PRIVATE, BROWSER-BASED AUDIO ANALYSIS
Online Spectrum Analyzer
Inspect bass, mids, treble, frequency peaks, and how sound changes over time directly in your browser.
Audio Spectrum Lab
ReadyChoose a microphone or local audio file, then watch the sound by frequency or over time. Use peak hold, exact frequency inspection, and display settings when you need more detail.
Live FFT Spectrum
Digital spectrum level · not calibrated SPL
Scrolling Spectrogram
Watch when frequencies appear and fade. Brighter color means more signal.
Waveform
How the signal level rises and falls over a short time.
Start a microphone or load a local audio file.
Top 5 Meaningful Peaks
The strongest distinct frequency spikes in the current signal.
- 01—Waiting for signal—
- 02—Waiting for signal—
- 03—Waiting for signal—
- 04—Waiting for signal—
- 05—Waiting for signal—
Mean Power by Frequency Band
Mean FFT-bin power in each band, expressed as a relative digital level; not total band energy.
Advanced Analysis SettingsFFT detail, smoothing, scale and visible range
Reading note. The graph shows the digital signal received by your browser. Microphone response, gain, noise suppression, echo cancellation, device processing, and the room can all change what appears. It is not a certified SPL or sound-safety reading.
This frequency-analysis and comparison tool shows signal content, not whether a sound is safe. It does not provide certified SPL measurements or medical or hearing diagnoses. Understand accuracy and limitations →
How to Use the Spectrum Analyzer
Start with the defaults. The controls become useful when you have a specific frequency question.
- 01
Choose an input
Start the microphone analyzer and allow browser permission, or select a local audio file.
- 02
Watch the live spectrum
Frequency runs from low bass to high treble; stronger parts of the signal rise higher on the graph.
- 03
Freeze or hold peaks
Use Freeze to inspect one moment, and Peak Hold to keep the strongest level seen at each frequency.
- 04
Adjust only when needed
FFT size changes frequency detail and response speed. Smoothing makes the display steadier.
- 05
Use readouts as clues
Compare peaks, bands, notes and the spectrogram with what you are actually hearing.
What Does a Spectrum Analyzer Show?
A spectrum analyzer shows which low and high sounds are present right now. On the graph, frequency runs from bass on the left to treble on the right. Amplitude, or level, is how strong the signal is: a taller line means more signal at that frequency. A peak is a frequency that stands out from its neighbors.
Bass is the low rumble or beat, midrange carries much of speech and music, and treble is the higher detail, such as hiss or sharp consonants. This online audio frequency analyzer helps you compare those parts of a signal. Time tells you when a sound appears or changes. Its level is a digital signal level, not a room loudness or safety reading.
- Waveform
- Signal level rising and falling over time.
- Spectrum
- Frequency across the graph; level shows how strong each part is now.
- Spectrogram
- Frequency and level repeated over time, so you can see changes.
Spectrum vs Spectrogram: Now vs Over Time
Spectrum: which frequencies are strong right now. Spectrogram: when those frequencies appear, continue, and fade. Both show the same audio in different ways.
How FFT works: The analyzer takes a tiny slice of audio and sorts it into low and high frequencies, much like separating notes in a chord. It repeats this many times each second to keep the picture moving. This FFT analyzer uses a calculation called a Fast Fourier Transform.
FFT Size: Detail vs Speed
FFT size sets the length of each audio window. A longer window separates close, steady tones more clearly, but reacts more slowly to sudden changes. Smoothing is separate: it blends recent readings to steady the display.
Read the Two Views
The spectrum is a snapshot. The spectrogram is a scrolling history of snapshots.
Left to right is low to high frequency. Taller peaks mean more signal at those frequencies.
Frequency runs up the side, time moves across, and brighter color means more signal. Watch a tone appear as a bright band and fade when it stops.
How to Read the Frequency Axis
Low frequencies are bass, the middle covers much of speech and music, and high frequencies are treble. The default log scale gives bass more room so you can see it clearly. These bands are useful guides, not fixed boundaries.
Sub-bass
Deep rumble, subwoofer content and some environmental vibration.
Bass
Bass instruments, low voice fundamentals and many room resonances.
Low mids
Warmth, body and lower instrument fundamentals.
Mids
A large part of speech and musical detail.
Upper mids
Definition, articulation and strong ear sensitivity.
Presence
Brightness, consonants, clicks and sibilant detail.
High treble / air
High harmonics and hiss; device and hearing limits matter strongly here.
Dominant Frequency, Harmonics and Notes
A stable tone often produces one obvious dominant frequency, but real sounds usually contain several peaks. A musical note can have a fundamental plus harmonics at higher frequencies. The strongest peak is not always the fundamental.
When a clear tonal peak is present, the analyzer shows the nearest equal-tempered note and cents offset. For broadband noise, speech, chords or complex audio, one note label may not be meaningful.
Finding Hum, Hiss and Narrow Peaks
The display can reveal repeatable patterns. Treat them as clues to compare under controlled conditions, not automatic diagnoses.
A persistent narrow peak near mains frequency—often with harmonics—can be worth investigating, but it does not prove the cause.
Hum
Look for a persistent narrow peak around 50 Hz or 60 Hz, often with harmonics.
Hiss
Broad high-frequency energy appears as a wide elevated region rather than one clean peak.
Resonance
A frequency that rises repeatedly with the same source or position can indicate resonance somewhere in the chain.
Whistles and tones
Narrow tonal sounds are easier to track with larger FFT sizes and peak hold.
Spectral Centroid: a Simple Brightness Reference
This analyzer calculates spectral centroid as the frequency average weighted by FFT-bin power. Higher values generally mean more power is concentrated at higher frequencies; lower values mean more is concentrated lower down.
Use it as a comparative descriptor, not a quality score. Two sounds can have similar centroids and still sound completely different.
Analyze a Local Audio File
File analysis lets you inspect a recording without adding microphone and room effects again. Load a supported local file, play or seek through it, and watch every view update through the same analyzer pipeline.
The file remains local to your browser. Codec support can vary by file format and device.
Microphone Analysis & Privacy
Microphone mode requires browser permission and access starts only after you press Start. The analyzer processes frames in memory; it does not need to record, upload or play microphone audio through your speakers.
When supported, the input selector appears after permission is granted. Read the Privacy Policy.
What Can You Use an Audio Spectrum Analyzer For?
Use the lab to compare, inspect and learn—not to replace calibrated acoustic hardware.
Voice & microphone checks
See where speech energy sits, compare microphones and spot obvious hum, hiss or processing changes.
Music & instruments
Inspect fundamentals, harmonics, changing notes and the spectral balance of a recording.
Speaker / room troubleshooting
Look for repeatable peaks or dips while changing position, source or playback conditions.
Noise investigation
Freeze narrow tones, broad hiss and repeating frequency patterns so you can compare them over time.
Audio education
See the difference between waveform, spectrum, harmonics, FFT size and spectrogram in real time.
Reference-signal inspection
Use the Tone Generator for a controlled input, then observe what your microphone or recording chain captures.
What This Analyzer Can and Cannot Tell You
The graph represents the digital signal reaching the browser. Microphone response, input gain, browser or device processing, and the room all influence it. A real-time spectrum analyzer can reveal patterns, but it cannot tell whether a sound is safe.
Useful for
- Relative frequency content of the browser input
- Dominant tones and harmonics
- Peak-hold comparisons
- Broad frequency-band comparisons
- Spectrogram changes over time
- Microphone vs local-file comparisons
Not proof of
- Calibrated acoustic SPL by frequency
- Certified microphone frequency response
- The exact cause of every hum, hiss or resonance
- RF / radio-frequency spectrum analysis
- Laboratory distortion or transfer-function measurement
- Hearing or medical diagnosis
Why this is not calibrated SPL. A laptop, headset or USB microphone can have its own frequency response, gain, filtering, automatic gain control, echo cancellation and noise suppression. Use this sound spectrum analyzer for relative frequency inspection and comparison, not certified acoustic measurement, medical or hearing diagnosis, or a safety decision. Read our Accuracy & Limitations, Validation, and Disclaimer.
Spectrum Analyzer Questions
Clear answers about FFT settings, digital levels, files, privacy and what the analyzer can reliably show. Read about accuracy and limitations, or see the methodology.
What is an online spectrum analyzer?
It shows which low and high frequencies are in a sound and how strong they are. This browser tool analyzes audio from a microphone or local file; it does not analyze radio frequencies.
What is the difference between a spectrum analyzer and a frequency analyzer?
For this browser audio tool, the terms overlap heavily. Both describe inspecting how much signal energy appears at different frequencies. “Spectrum analyzer” is the more common engineering term; “frequency analyzer” is a natural description of the same task here.
What is a spectrum analyzer used for?
Common uses include finding strong tones, hum and hiss, inspecting harmonics, comparing microphones or recordings, watching speech or music frequency balance, and learning how audio changes across the spectrum.
Is this a radio frequency spectrum analyzer?
No. This page analyzes audio-band signals handled by the browser. It does not analyze RF transmissions, wireless spectrum, antennas or electromagnetic signals.
What does FFT mean?
The analyzer takes a short slice of sound and works out which pitches are in it. FFT stands for Fast Fourier Transform: the calculation that sorts audio samples into frequency bins for the spectrum.
What FFT size should I use?
Use 4096 as a balanced default. Smaller sizes respond faster but have wider frequency bins. Larger sizes separate nearby stable frequencies more clearly but react more slowly.
Why does a larger FFT move more slowly?
A larger FFT uses more audio samples in each analysis window. That gives finer frequency spacing, but the window covers more time, so the display cannot respond as quickly to sudden changes.
What does smoothing do?
Smoothing averages the current spectrum with recent frames. Lower smoothing reacts quickly and looks more jumpy; higher smoothing is steadier but can hide fast changes.
Why is the frequency axis logarithmic?
Audio covers a very wide range from tens of hertz to tens of kilohertz. A logarithmic scale gives useful visual space to low frequencies while still showing the entire audible range.
What is dBFS on the spectrum?
dBFS is a digital level reference relative to the maximum representable digital signal. It is not the same as calibrated dB SPL in the room.
Can this measure real sound pressure level at each frequency?
Not by itself. The browser receives a microphone’s digital signal, and that signal is affected by microphone sensitivity, frequency response, gain, processing and the room. Calibrated acoustic measurement requires suitable measurement hardware and calibration.
What is peak hold?
Peak hold keeps the highest level seen at each frequency until you reset the peaks or change the FFT size. It helps reveal brief tones or harmonics that would otherwise disappear as soon as the live spectrum changes.
What is a spectrogram?
A spectrogram shows when low and high sounds appear and disappear. Frequency is shown on one axis, time scrolls along the other, and brighter color means more signal at that frequency and time.
What is the difference between a waveform and a spectrum?
A waveform shows how the signal changes over time. A spectrum shows which frequencies are present in that signal. They are two different views of the same underlying audio.
What is the dominant frequency?
It is the strongest useful spectral peak in the current analysis frame. With complex sounds, the strongest peak is not always the fundamental frequency.
Why does the note detection sometimes look wrong?
Note detection only makes sense when a stable tonal peak is present. Speech, noise, chords and complex instruments may contain several competing peaks, so one note label can be misleading.
What is spectral centroid?
Here it is the frequency average weighted by FFT-bin power. It describes whether spectral power is concentrated lower or higher in frequency, but it is not a quality score.
Can I use this to find 50 Hz or 60 Hz hum?
Yes, a persistent narrow peak near 50 or 60 Hz and its harmonics can be a useful clue. The analyzer cannot automatically prove what is causing that peak.
Can I analyze an audio file without uploading it?
Yes. The selected file is decoded and analyzed on your device without being sent to Decibel Meter Lab. Browser codec support can vary by file format and device. Read the Privacy Policy.
Does microphone audio get recorded?
No. The live analyzer processes microphone audio in memory and has no recording feature. It keeps derived spectrum history while the page is open, but does not save or upload the raw microphone audio. Read the Privacy Policy.
Why is my spectrum different on another device?
Microphones, input gain, noise suppression, automatic gain control, sample rate, room acoustics and operating-system processing can all change the captured signal.
Can I use this as an audio waveform analyzer?
Yes. The waveform view shows the time-domain signal, while the spectrum and spectrogram show its frequency content. The views are complementary.
Can the analyzer diagnose a damaged speaker or microphone?
It can reveal clues such as missing bands, unexpected harmonics, noise or repeatable peaks, but it cannot isolate every cause automatically. Use controlled reference signals and compare devices or positions before drawing conclusions.
Does the analyzer work on phones and tablets?
It can work in modern mobile browsers that support the required Web Audio and media APIs. Screen size, microphone processing and browser behavior vary by device.