Time-frequency audio analysis

Online Audio Spectrogram

See where tones, harmonics, speech, hum, noise, and short events occur—not just which frequencies exist. Analyze an audio file or follow live microphone input as it changes.

Open the spectrogram Learn how to read it
  • 20 Hz–20 kHzFrequency view
  • 5-second liveRolling window
  • PNG + CSVCurrent-view export
LIVE SPECTROGRAM
Current Online Spectrum Analyzer interface showing a real spectrogram, live measurements, and playback controls
Current analyzer interface · built-in 48 kHz diagnostic sample · five-second rolling view
Read the image

Three dimensions.
One useful picture.

A spectrum captures one moment. A spectrogram keeps those moments in order, making change visible.

01

Time runs left to right

Locate when a note begins, a click occurs, speech changes, or background noise enters the recording.

WHEN
02

Frequency runs low to high

Low-frequency rumble sits near the bottom. Fundamentals and harmonics form horizontal bands above it.

WHERE
03

Color represents signal level

Brighter areas indicate stronger energy under the current brightness, contrast, gamma, and color-map settings.

LEVEL
A practical workflow

From source to evidence.

Start with the defaults. Change a setting only when the question you are investigating requires it.

  1. 01

    Choose the source

    Upload MP3, WAV, FLAC, M4A, or OGG, try the sample, or start the microphone.

  2. 02

    Open Spectrogram

    Press Play for a file. The image advances in sync with the sound you hear.

  3. 03

    Begin balanced

    FFT 4096 and Log scale are a useful starting point for speech, music, hum, and noise.

  4. 04

    Inspect patterns

    Look for bands, harmonic spacing, broadband clouds, and short vertical transients.

  5. 05

    Confirm by listening

    Use the image to locate a passage, then listen around it before deciding what caused it.

  6. 06

    Export the view

    Download the current PNG or CSV. Pro adds the complete timeline and selected 4K export.

Choose the right view

Spectrum, spectrogram, or waveform?

01

Spectrum

Frequency now

Best for reading exact frequency and level at the current moment. Earlier moments disappear unless held or captured.

02

Spectrogram

Frequency over time

Best for changing tones, speech, intermittent noise, harmonics, and events that need a timeline.

03

Waveform

Amplitude over time

Best for edits, timing, waveform shape, and obvious clipping. It does not separate frequencies.

Go deeper: How to read a spectrogram·Spectrum vs Spectrogram·Reveal hidden spectrogram images

Resolution and limits

Sharper frequency detail always costs some time detail.

FFT size controls the trade-off. A larger transform separates nearby steady frequencies; a smaller transform follows brief changes more closely.

Color is not calibrated sound pressure. Brightness and contrast reveal structure, while the plotted digital level remains dBFS unless the complete signal chain is calibrated.

A pattern is a lead, not a cause. A line near 50 or 60 Hz may suggest mains hum, but fans, engines, music, handling, and room noise can occupy similar frequencies.

Read the complete measurement methodology →
Common questions

Before you read the colors.

Short answers to the decisions that most often change the result.

What does a spectrogram show?

A spectrogram shows time on the horizontal axis, frequency on the vertical axis, and signal level through color or brightness. It reveals when tones, harmonics, noise, speech, and short events occur.

Can I make a spectrogram from an audio file?

Yes. Load a supported MP3, WAV, FLAC, M4A, or OGG file, choose Spectrogram, and press Play. The analyzer also prepares complete-file spectrogram data for supported exports and Pro timeline tools.

Can I use a microphone for a live spectrogram?

Yes. Choose microphone input and allow access when prompted. The live view keeps the latest five seconds so changes remain readable while you listen and measure.

Which FFT size should I use for a spectrogram?

FFT 4096 is a balanced starting point. A larger FFT separates nearby frequencies more clearly but spans a longer slice of time; a smaller FFT follows short changes more closely with less frequency detail.

Why are some bands brighter than others?

Brighter colors represent stronger energy under the selected display settings. Bright horizontal bands often indicate sustained tones or harmonics, while short vertical marks often indicate clicks, attacks, or other transients.

Use your own audio

See the change, then hear it.

Open the analyzer, switch to Spectrogram, and follow the signal across time.

Start analyzing