I am trying to set up audio transcription using VOLK and Wave in Python. The final issue i have is that when I try to read the total frames in my audio file, the number is much higher than the actual frames in the file. For example, one audio file that I transcribed only contained 3440000 frames, but my program read 9779874 total frames. The console output was therefore "Progress: 3.52% Transcribed, processed 344000 / 9779874 frames[DEBUG] End of file reached."

My file processing function is below. Any help would be appreciated.

def process_file(self, file_path):
        print(f"[DEBUG] Started processing file: {file_path}")
        extension = os.path.splitext(file_path)[1].lower()
        print(f"[DEBUG] File extension: {extension}")
        
        #If it's an MP3, convert to WAV
        if extension == ".mp3":
            file_path = convert_mp3_to_wav(file_path)
            if not file_path:
                print("[ERROR] Conversion failed, skipping file.")
                return
        elif extension != ".wav":
            print(f"[ERROR] Unsupported file format: {extension}")
            return

        # Open the audio file
        try:
            wf = wave.open(file_path, "rb")
            print(f"[DEBUG] Opened WAV file: {file_path} ({wf.getnchannels()} channels, {wf.getsampwidth()} width, {wf.getframerate()} Hz)")
        except Exception as e:
            print(f"[ERROR] Failed to open file {file_path}: {e}")
            return
        
        # Check if audio is in the right format
        if wf.getnchannels() != 1 or wf.getsampwidth() != 2 or wf.getframerate() != 16000:
            print(f"[ERROR] Audio format is incorrect. Expected 1 channel, 2 bytes width, 16000 Hz sample rate.")
            return
        
        recognizer = KaldiRecognizer(model, wf.getframerate())
        total_frames = wf.getnframes()
        processed_frames = 0
        text_result = ""
        start_time = time.time()
        
        print(f"[DEBUG] Starting transcription...")

        while True:
            data = wf.readframes(4000)
            if len(data) == 0:
                print("[DEBUG] End of file reached.")
                break
            if recognizer.AcceptWaveform(data):
                processed_frames += len(data)
                percentage = (processed_frames / total_frames) * 100
                result = json.loads(recognizer.Result())
                text_result += result.get("text", "") + " "
                print(f" [DEBUG] Transcribed chunk: {result.get('text', '')}")
                sample_width = wf.getsampwidth()
                print(f"Sample width (bytes per sample): {sample_width}")
                total_frames = wf.getnframes()
                print(f"\rProgress: {percentage:.2f}% Transcribed, processed {processed_frames} / {total_frames} frames", end="")

        end_time = time.time()
        transcription_duration = end_time - start_time

        print(f"[DEBUG] Finished transcription in {transcription_duration:.2f} seconds.")
        ```