Free Online Spectrum Analyzer

How a Spectrum Analyzer Works: A Complete Guide for Producers and Mixers

Every time you hear a frequency build-up in the low mids or a harsh peak in the high end, you are experiencing a problem that a spectrum analyzer would have shown you instantly. This tool uses the Fast Fourier Transform (FFT) to convert your audio from the time domain into the frequency domain, displaying amplitude at every frequency from 20 Hz to 20,000 Hz simultaneously. The result is a real-time picture of exactly what your mix contains at any given moment.

What the FFT Actually Does

Audio is stored and processed as a stream of samples representing air pressure changes over time. A single audio sample tells you nothing about frequency content. The Fast Fourier Transform is a mathematical algorithm, first described by Cooley and Tukey in 1965 (building on earlier work by Gauss), that transforms a block of time-domain samples into a set of frequency bins, each showing how much energy is present at that frequency.

The Web Audio API's AnalyserNode handles this internally. It applies the FFT to a sliding window of audio samples and returns an array of magnitude values, one per frequency bin. The number of bins is half the FFT size you choose (because of the Nyquist limit). At an FFT size of 4096, you get 2048 frequency bins spread evenly from 0 Hz to half your sample rate. At 44,100 Hz sample rate, that means one bin covers roughly 10.8 Hz. At 48,000 Hz, one bin covers about 11.7 Hz.

FFT Size Trade-offs: Speed vs. Frequency Resolution

The FFT size setting controls a fundamental trade-off between frequency resolution and time resolution. There is no way to maximize both simultaneously. This is sometimes called the Heisenberg uncertainty principle of signal processing.

• 1024 bins (fastest response): At 44,100 Hz, each analysis window covers about 23 ms of audio. Transients and fast attacks show up quickly, but frequency resolution is coarse at approximately 43 Hz per bin. Bass frequencies below 100 Hz are hard to distinguish.
• 2048 bins: 46 ms window, 21 Hz per bin. A practical starting point for drums and percussive content where timing matters.
• 4096 bins (default): 93 ms window, 10.8 Hz per bin. Good general-purpose setting for music. Bass frequencies are reasonably resolved without too much lag.
• 8192 bins (highest detail): 186 ms window, 5.4 Hz per bin. Excellent for identifying specific tonal problems or checking individual bass notes. The display reacts more slowly to fast transients.

For most mixing tasks, 4096 is the right choice. Use 8192 when you need to resolve closely spaced low-frequency content, such as finding a room resonance between 60 Hz and 80 Hz. Use 1024 when you are analyzing a percussive source and need to watch attack transients in real time.

Logarithmic vs. Linear Frequency Scale

Human hearing does not perceive frequency linearly. We perceive pitch on a logarithmic scale. The distance from 100 Hz to 200 Hz (one octave) feels the same as the distance from 2,000 Hz to 4,000 Hz (also one octave), even though the first spans 100 Hz and the second spans 2,000 Hz.

A linear frequency scale places frequencies at equal horizontal spacing per Hz. This means nearly all of the useful musical content (20 Hz to 4,000 Hz) is compressed into the left 20% of the display, while the top octaves above 10,000 Hz take up half the screen. Linear scale is useful for scientific analysis of specific frequencies but unhelpful for music mixing.

The logarithmic scale in this analyzer places each octave at equal width. Sub-bass (20 to 80 Hz), bass (80 to 300 Hz), low mids (300 to 1,000 Hz), mids (1,000 to 4,000 Hz), and highs (4,000 to 20,000 Hz) each get a proportional share of the display. This is how professional hardware spectrum analyzers have worked since the 1970s. Use the Frequency Calculator if you want to look up the exact Hz value of any musical note to compare against what you see on the spectrum.

Reading the Display: What to Look For in Your Mix

A well-balanced mix rarely has a flat spectrum. Genre conventions, arrangement choices, and room acoustics all shape what a "healthy" spectrum looks like. That said, some common problems show up clearly on any spectrum analyzer.

Low-frequency build-up (20 to 150 Hz): If the sub-bass region is consistently brighter than everything else, the mix is likely too heavy below 60 Hz. This is often a headphones vs. speakers problem. Bass-heavy headphones flatter low-end content that sounds overwhelming on full-range speakers or in a club system.

Boxiness in the low mids (200 to 500 Hz): A pronounced peak anywhere in this range often sounds "muddy" or "boxy." Common causes include room modes, guitar amplifiers, and piano resonances. On the spectrum, this appears as a persistent hump that stays elevated even during quieter passages.

Harshness in the presence range (3,000 to 7,000 Hz): An overly bright peak here makes vocals and guitars sound strident. It is one of the most common problems in home studio mixes where the monitoring environment does not reflect the room's acoustic behavior accurately.

Air and sibilance (8,000 to 16,000 Hz): Excessive energy in this range gives vocals an unpleasant "ss" and "ts" sound. The spectrum will show a consistent peak during consonants. This is where a de-esser addresses the problem.

Use this tool alongside the LUFS Loudness Meter to check both frequency balance and overall loudness at the same time. Frequency problems and loudness problems are often connected. A mix with too much low-end energy will tend to measure louder in LUFS than it sounds perceptually.

The Smoothing Control and Why It Matters

The smoothing time constant (set between 0 and 0.95 in this tool) controls how quickly the display reacts to changes in the audio signal. At 0, every frame shows the raw FFT result with no averaging. This is extremely jumpy and useful mainly for very slow, sustained sounds. At 0.8 (the default), the display retains approximately 80% of the previous frame's value and blends in 20% of the new data. This creates a smooth, readable trace that still reacts quickly enough to see transients.

Higher smoothing (0.9 to 0.95) is useful when you want to see the average energy distribution of a track rather than individual transient spikes. Lower smoothing (0.3 to 0.5) reveals fast peaks that tell you about the transient character of drums, plucked strings, and staccato playing.

Peak Hold and Using It for EQ Decisions

When peak hold is enabled, the white markers at the top of each bar show the highest level reached at that frequency. The markers decay slowly rather than resetting immediately. This lets you see the peak frequency content across several bars or even an entire section of a song.

Play a chorus and watch where the peak markers settle. Then play a verse. The difference between peak positions across sections tells you which frequencies are only appearing at certain moments. This technique is useful for identifying a snare drum crack, a particular chord voicing that clusters in the 400 Hz range, or a synth pad that adds a lot of energy above 8 kHz only when the chorus arrives.

For mastering decisions, you can compare your unmastered mix against a reference track by loading each one and observing where the peak hold markers settle across the full range. This is not a replacement for proper reference monitoring, but it gives you a quick visual sanity check. The Decibel Meter complements this by giving you a real-time RMS and peak SPL reading at your monitoring position.

Microphone Mode: Analyzing Your Room

Switching to microphone input turns this tool into a room acoustic analyzer. Play a reference tone or use pink noise (which distributes equal energy per octave) and watch the spectrum. If your room has a 120 Hz resonance, you will see a consistent peak there whenever low-frequency content plays. If you have standing waves between parallel walls, you may see periodic dips and peaks across the bass range.

According to research published by the Audio Engineering Society, untreated rooms commonly show bass mode variations of 20 dB or more below 300 Hz. This is why professional studios invest in bass trapping and room treatment. Even identifying one or two dominant modes using a free browser-based analyzer can guide you toward practical treatment decisions such as placing absorbers in corners or moving your listening position away from the wall behind you.

Hover Readout and Note Names

Moving your cursor over the spectrum canvas shows the exact frequency in Hz or kHz, the approximate dB level at that position, and the nearest musical note name. This makes it possible to connect what you see on the spectrum to what you hear and what you are doing in your DAW.

For example, if you see a persistent peak at around 233 Hz during a chord, the readout will show "Bb3" or "A#3" (depending on enharmonic spelling). That tells you the peak is likely the fundamental of a Bb in the chord voicing, not a room mode or resonance. If the peak is at 247 Hz instead (B3), you are looking at a different note. This specificity helps you decide whether to cut with EQ, address the playing technique, or treat the room.

Connecting Spectrum Analysis to Your Other Tools

Spectrum analysis is most useful when combined with other measurements. After checking frequency balance here, run your finished master through the LUFS Loudness Meter to confirm integrated loudness targets for Spotify (-14 LUFS), Apple Music (-16 LUFS), and YouTube (-14 LUFS). If your mix has too much low-frequency energy, it will push integrated LUFS higher than the perceptual loudness suggests, because bass energy is heavily weighted in the K-weighting filter used by the LUFS standard.

If you are diagnosing a stereo balance problem or want to check whether your mix translates to mono, combine spectrum analysis with a phase correlation check. Heavy bass content that cancels in mono will disappear from the low end of the spectrum when you sum to mono. The Frequency Calculator can help you identify which musical notes correspond to problem frequencies, so you can go back to your DAW and target the right sources during your EQ work.