Python Desktop Audio Record Loopback

2024-07-21

筆記如何使用 Python 錄製作業系統桌面環境音效與迴路輸出。

說明

參考 tez3998 於 GitHub 所分享的專案 loopback-capture-sample。

主要透過 soundcard 以及 soundfile 來進行錄製與寫入的作業，同時因為錄製是指定時間長度，自行增加一個函數來清除結尾的空白。

import soundcard as sc
import soundfile as sf
import uuid
import numpy as np

OUTPUT_FILE_NAME = f'{uuid.uuid4()}.wav'    # file name.
SAMPLE_RATE = 48000                         # [Hz]. sampling rate.
RECORD_SEC = 20                             # [sec]. duration recording audio.

def read_and_cutoff(file_path, threshold=0.01):
    # Read audio file
    data, samplerate = sf.read(file_path)
    # Calculate the absolute value of the audio
    abs_data = np.abs(data)
    # Traverse from the end to find the first point that exceeds the threshold
    for i in range(len(abs_data) - 1, 0, -1):
        if np.any(abs_data[i] > threshold):
            break
    # Cut the audio data to remove the silent tail
    cutoff_data = data[:i+1]
    # Write to a new file
    sf.write(file_path, cutoff_data, samplerate)

def record_start():
    with sc.get_microphone(id=str(sc.default_speaker().name), include_loopback=True).recorder(samplerate=SAMPLE_RATE) as mic:
        # record audio with loopback from default speaker.
        data = mic.record(numframes=SAMPLE_RATE*RECORD_SEC)
        
        # change "data=data[:, 0]" to "data=data", if you would like to write audio as multiple-channels.
        sf.write(file=OUTPUT_FILE_NAME, data=data[:, 0], samplerate=SAMPLE_RATE)

record_start()
read_and_cutoff(OUTPUT_FILE_NAME)