Update app.py

app.py

@@ -4,30 +4,7 @@ import os
 import soundfile as sf
 from pydub import AudioSegment
 from io import BytesIO
-
-def main():
-    # Gradio Interface
-    with gr.Blocks() as app:
-        gr.Markdown(
-            """
-            # <div align="center"> diablofx Audio Interval Cutter (BETA) </div>
-            Want to [support](https://ko-fi.com/diablofx) me? Or [join AI HUB](https://discord.gg/aihub) for more Help!\n
-            """
-        )
-        with gr.Row():
-            with gr.Column():
-                audio_input = gr.Audio(type='filepath')
-                create_info_butt = gr.Button(value='Get Audio File Info', variant='primary')
-                split_audio_butt = gr.Button(value='Split Audio into 5s Clips', variant='success')
-            with gr.Column():
-                output_markdown = gr.Markdown(value="", visible=True)
-
-        create_info_butt.click(fn=get_audio_file_info, inputs=[audio_input], outputs=[output_markdown])
-        split_audio_butt.click(fn=split_audio_into_clips, inputs=[audio_input])
-
-        app.queue(max_size=1022).launch()
-
-
+import torchaudio
 
 def get_audio_file_info(audio_file):
     # Read the audio data from the file
@@ -66,28 +43,44 @@ def get_audio_file_info(audio_file):
 
 def split_audio_into_clips(audio_file):
     # Read the audio data from the file
-    audio_data, sample_rate = sf.read(audio_file)
+    audio_data, sample_rate = torchaudio.load(audio_file, normalize=True)
 
-    # Convert to mono if it's not mono
-    if len(audio_data.shape) > 1:
-        audio_data = np.mean(audio_data, axis=1)
+    # Calculate the duration of each 5-second segment
+    segment_duration = 5 * sample_rate
 
-    # Convert to Pydub AudioSegment
-    audio_segment = AudioSegment(audio_data.tobytes(), frame_rate=sample_rate, sample_width=audio_data.dtype.itemsize, channels=1)
+    # Split the audio into segments
+    segments = [audio_data[:, i:i+segment_duration] for i in range(0, audio_data.size(1), segment_duration)]
 
-    # Split the audio into 5s clips
-    clips = [audio_segment[i:i + 5000] for i in range(0, len(audio_segment), 5000)]
+    # Save each segment to a separate file
+    output_files = []
+    for i, segment in enumerate(segments):
+        output_file = f"output_segment_{i+1}.wav"
+        torchaudio.save(output_file, segment, sample_rate)
+        output_files.append(output_file)
+
+    return output_files
+
+def main():
+    # Gradio Interface
+    with gr.Blocks() as app:
+        gr.Markdown(
+            """
+            # <div align="center"> diablofx Audio Interval Cutter (BETA) </div>
+            Want to [support](https://ko-fi.com/diablofx) me? Or [join AI HUB](https://discord.gg/aihub) for more Help!\n
+            """
+        )
+        with gr.Row():
+            with gr.Column():
+                audio_input = gr.Audio(type='filepath')
+                create_info_butt = gr.Button(value='Get Audio File Info', variant='primary')
+                split_audio_butt = gr.Button(value='Split Audio into 5s Clips', variant='success')
+            with gr.Column():
+                output_markdown = gr.Markdown(value="", visible=True)
 
-    # Save each clip and create download buttons
-    download_buttons = []
-    for i, clip in enumerate(clips):
-        clip_data = clip.raw_data
-        clip_io = BytesIO(clip_data)
-        clip_name = f"clip_{i+1}.wav"
-        sf.write(clip_name, clip_io.read(), sample_rate)
-        download_buttons.append((f"Download Clip {i+1}", clip_name))
+        create_info_butt.click(fn=get_audio_file_info, inputs=[audio_input], outputs=[output_markdown])
+        split_audio_butt.click(fn=split_audio_into_clips, inputs=[audio_input])
 
-    return download_buttons
+        app.queue(max_size=1022).launch()
 
 # Create the Gradio interface
-main()
+main()