Update app.py

app.py

@@ -9,6 +9,14 @@ import subprocess
 # Check if CUDA is available
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
+# Check if sox is installed and install it if necessary
+try:
+    subprocess.run(["sox", "--version"], check=True, capture_output=True)
+except FileNotFoundError:
+    print("sox is not installed. Installing it now...")
+    subprocess.run(["apt-get", "install", "-y", "sox"], check=True)
+    print("sox has been installed.")
+
 # Define the inference function
 @spaces.GPU
 def inference(audio_file, model_name, vocals, drums, bass, other, mp3, mp3_bitrate):
@@ -57,12 +65,9 @@ def inference(audio_file, model_name, vocals, drums, bass, other, mp3, mp3_bitra
         # If only one stem is selected, just copy it
         os.rename(selected_stems[0], output_file)
     else:
-        # Otherwise, use ffmpeg to mix the stems
-        ffmpeg_cmd = ["ffmpeg", "-y"]
-        for stem in selected_stems:
-            ffmpeg_cmd.extend(["-i", f"\"{stem}\""])  # Quote file paths
-        ffmpeg_cmd.extend(["-filter_complex", f"amix=inputs={len(selected_stems)}:duration=longest", f"\"{output_file}\""])  # Quote file paths
-        subprocess.run(ffmpeg_cmd, check=True)
+        # Otherwise, use sox to mix the stems
+        sox_cmd = ["sox", "-m"] + selected_stems + [output_file]
+        subprocess.run(sox_cmd, check=True)
 
     return output_file
 
@@ -71,13 +76,13 @@ iface = gr.Interface(
     fn=inference,
     inputs=[
         gr.Audio(type="filepath"),
-        gr.Dropdown(["htdemucs", "htdemucs_ft", "htdemucs_6s", "hdemucs_mmi", "mdx", "mdx_extra", "mdx_q", "mdx_extra_q"], label="Model Name"),
+        gr.Dropdown(["htdemucs", "htdemucs_ft", "htdemucs_6s", "hdemucs_mmi", "mdx", "mdx_extra", "mdx_q", "mdx_extra_q"], label="Model Name", value="htdemucs_ft"),  # Set default value
         gr.Checkbox(label="Vocals", value=True),
         gr.Checkbox(label="Drums", value=True),
         gr.Checkbox(label="Bass", value=True),
         gr.Checkbox(label="Other", value=True),
-        gr.Checkbox(label="Save as MP3"),
-        gr.Slider(128, 320, step=32, label="MP3 Bitrate", visible=False),
+        gr.Checkbox(label="Save as MP3", value=False),  # Set default value to False
+        gr.Slider(128, 320, step=32, label="MP3 Bitrate", visible=True),  # Set visible to True
     ],
     outputs=gr.Audio(type="filepath"),
     title="Demucs Music Source Separation and Mixing",