The noise floor is the baseline level of unwanted sound or signal present in any audio system, and reducing it comes down to managing every link in your signal chain. From microphone choice to cable type to digital settings, each stage either adds noise or keeps it in check. The good news: most noise floor problems have straightforward fixes that don’t require expensive gear upgrades.
What Creates the Noise Floor
Every electronic component generates some amount of noise simply by existing. Resistors produce thermal noise from the random movement of electrons, and this happens at any temperature above absolute zero. Semiconductor junctions in your preamps, interfaces, and processors add their own noise through the flow of individual electrical charges. On top of that, material imperfections in components create low-frequency “flicker” noise. None of this can be completely eliminated, but it can be minimized and managed so it stays well below your actual signal.
The practical noise floor you hear is a combination of this inherent electronic noise, electromagnetic interference picked up from nearby devices and power lines, and any noise introduced by your cables, connectors, and grounding scheme. Tackling each of these sources separately is the most effective approach.
Set Your Gain Staging Correctly
Gain staging is the single most impactful technique for keeping your noise floor low. The idea is simple: at every point in your signal chain, the audio level should be strong enough to sit well above the noise floor but not so strong that it clips or distorts. When you under-gain a signal at the input stage and then boost it later, you’re amplifying all the noise that crept in during the quiet stages. When you over-gain, you get distortion that’s impossible to fix.
Start at your microphone preamp. Its job is to bring the mic’s weak output up to line level, which is the standard operating level for recording gear. A good target when recording digitally is around -12 dBFS on your DAW’s meters. This leaves enough headroom to handle peaks while keeping the signal far above the preamp’s noise floor.
Carry this principle through every stage. If a compressor reduces your signal level, use its output or makeup gain to bring it back up before it hits the next processor. If a plugin drops your level, compensate at that plugin’s output rather than cranking gain later in the chain. Every time a signal sits too low relative to the noise floor of the device processing it, you lose ground.
Choose the Right Bit Depth
In digital recording, bit depth directly determines how far your signal can sit above the theoretical noise floor. A 16-bit recording has a noise floor at -96.3 dBFS. A 24-bit recording pushes that down to -144.5 dBFS, giving you roughly 48 dB of extra dynamic range. In practice, no analog front end (microphone, preamp, converter) is quiet enough to take full advantage of 24-bit’s theoretical range, but recording at 24-bit gives you substantially more room to work with quiet signals without hearing quantization noise.
If your DAW supports 32-bit floating point processing, use it for mixing. The dynamic range at 32-bit float is so vast (over 1,500 dB) that internal clipping within your mix bus becomes essentially impossible, and rounding errors from plugin processing stay far below audibility. Record at 24-bit, mix at 32-bit float, and you’ve eliminated digital noise as a practical concern.
Use Balanced Cables and Connections
Balanced cables are one of the most reliable ways to reject noise picked up along cable runs. An unbalanced cable carries one copy of the signal plus a ground wire that doubles as a shield. A balanced cable carries two copies of the signal with reversed polarity, plus a separate ground shield. Any electromagnetic interference that hits the cable affects both signal copies equally. At the receiving end, the gear flips one copy back to its original polarity, which means the signal adds up while the noise cancels out.
This is why XLR and TRS connections are standard in professional audio. For cable runs longer than about 15 to 20 feet, balanced connections make a noticeable difference. Even on shorter runs, balanced cables provide better noise rejection than their unbalanced equivalents, particularly in environments with lots of nearby electronics or power cables. If your gear has balanced outputs and inputs, use them.
Eliminate Ground Loops
A persistent hum at 50 Hz or 60 Hz (depending on your country’s power frequency) almost always points to a ground loop. This happens when two or more pieces of gear are connected to a common ground through different electrical paths. Small voltage differences between those paths create current flow that shows up as an audible hum in your signal.
To diagnose a ground loop, start by isolating the source. If the hum doesn’t change when you adjust volume on your preamp or receiver, the problem is downstream of that point. Try selecting different inputs: if the hum disappears on one input, the source device on the noisy input is involved. Cable TV and cable internet connections are frequent culprits because they introduce a separate ground path into your system.
The two ways to fix a ground loop are to remove one of the redundant ground paths or to break the loop with an isolation transformer. An isolation transformer passes the audio or video signal through but interrupts the ground connection between devices, so the noise voltage appears across the transformer windings instead of at your circuit’s input. For most consumer and project studio setups, a small isolation transformer on the offending connection is the cheapest and most practical fix. Plugging all your audio gear into the same power strip or outlet can also help by reducing ground potential differences between devices.
Reduce Electromagnetic Interference
Electromagnetic interference from nearby electronics, power supplies, motors, LED dimmers, and wireless devices raises the noise floor in ways that gain staging alone can’t fix. Physical distance is the first line of defense: keep audio cables away from power cables, and don’t coil them together. Cross power and audio cables at right angles when they must share space, which minimizes the length of cable exposed to the interference field.
For persistent RF or EMI problems, shielding becomes important. Metal enclosures naturally block interference, which is one reason professional audio gear uses steel or aluminum chassis. Plastic enclosures can be treated with conductive coatings (silver, copper, or nickel-based sprays) or lined with conductive films and metalized fabrics. Gaps and seams in enclosures are weak points, and conductive gaskets made from silicone or foam embedded with metal particles can seal them.
Ferrite chokes, the cylindrical lumps you sometimes see on cables, are another simple tool. They suppress high-frequency noise traveling along a cable’s outer shield and cost very little.
Invest Where It Matters Most
Your noise floor is largely set by the first gain stage in the chain, which for most recording setups means the microphone preamp. A preamp’s equivalent input noise (EIN) tells you how much noise it adds to the signal. The theoretical best possible EIN is -133 dBV, which represents only the thermal noise from a 150-ohm resistor (the standard test impedance). Quality standalone preamps and modern audio interfaces get close to this limit.
Top-tier interfaces now achieve signal-to-noise ratios above 113 dB on mic inputs and above 116 dB on line inputs. At these levels, the converter and preamp noise is low enough that your room noise, cable quality, and gain staging decisions matter far more than the interface itself. If you’re using a reasonably modern interface from a reputable manufacturer, upgrading to a more expensive one will yield diminishing returns. Your money is better spent on acoustic treatment, balanced cabling, and solving any ground loop or interference problems first.
Software Noise Reduction as a Last Step
When you’ve done everything you can at the hardware and signal chain level, software noise reduction can clean up what remains. Modern noise reduction plugins work by sampling a section of “silence” in your recording to build a profile of the noise, then subtracting that profile from the full signal. This works well for steady-state noise like hiss, hum, or fan noise.
The key is to use noise reduction sparingly. Aggressive settings introduce artifacts, often described as a “watery” or “metallic” quality, that can sound worse than the original noise. Apply just enough reduction to push the noise below the listener’s attention threshold rather than trying to eliminate it entirely. And treat software noise reduction as a safety net, not a substitute for getting the signal chain right in the first place. A clean recording with good gain staging will always sound better than a noisy recording that’s been processed after the fact.