AI Tone Match: Automated Spectral EQ
Capture a reference spectrum, capture your sound, and let a 16-band parametric EQ automatically correct the difference.
Spectral Fingerprints
Every sound has a unique spectral shape, a "fingerprint" of how energy is distributed across frequencies. A muffled recording has less high-frequency energy than a bright one. A nasal tone has a peak in the midrange. If you can measure these differences, you can correct them with EQ.
This is how professional mastering engineers match the tonal balance of an album to a reference track. It is how film sound mixers ensure dialogue recorded in different locations sounds consistent. And it is how live sound engineers tune a PA system to flatten room resonances. The technique is always the same: capture a reference spectrum, capture the source, compute the difference, and apply corrective EQ.
How It Works
The tool below captures one second of audio from each source, averages many FFT frames to get a stable spectral profile, then computes a per-band dB difference. That difference drives a bank of 16 peaking EQ filters spaced logarithmically from 60 Hz to 12 kHz. Each filter boosts or cuts by up to ±12 dB to nudge the source toward the reference.
Sixteen bands is a practical compromise. Fewer bands miss narrow resonances; more bands risk correcting noise rather than tonal character. The "Amount" slider lets you dial the correction from 0% (bypass) to 100% (full match), so you can hear each filter gradually reshape the spectrum.
Real-World Applications
In mastering, engineers load a commercially released track as the reference and their mix as the source. The EQ match gives them a starting point that they then refine by ear. In live sound, pink noise through the PA is the reference and a measurement microphone captures the room response. The correction curve flattens the room's coloration. In game audio, spectral matching ensures that sounds recorded from different sources blend naturally in the same virtual environment.
Try it: select a reference preset and a different sound for "Your Sound." Capture both spectra, then sweep the Amount slider to hear the correction reshape the tone in real time.