music-timbre

📁 benzema216/dreamina-claude-skills 📅 8 days ago
4
总安装量
4
周安装量
#50515
全站排名
安装命令
npx skills add https://github.com/benzema216/dreamina-claude-skills --skill music-timbre

Agent 安装分布

opencode 3
amp 2
kimi-cli 2
codex 2
github-copilot 2
claude-code 2

Skill 文档

/music-timbre — Timbre & Spectral Analysis

Extract timbral and spectral characteristics: MFCC coefficients, spectral centroid/bandwidth/rolloff, loudness (LUFS), dynamic range, brightness, warmth, and optionally separate audio sources.

Usage

/music-timbre <audio_file_path>

Steps

  1. Validate the audio file path
  2. Run timbre analysis:
python3 -m music_analyzer timbre "<audio_file_path>"

Add --no-separation to skip Demucs source separation.

  1. Present results:
    • Brightness: Score 0-1 (higher = brighter timbre)
    • Warmth: Score 0-1 (higher = more low-frequency energy)
    • Dynamic Range: Estimated in dB
    • Loudness: LUFS (if pyloudnorm available)
    • MFCC Summary: 13-coefficient means
    • Stems: Paths to separated vocals/drums/bass/other (if demucs ran)

Output Fields

Field Description
mfcc MFCC means and stds
spectral Centroid, bandwidth, rolloff, ZCR
loudness_lufs Integrated loudness in LUFS
dynamic_range_db Dynamic range in dB
brightness Brightness score 0-1
warmth Warmth score 0-1
stems Separated stem file paths (if available)