[
  "rms_energy",
  "rms_std",
  "spectral_centroid_mean",
  "spectral_centroid_std",
  "spectral_flatness_mean",
  "spectral_flatness_std",
  "spectral_bandwidth_mean",
  "spectral_bandwidth_std",
  "spectral_rolloff_mean",
  "spectral_rolloff_std",
  "spectral_contrast_mean",
  "spectral_contrast_std",
  "mfcc_variance",
  "mfcc_delta_var",
  "mfcc_delta2_var",
  "mel_flatness",
  "tempo_bpm",
  "tempo_stability",
  "tempo_cv",
  "zero_crossing_rate",
  "zero_crossing_std",
  "onset_strength_mean",
  "onset_strength_std",
  "rms_dynamic_range",
  "beat_count",
  "chroma_entropy",
  "chroma_std",
  "chroma_transition_rate",
  "harmonic_ratio",
  "tonnetz_std",
  "spectral_regularity",
  "temporal_patterns",
  "harmonic_structure",
  "has_vocals",
  "vocal_confidence",
  "vocal_ai_score",
  "pitch_stability_score",
  "vibrato_regularity_score",
  "formant_consistency_score",
  "breath_pattern_score",
  "vocal_texture_score",
  "pitch_mean_hz",
  "pitch_std_cents",
  "vibrato_rate_hz",
  "vibrato_extent_cents",
  "vocal_harmonic_ratio",
  "vocal_energy_ratio"
]