Using Audiostream's read() and write() methods simultaneously can result in noise.

[Kevin-Kevin]

here is my code

    with AudioStream(num_output_channels=2,
                     output_device_name=AudioStream.default_output_device_name,
                     buffer_size=20000) as out:
        with AudioStream(
                         input_device_name=AudioStream.input_device_names[2],
                         buffer_size=20000,
                         sample_rate=TARGET_SAMPLE_RATE) as ins:
            try:
                while True:
                    print("reading...")
                    # 88-882
                    buffer= ins.read(1024)
                    buffer = board.process(buffer,sample_rate=SAMPLE_RATE)
                    out.write(buffer,sample_rate=TARGET_SAMPLE_RATE)

but if i only use one audiostream，it behaves well

def test2():
    with AudioStream(
            input_device_name=AudioStream.input_device_names[2],  # Guitar interface
            output_device_name=AudioStream.default_output_device_name
    ) as stream:
        # Audio is now streaming through this pedalboard and out of your speakers!
        stream.plugins = Pedalboard([
            Reverb(room_size=0.25),
            PitchShift(semitones=10)
        ])
        input("Press enter to stop streaming...")

the reason why i use write() and read() is i need change pitch by code without interrupt voice output, board.append() and board.remove() will interrupt output