Low Frequency Noise Detector

Hunt down that low, droning rumble with this low frequency noise detector. It zooms your microphone in on the sub-100 Hz band — a low-frequency spectrum, automatic 50/60 Hz mains-hum detection, a band breakdown (infrasound, deep sub-bass, low hum), and a reference for “the Hum” phenomenon — so you can see what frequency a complaint actually sits at.

Read this first: your mic almost certainly can’t hear the very low end. Consumer laptop and phone microphones roll off steeply below ~50–100 Hz and essentially cannot capture true infrasound (<20 Hz) at all. Anything shown below ~50 Hz is mic-limited and unreliable. This is an uncalibrated, approximate estimate from a consumer mic — not a calibrated sound-level meter and not legal or complaint evidence. And a rumble you feel but the tool doesn’t show may be structure-borne vibration, not airborne sound. Nothing is recorded or uploaded.

Microphone

Microphone access is requested only when you press Start. Audio is analyzed live and never recorded or sent anywhere.

Idle — press Start to listen for low-frequency noise.

Mains frequency

Europe / Asia / Africa / Australia use 50 Hz mains; North America and parts of South America/Japan use 60 Hz. Auto picks whichever shows a stronger hum peak.

SPL calibration

Not calibrated — levels shown in dBFS (relative). To estimate dB SPL, read the current level on a real meter or calibrated phone app while this is running, type it above, and press Set. The offset is shared across every Noise Analysis tool.

Low-band spectrum (0–120 Hz)

Dominant low frequency: · overall level

Band breakdown

Relative band energy (dBFS) — not affected by SPL calibration.

Infrasound
< 20 Hz
below mic range
Deep sub-bass
20–50 Hz
mic-limited
Low hum
50–100 Hz
most reliable here

Mains-hum check

Listening for a power-line hum and its harmonics…

Share or embed this tool

Free to use on your own website — WordPress, Wix, or any platform. Paste one line and it works instantly, resizing to fit.


How It Works

When you press Start, the tool opens your microphone with automatic gain, noise suppression and echo cancellation switched off (those would distort any noise measurement), runs a large FFT, and throws away everything above ~120 Hz. The result is a zoomed low-frequency spectrum: the horizontal axis covers roughly 0–120 Hz so a faint 47 Hz drone that would be one invisible pixel on a full-range analyzer becomes a clear, labelled peak. It marks the dominant low-frequency component and breaks the energy into three bands — infrasound (<20 Hz), deep sub-bass (20–50 Hz) and low hum (50–100 Hz).

It also runs a dedicated mains-hum check. Electrical equipment, transformers, dimmers and motors radiate a tone at the power-line frequency — 50 Hz in most of the world, 60 Hz in North America — plus harmonics. Within the zoomed band (up to ~120 Hz) that means a sharp peak at 50 and 100 Hz for 50 Hz mains, or 60 and 120 Hz for 60 Hz mains. The tool looks for sharp peaks at those frequencies and tells you whether your rumble is most likely electrical hum versus broadband mechanical noise. Frequency content like this — what frequency, the harmonic spacing, the spectral shape — is calibration-independent, so it is genuinely meaningful even though the absolute loudness number is not.

Common sources of low-frequency noise in residential and commercial buildings include HVAC systems and air-handling units (typically 40–80 Hz duct rumble), industrial machinery and diesel generators nearby, traffic and railway vibration (broadband, usually below 100 Hz), and transformer whine from sub-stations or building plant rooms, which produces a persistent low frequency hum at 50 or 60 Hz and its harmonics. Identifying whether the drone is tonal (sharp spectral spike) or broadband (wide energy hump) and whether it varies with time of day is often more diagnostic than the absolute level — information this spectrum view makes easy to see.

Levels are an uncalibrated estimate

A browser cannot know your microphone’s real-world sensitivity, so the level it computes is dBFS (decibels relative to digital full scale, always ≤ 0), not true environmental dB SPL. If you have a real sound-level meter or a calibrated phone app, read the current level while the tool is running, type it into the calibration box, and press Set — the tool stores an offset (offset = your reading − the current dBFS) and shows an estimated SPL afterwards. That offset is saved in your browser under the shared key fd-noise-cal, so calibrating here also calibrates every other Noise Analysis tool. Until you calibrate, levels stay in honest dBFS. Even calibrated, this remains an approximation, not certified data.

Vibration vs. airborne sound, and the mic’s low-end limit

This is the part to take seriously. Small consumer microphones have a steep low-frequency roll-off: they lose sensitivity rapidly below ~50–100 Hz and are effectively deaf to true infrasound below 20 Hz. So if you can feel a deep rumble in your chest or through the floor but this tool shows little or nothing below 50 Hz, that does not mean the rumble isn’t real — it may be structure-borne vibration travelling through the building, or genuine infrasound your mic simply can’t register. Airborne low-frequency sound that the tool can see (roughly 50–100 Hz) is the reliable zone. For the band below that, treat any reading as a weak hint at best, and consider a contact/vibration sensor or a professional survey.

For context on the wider frequency picture above the low band, the noise frequency analyzer shows the full audible spectrum up to 20 kHz alongside the low-frequency content you see here — useful when you suspect a noise complaint spans multiple ranges.

Frequently Asked Questions

Why does the tool show almost nothing below 50 Hz even when I hear a rumble?
Because consumer laptop and phone microphones roll off steeply below ~50–100 Hz and essentially cannot capture true infrasound (<20 Hz). The low end you feel may be there in the room but below your mic’s usable range, or it may be structure-borne vibration travelling through the building rather than airborne sound. A flat reading below 50 Hz means “my mic can’t hear it,” not “it isn’t there.” The 50–100 Hz low-hum band is the part you can trust most.
Can I use this as proof for a noise complaint?
No. This is an uncalibrated, approximate estimate from a consumer microphone — it is not a certified Type 1/2 sound-level meter and is not valid as legal, complaint or compliance evidence. What it is good for is identifying the frequency and character of a noise (for example, “a 50 Hz electrical hum” or “broadband rumble around 35 Hz”) so you can describe the problem and decide whether to arrange a proper measured survey.
What is “the Hum”?
“The Hum” is the name given to a worldwide phenomenon in which a minority of people report a persistent, low-pitched humming or droning noise that others nearby often cannot hear. Reported pitches commonly fall in roughly the 30–80 Hz range. Investigations have pointed variously to industrial equipment, ventilation and transformers, distant traffic, and in some cases an internal/auditory cause — there is no single confirmed explanation. This tool can help you check whether a measurable low-frequency tone is present, but a silent spectrum doesn’t rule the experience out, especially given the mic’s low-end limits.
How do I know if it’s electrical mains hum?
Electrical hum sits at a sharp, stable peak at the power-line frequency — 50 Hz in most of the world, 60 Hz in North America — usually with harmonics. Within the zoomed band (up to ~120 Hz) the mains-hum check looks for those frequencies: 50 and 100 Hz for 50 Hz mains, or 60 and 120 Hz for 60 Hz mains. A clear narrow spike at 50 or 60 Hz strongly suggests electrical hum from transformers, lighting dimmers, chargers or motors; a broad, hump-shaped rise without a sharp line is more likely mechanical or aerodynamic noise.
Why must automatic gain and noise suppression be off?
Automatic gain control, noise suppression and echo cancellation are designed for voice calls. They actively attenuate steady low-frequency sounds — exactly the rumble and hum you are trying to detect — and constantly change the gain, which makes any level reading meaningless. The tool requests the microphone with all three switched off. If your operating system or browser forces them on anyway, treat the reading as unreliable.
Is my audio recorded or uploaded?
No. The microphone signal is analyzed live in your browser to draw the spectrum and compute the bands, and is never recorded, saved or transmitted. The only thing stored is the optional calibration offset, kept locally in your browser. The microphone is released when you press Stop or close the tab.
Is a phone microphone better than a laptop mic for detecting low-frequency noise?
Neither has a significant advantage for low-frequency noise detection — both roll off steeply below about 50–100 Hz due to the tiny capsule and its sealed housing. Dedicated external USB microphones with a large diaphragm and a flat frequency response down to 20 Hz perform measurably better in the 20–60 Hz range. If you need trustworthy sub-50 Hz readings, a proper measurement microphone (such as an IEC 61672 Type 1/2 calibrated meter) is the only reliable option.
I feel the vibration in the floor or walls but the tool shows nothing — what does that mean?
It almost certainly means the energy is structure-borne, not airborne. Low-frequency vibration from HVAC plant, industrial machinery, or traffic can couple directly into a building’s structure and travel as a seismic wave through floors and walls without radiating much audible airborne sound at all. A microphone measures air pressure — it cannot detect vibration transmitted through solid materials. A contact accelerometer or geophone placed on the floor or wall is the right sensor for structure-borne low-frequency vibration.
What are the most common sources of a 50–100 Hz low-frequency drone or rumble indoors?
The most common causes are: (1) HVAC and air-handling units — fan blades produce tonal hum in the 40–80 Hz range plus duct rumble; (2) electrical transformers inside or near the building, which hum at the mains frequency (50 or 60 Hz) and its harmonics; (3) refrigeration compressors, which produce low broadband rumble plus a tonal motor hum; and (4) external traffic or construction, which generates broadband low-frequency energy that can penetrate building envelopes. The mains-hum check in this tool helps distinguish case 2 from the others.