app.py

# app.py
import gradio as gr
from pydub import AudioSegment
from pydub.effects import compress_dynamic_range
from pydub.silence import split_on_silence

def remove_silence(audio_file, silence_thresh, min_silence_len, compression_threshold, amplification):
    # Load the audio file
    audio = AudioSegment.from_file(audio_file)

    # Apply compression to the audio file
    compressed_audio = compress_dynamic_range(audio, threshold=compression_threshold)

    # Split the audio into parts without silence
    chunks = split_on_silence(compressed_audio,
        # Minimum silence length (in milliseconds)
        min_silence_len=min_silence_len,

        # Silence threshold (in dBFS)
        silence_thresh=silence_thresh
    )

    # Check if any chunks were found
    if not chunks:
        return audio_file  # Return the original file if no silence chunks were found

    # Combine chunks without silence
    audio_without_silence = sum(chunks)

    # Amplify the audio
    amplified_audio = audio_without_silence + amplification

    # Export the amplified audio file
    output_file = "audio_without_silence.wav"
    amplified_audio.export(output_file, format="wav")

    return output_file

# Create Gradio interface
iface = gr.Interface(
    fn=remove_silence,
    inputs=[
        gr.Audio(sources=["upload", "microphone"], type="filepath", label="Upload or Record Audio File"),
        gr.Slider(minimum=-100, maximum=0, step=1, value=-16, label="Silence Threshold (dBFS): The threshold in decibels (dBFS) below which audio is considered silence."),
        gr.Slider(minimum=10, maximum=5000, step=10, value=500, label="Minimum Silence Length (ms): The minimum length of silence (in milliseconds) to be detected."),
        gr.Slider(minimum=-100, maximum=0, step=1, value=-20, label="Compression Threshold (dBFS): The threshold in decibels (dBFS) for audio compression. Lower values compress more."),
        gr.Slider(minimum=0, maximum=30, step=1, value=0, label="Amplification (dB): Amplify the volume of the processed audio. Range from 0 dB (no change) to 30 dB.")
    ],
    outputs=gr.Audio(type="filepath", label="Processed Audio Without Silence"),
    title="Remove Silence from Audio",
    description="Upload an audio file or record audio and get the audio with silence removed. Adjust the silence threshold, minimum silence length, compression threshold, and amplification."
)

# Run the app
if __name__ == "__main__":
    iface.launch()