Create utils/audio_processor.py

utils/audio_processor.py

@@ -0,0 +1,73 @@
+"""
+Audio processing utilities
+"""
+import numpy as np
+import librosa
+import soundfile as sf
+from typing import Tuple, Optional
+import logging
+
+logger = logging.getLogger(__name__)
+
+class AudioProcessor:
+    def __init__(self):
+        self.sample_rate = 16000  # Default sample rate for models
+    
+    def load_audio(self, file_path: str) -> Tuple[np.ndarray, int]:
+        """Load audio file and convert to appropriate format"""
+        try:
+            audio, sr = librosa.load(file_path, sr=self.sample_rate)
+            return audio, sr
+        except Exception as e:
+            logger.error(f"Failed to load audio: {str(e)}")
+            raise
+    
+    def save_audio(self, audio_array: np.ndarray, output_path: str, sample_rate: int = None):
+        """Save audio array to file"""
+        try:
+            sr = sample_rate or self.sample_rate
+            sf.write(output_path, audio_array, sr)
+            logger.info(f"Audio saved to {output_path}")
+        except Exception as e:
+            logger.error(f"Failed to save audio: {str(e)}")
+            raise
+    
+    def normalize_audio(self, audio: np.ndarray) -> np.ndarray:
+        """Normalize audio to [-1, 1] range"""
+        return audio / np.max(np.abs(audio))
+    
+    def trim_silence(self, audio: np.ndarray, threshold: float = 0.01) -> np.ndarray:
+        """Remove silence from beginning and end"""
+        return librosa.effects.trim(audio, top_db=20, frame_length=512, hop_length=256)[0]
+    
+    def change_speed(self, audio: np.ndarray, speed_factor: float) -> np.ndarray:
+        """Change playback speed without changing pitch"""
+        return librosa.effects.time_stretch(audio, rate=speed_factor)
+    
+    def change_pitch(self, audio: np.ndarray, n_steps: float) -> np.ndarray:
+        """Change pitch by n semitones"""
+        return librosa.effects.pitch_shift(audio, sr=self.sample_rate, n_steps=n_steps)
+    
+    def get_spectrogram(self, audio: np.ndarray) -> np.ndarray:
+        """Generate spectrogram for visualization"""
+        return librosa.stft(audio)
+    
+    def get_tempo(self, audio: np.ndarray) -> float:
+        """Estimate tempo (BPM)"""
+        tempo, _ = librosa.beat.beat_track(y=audio, sr=self.sample_rate)
+        return tempo
+    
+    def apply_fade(self, audio: np.ndarray, fade_in: float = 0.1, fade_out: float = 0.1) -> np.ndarray:
+        """Apply fade in/out"""
+        fade_in_samples = int(fade_in * self.sample_rate)
+        fade_out_samples = int(fade_out * self.sample_rate)
+        
+        if fade_in_samples > 0:
+            fade_in_curve = np.linspace(0, 1, fade_in_samples)
+            audio[:fade_in_samples] *= fade_in_curve
+        
+        if fade_out_samples > 0:
+            fade_out_curve = np.linspace(1, 0, fade_out_samples)
+            audio[-fade_out_samples:] *= fade_out_curve
+        
+        return audio