What is Spectrum Analyzer?

Spectrum Analyzer shows the frequency content of audio in real time using FFT analysis. Upload an audio file or use your microphone to see frequency distribution, peak levels, and spectral patterns as they happen.

FFT size goes from 256 up to 32768 bins, giving you the choice between fast time response and fine frequency resolution. Pick the bars view to read individual notes, the line view to spot harmonic peaks, or the waterfall to watch how the spectrum evolves over seconds. Supports MP3, WAV, OGG, FLAC and AAC up to 50 MB.

How to use

1. Upload an audio file or select microphone as the input source.
2. Adjust FFT size, frequency range, and visualization style (bars, line, waterfall).
3. Play the audio and observe the real-time frequency spectrum display.

When to use

• Tuning a room or pair of monitors and looking for resonant peaks to cut with EQ.
• Studying how an instrument's harmonics build up after the initial attack.
• Checking that a podcast voice stays inside the 80 Hz to 12 kHz range your codec keeps.

Result

A sound engineer uploads a vocal recording and identifies a resonance peak at 3.2 kHz that needs EQ correction, visible as a persistent spike in the spectrum display.

FAQ

Why does a higher FFT size make the readout sharper but slower?

Each FFT bin covers sample-rate divided by FFT size. At 44.1 kHz and 2048 bins, each bar is 21.5 Hz wide. Doubling to 4096 halves that to 10.7 Hz, but the analyzer has to gather twice as many samples, so the display lags more.

What's the difference between the bars, line and waterfall views?

Bars give a classic per-band readout, like a hardware spectrum analyzer. Line draws a smooth curve through the bin tops, which makes harmonic peaks easier to follow. Waterfall scrolls the spectrum downward over time so you can see how a sound evolves.

Can I record from my microphone instead of a file?

Yes, pick Microphone as the input source and the browser will ask for mic permission. The audio is read by the Web Audio API directly and never leaves your device, so you can analyze a live instrument or your room's noise floor in real time.

Why does my speech show energy mostly under 4 kHz?

Voice fundamentals sit between roughly 85 Hz (male) and 255 Hz (female), with formants and consonants adding content up to 8 kHz. Most of the perceived loudness lands below 4 kHz, which is why telephone codecs cut off there.

What does the peak hold show me?

It's a slower-falling indicator that tracks the loudest level each bin reached over the last second or so. Useful for catching transient peaks (snare hits, sibilance) that the main bars would jump past too quickly to read.

Related Tools