Click Start to begin real-time spectrum analysis
Click Start above
Browser will request mic access
Allow permission
Click "Allow" when prompted
Play any audio nearby
Music, voice, instruments…
Bar Graph
Frequency amplitude per bin
Waveform
Time-domain signal shape
Spectrogram
Freq × time heatmap
Resolution vs Speed
Higher FFT (4096) = finer frequency detail but slower update. Lower (1024) = faster response but coarser bins.
Free Online Spectrum Analyzer
Visualize the frequency content of any sound in real time. Use your microphone to analyze music, room acoustics, voice, or instruments — directly in your browser.
Real-time FFT Analysis
Uses the Web Audio API's AnalyserNode with configurable FFT sizes (1024–4096) to compute the frequency domain representation of your microphone input with up to 2048 frequency bins.
Three Visualization Modes
Switch between Bar Graph (frequency amplitude by bin), Waveform (time-domain signal shape), and Spectrogram (frequency-vs-time heatmap). Each reveals different aspects of your audio signal.
No Install Required
Works entirely in your browser using the Web Audio API and Canvas. Requires only microphone permission. No plugins, no downloads, no accounts — 100% free and instant.
What is a
Spectrum Analyzer?
Fast Fourier Transform (FFT)
The FFT is a mathematical algorithm that transforms a time-domain audio signal into its frequency components. It reveals which frequencies are present and how loud each one is — the foundation of spectrum analysis.
Frequency Bins
The FFT output is divided into "bins" — discrete frequency ranges. With a 2048-point FFT and 44.1kHz sample rate, each bin covers about 21.5Hz. Higher FFT sizes = more bins = better frequency resolution.
Reading the Spectrum
The X-axis represents frequency (20Hz to 20kHz, the range of human hearing). The Y-axis shows amplitude (loudness). A flat response across most frequencies typically indicates a flat-response microphone or room.
Smoothing
Raw FFT data fluctuates rapidly. Smoothing (set to 0.8 here) averages recent frames to produce a more stable, readable display. Less smoothing = more responsive but jittery, more smoothing = calmer but slower.
How to Read Your Spectrum
Understanding what you see in the analyzer.
High Peaks = Dominant Frequencies
Tall bars or bright spots on the spectrogram indicate the dominant frequencies in your audio. A singer's fundamental note will show a tall peak, with smaller peaks at each harmonic above it.
Flat Response = Balanced Sound
A spectrum that's roughly flat across all frequencies indicates a balanced audio signal. Large bumps in the bass (left side) suggest bass-heavy audio; bumps on the right indicate harsh treble.
Spectrogram = History
The spectrogram scrolls left over time. Horizontal bands (same frequency persisting) indicate continuous tones. Diagonal patterns show pitch changes. Vertical lines indicate brief transients like claps.
Frequently Asked Questions
Q.Does the analyzer need internet to work?
No. Once the page is loaded, the spectrum analyzer runs entirely in your browser using the Web Audio API. No audio data is sent to any server — all processing happens locally on your device.
Q.Why does my spectrum look flat or empty?
Your microphone may need to be closer to the sound source, the input volume may be too low, or the sound source may be too quiet. Try speaking directly into your mic or playing music nearby. Also check your OS microphone volume settings.
Q.What's the difference between bars and spectrogram?
Bar graph shows amplitude vs frequency at the current moment. Spectrogram shows amplitude as color across frequency (vertical) over time (horizontal), scrolling left — like an X-ray of how your audio's frequency content changes over time.
Q.What FFT size should I use?
2048 is a good default — fast enough for real-time response and detailed enough for most use cases. Use 4096 for detailed classical music analysis, 1024 for fast percussion or rhythm visualization.
Q.Can I use this to tune an instrument?
You can use the spectrum analyzer to see dominant frequencies. A guitar A string (440Hz) would show a peak at 440Hz. However, for precise tuning, a dedicated chromatic tuner is more accurate. This is better for broad audio analysis.
Q.Why are very low frequencies (< 20Hz) not shown?
Human hearing begins at approximately 20Hz. Frequencies below this (infrasound) are not displayed as they're outside the audible range. Additionally, most microphones and speakers have limited response below 50-80Hz.