Choose FFT sizes from 1,024 to 8,192 and inspect exact frequency, dBFS, spectral center, and multiple dominant peaks.
FFT · Hz · dBFS · Peak detectionAudio Frequency Analysis Tool
Online Spectrum Analyzer
Upload audio or use your microphone to inspect frequencies, peaks, and a live spectrogram. Everything runs privately on your device.
No server uploadMP3 · WAV · FLAC · M4A
Nothing to analyze yetUpload a file or enable your microphone.
0:00 / 0:00
Upload audio fileClick or drag a file here · MP3, WAV, FLAC, M4A, OGG
Upload audio, try the sample, or use your microphone to begin.
Analysis capabilities
Professional tools for closer signal inspection.
Move from a quick visual check to measurable frequency data without changing applications.
Compare an instantaneous frequency spectrum with a scrolling time-frequency view using linear or logarithmic scaling.
Time × frequency · Linear · LogFreeze a moment, hold peak values, save a reference trace, isolate frequency ranges, and export evidence as PNG or CSV.
Peak hold · Reference · PNG · CSVHow it works
From sound to signal, without the guesswork.
A spectrum shows the frequencies present at one moment. A spectrogram maps those frequencies over time, making tones, harmonics, noise, speech, and hidden images visible.
- 01Choose an input
Upload an audio file or listen through your microphone.
- 02Inspect the signal
Hover for exact readings and tune FFT or image controls.
- 03Export your evidence
Download the current visualization or its frequency data.
Signal generators
Create a signal, then see what it contains.
Generate fixed tones, frequency sweeps, or broadband noise locally in your browser.
Frequently Asked Questions
Is my audio uploaded anywhere?
No. Audio decoding, FFT analysis, rendering, and export happen in your browser. The selected file and microphone signal are not sent to an analysis server.
What is the difference between Spectrum and Spectrogram?
Spectrum shows the frequency content at the current moment. Spectrogram adds time as a horizontal dimension, so you can see how tones, harmonics, speech, noise, and encoded images change throughout a recording.
What is the best FFT size?
4096 is a balanced default. Larger FFT sizes separate nearby frequencies more clearly but respond more slowly to short events. Smaller values improve time resolution for rapidly changing sounds.
Why do the levels use dBFS instead of sound-pressure dB?
dBFS measures level inside a digital audio signal, where 0 dBFS is the maximum before clipping. A browser cannot provide reliable dB SPL measurements without a calibrated microphone and device-specific correction.
Which audio formats can I analyze?
The tool accepts common browser-supported formats including MP3, WAV, FLAC, M4A, and OGG. Actual decoding support can vary slightly by browser and operating system.
Why does the live Spectrogram show a five-second window?
A focused five-second window keeps changes readable during playback and microphone input. The window advances with the audio, while file analysis can also export the complete Spectrogram as a separate PNG.
Can this reveal hidden Spectrogram images?
Yes. Use Reveal hidden image as a starting preset, then adjust brightness, contrast, gamma, frequency scale, and color map. Results depend on how the image was encoded in the source audio.
What do Peak Hold and Save Reference do?
Peak Hold keeps the highest observed level at each frequency. Save Reference stores the current spectrum as an overlay, making it easier to compare another moment, channel, microphone position, or audio file.
What is included in Export?
You can export the current visible graph as PNG, the current spectrum data as CSV, or the complete file Spectrogram as PNG after full-file analysis has finished.