Update app.py

app.py

@@ -1,8 +1,11 @@
+
 import gradio as gr
-from pydub import AudioSegment
+import numpy as np
 import os
+import soundfile as sf
+from pydub import AudioSegment
 from io import BytesIO
-import math
+import torchaudio
 
 def get_audio_file_info(audio_file):
     # Read the audio data from the file
@@ -39,31 +42,25 @@ def get_audio_file_info(audio_file):
     # Return the info table
     return info_table
 
-def split_audio(input_file, duration):
+def split_audio_into_clips(audio_file):
     # Read the audio data from the file
-    audio = AudioSegment.from_file(input_file)
+    audio_data, sample_rate = torchaudio.load(audio_file, normalize=True)
 
-    # Calculate the total length and the number of parts
-    total_length = len(audio)
-    num_parts = math.ceil(total_length / (duration * 1000))
+    # Calculate the duration of each 5-second segment
+    segment_duration = 5 * sample_rate
 
-    # Create a folder to store the split audio files
-    output_folder = "output"
-    os.makedirs(output_folder, exist_ok=True)
+    # 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 and save each part
+    # Save each segment to a separate file
     output_files = []
-    for i in range(num_parts):
-        start = i * duration * 1000
-        end = (i + 1) * duration * 1000
-        split_audio = audio[start:end]
-        output_path = os.path.join(output_folder, f"part_{i + 1}.mp3")
-        split_audio.export(output_path, format="mp3")
-        output_files.append(output_path)
+    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:
@@ -72,17 +69,19 @@ def main():
             # <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', label="Upload Audio File", file_extensions=["mp3", "wav"])
+                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_text = gr.Textbox(value="", label="Output Files", disabled=True)
+                output_markdown = gr.Markdown(value="", visible=True)
 
-        split_audio_butt.click(fn=split_audio, inputs=[audio_input], outputs=[output_text])
+        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()
 
 # Create the Gradio interface
-main()
+main()