声音设备输出溢出

Question

我在使用 Python 的 sounddevice 模块时遇到了未知的麻烦。 下面的代码输出一个 GUI，您可以在其中输出用户定义频率的正弦波。此外，还可以使用用户定义的频率和幅度/量的正弦波来调制其幅度。 按下“播放”按钮后，我收到输出溢出并且音频立即开始滞后。 如果我删除代码中的幅度调制选项并且听到平滑的正弦信号，则不会出现此问题。

我知道这段代码可能会使用太多的 CPU 时间来保证音频信号的流畅。

有人在类似的项目中遇到过类似的问题吗？或者有人对声音设备模块足够熟悉来帮助我吗？

代码的很大一部分是从 Github 上的这个示例复制粘贴的。

我绝不是一个经验丰富的程序员。

非常感谢任何帮助！

import argparse
import sys

import tkinter as tk
import sounddevice as sd
import numpy as np

# the following function and parsing part is copy-paste from the link above 
# and with a few modifications of the remaining code wouldn't be crucial I guess.
# I left it there because I wanted to spare myself the modifications.

def int_or_str(text):
    """Helper function for argument parsing."""
    try:
        return int(text)
    except ValueError:
        return text

parser = argparse.ArgumentParser(add_help=False)
parser.add_argument(
    '-l', '--list-devices', action='store_true',
    help='show list of audio devices and exit')
args, remaining = parser.parse_known_args()
if args.list_devices:
    print(sd.query_devices())
    parser.exit(0)
parser = argparse.ArgumentParser(
    description=__doc__,
    formatter_class=argparse.RawDescriptionHelpFormatter,
    parents=[parser])
parser.add_argument(
    'frequency', nargs='?', metavar='FREQUENCY', type=float, default=500,
    help='frequency in Hz (default: %(default)s)')
parser.add_argument(
    '-d', '--device', type=int_or_str,
    help='output device (numeric ID or substring)')
parser.add_argument(
    '-a', '--amplitude', type=float, default=0.2,
    help='amplitude (default: %(default)s)')
args = parser.parse_args(remaining)

start_idx = 0

def play():
    samplerate = sd.query_devices(args.device, 'output')['default_samplerate']

    # the callback function is basically copy-paste from the link abouve 
    # except I added the modulating sine wave
    def callback(outdata, frames, time, status): 
        if status:
            print(status, file=sys.stderr)
        global start_idx
        t = (start_idx + np.arange(frames)) / samplerate
        t = t.reshape(-1, 1)
        am_amount = float(am_amt_1.get())
        a_modulator = np.sin(2*np.pi*float(am_freq_1.get())*t)*am_amount+1-am_amount
        carrier = np.sin(2 * np.pi * float(freq_1.get()) * t)
        outdata[:] = carrier*a_modulator
        start_idx += frames

    with sd.OutputStream(device=args.device, channels=1, callback=callback,
                         samplerate=samplerate):
        input() #don't really know what this is doing but there's no audio at all without it

# setting up the GUI
  
main = tk.Tk()
main.title('Test')
main.geometry('10000x10000')

# entries for frequency in hz and am amount

freq_1 = tk.Entry(main)
freq_1.grid(row = 0, column = 1)

am_freq_1 = tk.Entry(main)
am_freq_1.grid(row=1,column=1)

am_amt_1 = tk.Entry(main)
am_amt_1.grid(row=2,column=1)

# labels

f1 = tk.Label(main,text='Frequency')
f1.grid(row=0,column=0)

amf_1 = tk.Label(main,text='AM Frequency')
amf_1.grid(row=1,column=0)

amamt_1 = tk.Label(main,text='AM Amount')
amamt_1.grid(row=2,column=0)

# play button executing the above play function

pl = tk.Button(main, text='Play',command = play)
pl.grid(row=3,column=1)

main.mainloop()

我使用笔记本电脑的内置输出设备和 M-Audio Fast Track Pro 尝试了该程序。

Answer 1

正如我所怀疑的，只要您不在回调中调用 tkinter 函数，问题就会解决；如果将值传递给 play 函数，一切似乎都正常。

为了简洁起见，我还对其余部分进行了一些重构。您需要在

input()

 中添加 

with

 的原因是设备会立即关闭；我在这里用 sleep 代替了它，所以声音播放 2 秒然后退出。

如果您的声音设备名称中不包含

Speakers

，您需要在 

device=...

 行中进行更改。

import sys
import time
import tkinter as tk

import numpy as np
import sounddevice as sd


def play(*, device, freq: float, am_freq: float, am_amount: float):
    samplerate = sd.query_devices(device, "output")["default_samplerate"]
    start_idx = 0

    def callback(outdata, frames, time, status):
        nonlocal start_idx
        if status:
            print(status, file=sys.stderr)
        t = (start_idx + np.arange(frames)) / samplerate
        t = t.reshape(-1, 1)
        a_modulator = (
            np.sin(2 * np.pi * am_freq * t) * am_amount + 1 - am_amount
        )
        carrier = np.sin(2 * np.pi * freq * t)
        outdata[:] = carrier * a_modulator
        start_idx += frames

    with sd.OutputStream(
        device=device,
        channels=1,
        callback=callback,
        samplerate=samplerate,
    ):
        time.sleep(2)  # Play for a while


def make_box(parent, label, value, y):
    box = tk.Entry(parent)
    box.insert(0, value)
    label = tk.Label(parent, text=label)
    label.grid(row=y, column=0)
    box.grid(row=y, column=1)
    return (label, box)


def main():
    root = tk.Tk()
    root.title("Test")
    freq_label, freq_box = make_box(root, "Frequency", "500", 0)
    am_freq_label, am_freq_box = make_box(root, "AM Frequency", "25", 1)
    am_amt_label, am_amt_box = make_box(root, "AM Amount", "0.5", 2)

    pl = tk.Button(
        root,
        text="Play",
        command=(
            lambda: play(
                device="Speakers",
                freq=float(freq_box.get()),
                am_freq=float(am_freq_box.get() or 0),
                am_amount=float(am_amt_box.get() or 0),
            )
        ),
    )
    pl.grid(row=3, column=1)

    root.mainloop()


if __name__ == "__main__":
    main()