zfff

app.py

@@ -37,10 +37,12 @@ def send_to_deepgram(audio_file_path):
         audio_data = audio_file.read()
     
     response = requests.post(endpoint, headers=headers, data=audio_data)
-    print(response.json())  # Print the response here
-    return response.json()
+    response_json = response.json()
+    print("Deepgram API Response:", response_json)  # Log the response
+    return response_json
 
 def analyze_audio(audio_file_path):
+    print("Analyzing audio...")  # Log start of analysis
     last_frame = None
     y, sr = librosa.load(audio_file_path)
     chunk_length = 10 * sr # 10 seconds
@@ -55,6 +57,7 @@ def analyze_audio(audio_file_path):
         moods.append(mood)
 
     for i, start in enumerate(range(0, len(y), chunk_length)):
+        print(f"Analyzing chunk {i + 1}...")  # Log chunk analysis
         chunk = y[start:start + chunk_length]
         lyrics_summary = lyrics_chunks[i] if i < len(lyrics_chunks) else 'Instrumental or silence'
         previous_mood = moods[i - 1] if i > 0 else None
@@ -63,8 +66,11 @@ def analyze_audio(audio_file_path):
         _, tempo, chroma_mean, spectral_contrast_mean, zero_crossing_rate_mean, mfcc_mean = analyze_chunk(chunk)
         prompt = generate_video_prompt(previous_mood, current_mood, next_mood, tempo, chroma_mean, spectral_contrast_mean, zero_crossing_rate_mean, mfcc_mean, lyrics_summary)
         description = f"Chunk starting at {start / sr} seconds:<br>Mood: {current_mood}<br>Video Prompt: {prompt}<br><br>"
+        print(f"Generating video for chunk {i + 1}...")
         video = generate_video(prompt, last_frame)
         last_frame = extract_last_frame(video)
+        print(f"Description for chunk {i + 1}: {description}")
+        print(f"Video for chunk {i + 1}: {video}")
 
         # Yield the result for this chunk
         yield (description, video)