Menu
Advanced audio analysis and feature extraction using librosa with optional Vamp plugin support via sonic-annotator
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Use when working with the Sleepy Circuits Hypno 2 video synthesizer/resampler — 2-channel video mixer/looper, shader engine, MIDI CC mapping, CV/modulation, sampler. Firmware v0.0.163+.
Hardware instrument skills — reference guides for each device in the music studio setup.
GLSL fragment shader techniques adapted for the Hypno 2 video synthesizer — single-pass generative visuals and video effects optimized for Raspberry Pi 5, 5-uniform limit, CC-mapped parameters
Korg KAOSS DJ — USB DJ controller with X-Y touchpad, Serato DJ Intro integration, KAOSS effects, EQ, crossfader, and MIDI on Linux
Korg KAOSSILATOR dynamic phrase synthesizer (2007 model) — X-Y touchpad, 100 programs, gate arpeggiator, phrase loop recording, USB MIDI on Linux
Use when working with the Korg kaossilator 2S dynamic phrase synthesizer — X-Y touchpad, 150 programs, 50 arp patterns, loop recording, audio player, master recorder, microSD storage, and USB MIDI on Linux.
| name | daw-master:audio-analyzer |
| version | 1.0.0 |
| description | Advanced audio analysis and feature extraction using librosa with optional Vamp plugin support via sonic-annotator |
| author | Hermes Agent |
| license | MIT |
| metadata | {"hermes":{"tags":["audio","analysis","feature-extraction","librosa","vamp","statistics"],"provides":["analyze","probe","extract_batch"],"requires":["python","librosa","soundfile","numpy"],"optional":["sonic-annotator"]}} |
Comprehensive audio analysis skill for extracting musical features, spectral characteristics, and metadata from audio files. Integrates with the daw-master meta-skill pipeline.
| Operation | Description | Parameters |
|---|---|---|
analyze | Extract comprehensive feature set from a single audio file | file, features (list, optional subset) |
probe | Basic file metadata (duration, sample rate, channels, format) | file |
extract_batch | Recursively analyze directory, write results to CSV/JSON | directory, pattern, output_format, output_file |
Required (Python):
pip install librosa soundfile numpy
Optional (Vamp plugins):
sudo apt install sonic-annotator
# Install Vamp plugin packs:
# qm-vamp-plugins: tempo tracking, chroma, onset detection
# vamp-example-plugins: basic features
from skills.daw_master.audio_analyzer import analyze, probe, extract_batch
# Single file analysis
features = analyze("input.wav")
print(f"Tempo: {features['tempo']} BPM")
print(f"Key: {features['key']}")
# Batch processing
extract_batch(
directory="./samples",
pattern="*.wav",
output_format="csv",
output_file="features.csv"
)
analyze() returns a dictionary containing:
{
"duration": 180.5,
"sample_rate": 44100,
"channels": 2,
"tempo": 128.0,
"tempo_confidence": 0.85,
"key": "C major",
"key_confidence": 0.72,
"loudness": -12.3,
"spectral_centroid": 2450.0,
"spectral_rolloff": 5800.0,
"spectral_flatness": 0.15,
"mfcc": [mean_vector, std_vector],
"harmonic_ratio": 0.68,
"percussive_ratio": 0.32,
"onset_count": 240,
"zero_crossing_rate": 0.08
}
All audio-analyzer operations are pipeable:
# Chain with sox-engine for preprocessing
from skills.daw_master.sox_engine import trim, normalize
from skills.daw_master.audio_analyzer import analyze
# Normalize then analyze
normalize("raw.wav", "normalized.wav")
features = analyze("normalized.wav")
pattern="**/*.wav" for recursive directory scanningvamp_plugins key in output