# Genre Classifier AI Classify a track into genres from measured evidence: GTZAN timbral-texture features, beat-grid pulse shape, band balance, key and dynamics — with a ranked top-3 and the reason behind every pick. > Canonical page: https://elysiatools.com/en/tools/genre-classifier-ai - **Category:** Media - **Keywords:** genre classification, music genre, genre identifier, audio analysis, gtzan features, spectral centroid, bpm detection, key detection, music information retrieval, four on the floor, tempo, dance music analysis ## Overview The classifier implements the classic recipe that founded music-genre classification: Tzanetakis & Cook (2002) timbral texture — means and variances of spectral centroid, 85% spectral rolloff, spectral flux and zero-crossing rate over 23 ms windows, plus the low-energy descriptor (the share of windows quieter than the mean) — extended with the two other GTZAN groups: rhythmic content (kick/snare/hi-hat onset densities, four-on-the-floor regularity from the beat grid, tempo via autocorrelation with half-time and double-time candidate testing) and pitch content (Krumhansl-Schmuckler key detection on a chroma profile). Seventeen genre profiles — house, techno, drum-and-bass, dubstep, hip-hop, trap, pop, rock, metal, punk, reggae, funk/disco, jazz, classical, folk, ambient, lo-fi hip-hop — score every feature against windows from published tempo guides (Ableton: house 115-130, techno 120-140, dnb 160-180 BPM; hip-hop 60-100 with half-time at 130-160). The report shows the winning genre, a top-3 ranking with confidences, and the concrete measurements behind each choice, plus a feature snapshot (BPM, key, brightness, bass share, dynamic swing) that doubles as DJ metadata. Honest expectation-setting: the original GTZAN classifier scored 61% on 10 classes; treat this heuristic scorer as a well-informed first opinion, not a label-machine. ## Inputs - **Music track** (file) - **Tempo feel** (select) ## When to use - When organizing unlabelled audio files and needing preliminary genre tags alongside BPM and musical key. - When analyzing track stems or production demos to check rhythmic pulse regularity and spectral brightness against genre standards. - When auditing music libraries where half-time or double-time tempo ambiguity needs guided heuristic evaluation. ## How it works - Upload an audio file (up to 100 MB) and select whether tempo detection should evaluate standard, half-time, or double-time feel. - The analysis engine extracts timbral features across 23 ms windows (spectral centroid, 85% rolloff, flux, zero-crossing rate, low-energy share), detects rhythm onsets (kick, snare, hi-hat), and calculates key via Krumhansl-Schmuckler chroma profiling. - Extracted metrics are scored against 17 genre profiles based on tempo windows, four-on-the-floor regularity, dynamic swing, and frequency distribution. - An interactive HTML report displays the top-3 genre candidates with confidence scores, explanatory evidence for each pick, and a feature snapshot summary. ## Use cases - DJs and crate diggers generating key, tempo, and genre metadata snapshots for digital audio collections. - Electronic music producers evaluating kick-snare onset density and four-on-the-floor alignment in dance tracks. - Audio archivists automatically categorizing untagged instrumental and ambient recordings using timbral texture data. ## Frequently asked questions ### What audio formats and file sizes are supported? The classifier accepts standard audio formats up to a maximum file size of 100 MB per upload. ### How does the tempo feel selector affect classification? The 'Auto' setting evaluates half-time and double-time candidates automatically, while 'Half-time' or 'Double-time' manually locks the tempo evaluation window to prevent octave tempo errors. ### Which musical genres can the tool detect? It evaluates 17 profiles: house, techno, drum-and-bass, dubstep, hip-hop, trap, pop, rock, metal, punk, reggae, funk/disco, jazz, classical, folk, ambient, and lo-fi hip-hop. ### What metadata metrics are included in the feature snapshot? The report provides detected BPM, musical key, spectral brightness (centroid), bass frequency share, dynamic swing, and four-on-the-floor regularity percentage. ### Why does the tool show a top-3 ranking instead of a single label? Music frequently blends characteristics of multiple styles. The top-3 ranking shows secondary genre influences along with the concrete acoustic reasons behind each match. ## Related tools - [Podcast Chapter Marker Builder](https://elysiatools.com/en/tools/podcast-chapter-marker-builder): Build every podcast chapter format from one timecoded list: Podcasting 2.0 JSON + RSS tag, ID3v2.4 CHAP+CTOC burned into an MP3, Vorbis comments, mp4chaps, YouTube timestamps and SRT, with a per-player support matrix. - [Instrument Identifier AI](https://elysiatools.com/en/tools/instrument-identifier-ai): Detect which instruments play in a track: onset-split note analysis scores every event against 13 timbre profiles and reports ranked instruments, families and per-note evidence. - [Image Metadata Extractor](https://elysiatools.com/en/tools/image-metadata): Extract comprehensive metadata from images including EXIF, IPTC, XMP, and technical information - [Karaoke Vocal Reducer](https://elysiatools.com/en/tools/audio-karaoke-track): Attempt to remove vocals from a stereo track - [Convert GIF to WebM](https://elysiatools.com/en/tools/gif-to-webm): Convert GIF images to WebM video format with modern compression and web optimization - [Mastering Chain Pro](https://elysiatools.com/en/tools/mastering-chain-pro): Run a full music mastering chain on one track: sub-rumble high-pass, corrective EQ, glue compression, character EQ, saturation, stereo widening, true-peak limiter, and two-pass linear loudness normalization to 2026 platform targets. - [Convert GIF to Raw Pixel Buffer](https://elysiatools.com/en/tools/gif-to-raw): Export GIF frames as raw pixel buffer data for analysis, rendering pipelines, and low-level image processing. - [Audio Click Removal](https://elysiatools.com/en/tools/audio-click-removal): Remove clicks and pops from vinyl recordings ## Samples - [Copyright-Free MP3 Audio Samples](https://elysiatools.com/en/samples/mp3-samples): Collection of royalty-free audio samples for testing and development purposes including nature sounds, meditation music, and ambient audio - [Copyright-Free FLAC Audio Samples](https://elysiatools.com/en/samples/flac-samples): Lossless FLAC audio samples for testing and development, mirrored from MP3 set with nature sounds and meditation music - [Copyright-Free WAV Audio Samples](https://elysiatools.com/en/samples/wav-samples): Uncompressed PCM WAV audio samples for testing and development, mirrored from MP3 set with nature sounds and meditation music - [Copyright-Free Raw PCM Audio Samples](https://elysiatools.com/en/samples/pcm-samples): Raw PCM s16le audio samples featuring nature ambience and relaxing music for waveform, playback, and conversion workflows