sit at -100 dBFS.
-
Quantization noise: In digital audio, the bit depth sets a theoretical noise floor. 16-bit audio has a noise floor of -96 dBFS. 24-bit audio has a noise floor of -144 dBFS. This is why 24-bit recording gives you a lower noise floor than 16-bit, regardless of your hardware.
-
Acoustic noise: The room itself contributes to the noise floor. HVAC systems, computer fans, traffic, and electrical hum all raise the effective noise floor of a recording. A typical untreated bedroom studio has an ambient noise floor of 35 to 45 dB SPL. A treated vocal booth achieves 20 to 25 dB SPL.
How to Measure Your Noise Floor
In your DAW, arm a track with your microphone connected, set the preamp gain to the level you use for recording, and record 10 seconds of silence. Play back the recording and read the average level on a meter. That number is your effective noise floor for that recording chain.
If the noise floor reads above -60 dBFS, something is wrong. Check for ground loops, gain staging issues, or a noisy preamp. A well-configured 24-bit recording chain should produce a noise floor between -75 and -90 dBFS.
Real-World Example
You record a vocal at 24-bit with a Rode NT1 microphone into a Universal Audio Apollo interface. The preamp gain is set so the vocal peaks at -12 dBFS. You record 10 seconds of silence and check the level: the noise floor sits at -82 dBFS. This means your dynamic range is 82 dB (from -82 dBFS to 0 dBFS). The vocal, peaking at -12 dBFS, sits 70 dB above the noise floor. The recording is clean.
Now you apply heavy compression to the vocal: 10 dB of gain reduction with 10 dB of makeup gain. The compression brings the quietest vocal passages up by 10 dB, but it also brings the noise floor up by 10 dB. The noise floor is now at -72 dBFS. During quiet passages between vocal phrases, the listener can hear a faint hiss. This is why gain staging and noise floor management matter before you reach the compression stage.
You switch to a budget microphone with a self-noise of 22 dB-A into a budget interface preamp. The noise floor jumps to -65 dBFS. After the same compression, the noise floor is at -55 dBFS. The hiss is clearly audible between phrases. The recording sounds amateur.
Why It Matters for Independent Artists
Your noise floor determines the maximum dynamic range of your recording and how much processing you can apply before noise becomes audible. Three practices keep your noise floor low:
-
Record at 24-bit. This gives you a theoretical noise floor of -144 dBFS, which is far below the noise of any microphone or preamp. The actual noise floor will be determined by your hardware, not the recording format. Never record at 16-bit unless you have no other option.
-
Set gain correctly. Aim for recording levels averaging -18 to -12 dBFS with peaks no higher than -6 dBFS. This gives plenty of headroom while keeping the signal well above the noise floor. Do not try to record as hot as possible. At 24-bit, there is no benefit to recording near 0 dBFS, and you risk clipping.
-
Treat your room. Acoustic noise is often the highest contributor to your effective noise floor. Recording in a room with an HVAC vent running adds a constant 40 dB SPL of low-frequency rumble. Turn off fans and HVAC during recording. Use acoustic treatment or a reflection filter to reduce room noise. Even a portable vocal booth like the Aston Halo or sE Electronics Reflexion Filter makes a measurable difference.
-
Clean up before processing. If your recording has a noise floor at -70 dBFS, apply a noise gate or noise reduction plugin before heavy compression. Tools like iZotope RX 11, Waves NS1, and Cedar DNS One can reduce noise floor by 10 to 20 dB without damaging the signal. This prevents compression from amplifying the noise.
In 2026, AI-powered noise reduction tools have become standard in post-production. iZotope RX 11 includes a Dialogue Isolate module that uses machine learning to separate voice from background noise with remarkable accuracy. Adobe Podcast AI (formerly Project Shasta) offers free browser-based noise reduction that rivals professional plugins. These tools are useful for rescue work but should not replace good recording practices. The best noise floor is the one you never create.
For hardware recommendations that minimize noise floor, read our guide on the best audio interfaces for home studio recording in 2026 and our home studio budget guide. For understanding how noise floor interacts with loudness standards, see our mastering for streaming platforms article.
Related Terms
- Bit Depth - Sets the theoretical digital noise floor of a recording
- Dither - Adds noise at the noise floor level to prevent quantization distortion
- Compression - Raises the effective noise floor by amplifying quiet passages
- DAW (Digital Audio Workstation) - Where noise floor is measured and noise reduction is applied
- Mastering - The stage where noise floor is finalized before distribution