rs-audio-stats 1.1.8

Professional-grade audio analysis tool with EBU R128 loudness measurement

rs_audio_stats

Professional-grade audio analysis tool with EBU R128 loudness measurement for Python.

概要

rs_audio_statsは、EBU R128標準（ITU-R BS.1770-4）に準拠したラウドネス測定、真のピーク検出、RMS計算、音声正規化機能を提供する包括的な音声解析ライブラリです。

インストール

pip install rs_audio_stats

クイックスタート

import rs_audio_stats as ras

# 音声ファイルを解析
info, results = ras.analyze_audio("audio.wav", True, False, False, False, True, False, False)
print(f"統合ラウドネス: {results.integrated_loudness:.1f} LUFS")
print(f"真のピーク: {results.true_peak:.1f} dBFS")

# ディレクトリの一括解析
results = ras.batch_analyze_directory("audio_folder/", True, False, False, False, True, False, False)
for file_path, (info, analysis) in results.items():
    print(f"{file_path}: {analysis.integrated_loudness:.1f} LUFS")

📖 完全ドキュメント

COMPLETE_API_REFERENCE.md で全機能の詳細な使用方法とサンプルコードをご覧いただけます。

Core Features

📊 Audio Information

• ✅ Sample rate, channels, bit depth (-sr, -ch, -bt)
• ✅ Duration in seconds and formatted time (-du, -tm)
• ✅ Total samples and format detection (-f, -fe, -fea)

🎚️ Loudness Analysis (EBU R128)

• ✅ Integrated loudness measurement (-i)
• ✅ Short-term loudness (-s)
• ✅ Momentary loudness (-m)
• ✅ Loudness range (LRA) (-l)
• ✅ True peak detection (-tp)
• ✅ RMS maximum and average (-rm, -ra)

🎛️ Audio Normalization

• ✅ True peak normalization (-norm-tp)
• ✅ Integrated loudness normalization (-norm-i)
• ✅ Short-term loudness normalization (-norm-s)
• ✅ Momentary loudness normalization (-norm-m)
• ✅ RMS maximum normalization (-norm-rm)
• ✅ RMS average normalization (-norm-ra)

📁 Export Formats

• ✅ CSV export (-csv)
• ✅ TSV export (-tsv)
• ✅ XML export (-xml)
• ✅ JSON export (-json)

🔄 Batch Processing

• ✅ Single file analysis
• ✅ Directory batch processing
• ✅ Recursive file discovery
• ✅ Multiple format support

API Reference

analyze_audio()

info, results = analyze_audio(
    file_path: str,
    integrated_loudness: bool = False,
    short_term_loudness: bool = False,
    momentary_loudness: bool = False,
    loudness_range: bool = False,
    true_peak: bool = False,
    rms_max: bool = False,
    rms_average: bool = False
)

batch_analyze_directory()

results = batch_analyze_directory(
    directory_path: str,
    integrated_loudness: bool = False,
    short_term_loudness: bool = False,
    momentary_loudness: bool = False,
    loudness_range: bool = False,
    true_peak: bool = False,
    rms_max: bool = False,
    rms_average: bool = False
)

Normalization Functions

# True peak normalization
normalize_true_peak(input_path: str, target_dbfs: float, output_path: str)

# Integrated loudness normalization
normalize_integrated_loudness(input_path: str, target_lufs: float, output_path: str)

# Short-term loudness normalization
normalize_short_term_loudness(input_path: str, target_lufs: float, output_path: str)

# Momentary loudness normalization
normalize_momentary_loudness(input_path: str, target_lufs: float, output_path: str)

Export Functions

# Export analysis results
export_to_csv(results, output_path: str)
export_to_tsv(results, output_path: str)
export_to_xml(results, output_path: str)
export_to_json(results, output_path: str)

Use Cases

Broadcasting

# Check compliance with broadcast standards
info, results = ras.analyze_audio("broadcast.wav", True, False, False, True, True, False, False)
if results.integrated_loudness < -23.0:
    print("Meets EBU R128 broadcast standard")

Music Production

# Analyze dynamics and prepare for mastering
info, results = ras.analyze_audio("song.wav", True, True, True, True, True, True, True)
print(f"Dynamic Range: {results.loudness_range:.1f} LU")

Podcast Processing

# Batch normalize podcast episodes
episodes = ras.batch_analyze_directory("episodes/", True, False, False, False, True, False, False)
for file_path, (info, analysis) in episodes.items():
    if analysis.integrated_loudness < -16.0:
        output_path = file_path.replace(".wav", "_normalized.wav")
        ras.normalize_integrated_loudness(file_path, -16.0, output_path)

Supported Formats

• WAV (all PCM variants)
• FLAC
• MP3
• AAC/M4A
• OGG/Vorbis
• And many more via Symphonia decoder

Requirements

• Python 3.10+
• Windows, macOS, or Linux

License

MIT License