可以使用AudioSegment.silent()、AudioSegment.empty()和AudioSegment concatenation的组合：

from pydub import AudioSegment
from pydub.playback import play

sounds = []
sounds.append(AudioSegment.from_file("Downloads/my_file.mp3"))
for i in range(4):
    sounds.append(sounds[i].overlay(sounds[0], position=delays[i+1]))

my_custom_sound = AudioSegment.empty()
for i in range(len(sounds)):
    counter = 0
    while counter <= 2:
        delay = AudioSegment.silent(duration=delays[i]*1000) # Multiply by 1000 to get seconds from delays[i]. silent(duration=1000) is 1 second long
        my_custom_sound += sounds[I] + delay
        play(sounds[i])
        time.sleep(delays[i])
        counter = counter + 1

my_custom_sound.export("/path/to/new/my_custom_sound.mp4", format="mp4")

为此，您可以连接正在生成的音频文件，然后导出此类音频。例如，您可以将脚本修改为：

from pydub import AudioSegment
from pydub.playback import play

track = None
sounds = []
sounds.append(AudioSegment.from_file("Downloads/my_file.mp3"))
for i in range(4):
    sounds.append(sounds[i].overlay(sounds[0], position=delays[i+1]))

for i in range(len(sounds)):
    counter = 0
    while counter <= 2:
        if track is not None:
            track = track + sounds[I]
        else:
            track = sounds[I]
        play(sounds[i])
        time.sleep(delays[i])
        counter = counter + 1

track.export("/path/to/new/my_custom_sound.mp4", format="mp4")