Instrument Frequency Detector

Identify musical instruments by analyzing their harmonic profile and frequency range. See confidence scores for each instrument match, harmonic timbre bar charts, tuning verification, and a reference database covering 15+ instruments. Upload an audio file or use your microphone. 100% browser-based, no audio uploaded.

Instrument Frequency Detector Tool

Understanding Your Results

Instrument Identification via Harmonic Profile

Every musical instrument produces a unique pattern of harmonic overtones when playing a note. A flute is nearly pure fundamental with weak overtones, while a trumpet has many strong harmonics creating its bright, brassy character. This tool measures the relative amplitudes of harmonics H1 through H10 and compares this "fingerprint" against stored templates for each instrument using cosine similarity — a mathematical measure of how closely two profiles point in the same direction in harmonic space.

Confidence Scoring

The confidence percentage represents how closely your sound's harmonic profile matches the best instrument template. 90%+ is a very strong match — the harmonic structure closely resembles the template. 70–89% is a good match with some deviation (possibly due to playing technique, room acoustics, or microphone coloration). 50–69% is a weak match — the sound may be a different variant of the instrument or a different instrument entirely. Below 50% means the harmonic profile does not closely match any template.

Frequency Range Matching

In addition to harmonic profile matching, the tool checks whether the detected fundamental frequency falls within each instrument's known playing range. A note at 262 Hz (middle C) could be a piano, flute, violin, or clarinet — but a note at 41 Hz is almost certainly a bass guitar or cello, because most instruments cannot produce that frequency. Range matching provides a secondary filter that improves identification accuracy.

Tuning Verification

The tuning checker identifies the nearest standard pitch for common tuning systems. For guitar, it checks against standard tuning frequencies (E2=82.4 Hz, A2=110 Hz, D3=146.8 Hz, G3=196 Hz, B3=246.9 Hz, E4=329.6 Hz). For orchestral instruments, it compares against A4=440 Hz concert pitch. It reports the cents deviation — within ±5 cents is considered in tune, ±5–15 cents is slightly off, and beyond ±15 cents is noticeably out of tune.

Timbre Comparison Visualization

The timbre comparison canvas overlays your detected harmonic profile (green bars) against the matched instrument's template profile (cyan outline). When the two patterns closely align, you have a strong match. Harmonics where the detected amplitude significantly exceeds the template may indicate resonance reinforcement, room coloration, or playing technique effects. Harmonics weaker than the template suggest damping or a different instrument variant.

Technical Background

How Instrument Detection Works

The Instrument Frequency Detector uses a multi-stage pipeline to identify instruments from audio. First, the fundamental frequency (F0) is detected using time-domain autocorrelation on the raw audio samples. Autocorrelation finds the signal's period by measuring self-similarity at different time lags — the lag with the strongest correlation peak corresponds to the period of the fundamental pitch. This method correctly identifies the true pitch even when the fundamental is weak or absent (the "missing fundamental" phenomenon).

Once F0 is established, the tool performs an FFT (Fast Fourier Transform) with an 8192-sample window and measures the amplitude at each integer multiple of F0 — specifically H1 (fundamental) through H10. Parabolic interpolation across the three FFT bins nearest each harmonic frequency provides sub-bin amplitude accuracy. The resulting 10-element amplitude vector is normalized relative to H1 to create a scale-independent harmonic profile.

This profile is then compared against a database of 18 instrument templates using cosine similarity: cos(theta) = (A dot B) / (|A| * |B|), where A is the detected profile and B is the template. Cosine similarity ranges from 0 (completely different) to 1 (identical shape), regardless of absolute amplitude. The instrument with the highest similarity score is reported as the primary match. A secondary frequency range filter checks whether the detected F0 falls within each instrument's known fundamental frequency range, penalizing instruments where F0 is outside their playable range. The combined score (weighted cosine similarity + range bonus) determines the final ranking.