问题描述
我正在尝试在 .mp4 文件中实现惊人的爆发。我已经能够成功地做到这一点，但是我如何改变原始音频结果在一个数组中。我遇到的问题是“write_videofile”函数似乎只采用特定类型的结构来写入音频文件。到目前为止，这就是我所拥有的任何帮助都会很棒。
#Necessary imports
import os as os
import moviepy.editor as mp
from playsound import playsound
import librosa as librosa
import librosa.display
import sounddevice as sd
import numpy as np
from moviepy.audio.AudioClip import AudioArrayClip
#Set working directory
os.chdir('/Users/matthew/Desktop/Python_startle')
#Upload the video and indicate where in the clip the burst should be inserted
vid = mp.VideoFileClip("37vids_balanced/Baby_Babble.mp4")
audio,sr = librosa.load('37vids_balanced/Baby_Babble.mp4')
librosa.display.waveplot(audio,sr = sr)
sd.play(audio,sr)
audio_duration = librosa.get_duration(filename = '37vids_balanced/Baby_Babble.mp4')
start_point_sec = 6
start_point_samp = start_point_sec * sr
burst_samp = int(sr * .05) #how many samples are in the noise burst
burst_window = audio[start_point_samp : start_point_samp + burst_samp].astype(np.int32) #Isolating part of the audio clip to add burst to
#Creating startle burst
noise = np.random.normal(0,1,len(burst_window))
#Inserting the noise created into the portion of the audio that is wanted
audio[start_point_samp : start_point_samp + burst_samp] = audio[start_point_samp : start_point_samp + burst_samp].astype(np.int64) + noise #Puts the modified segment into the original audio
librosa.display.waveplot(audio,sr)
new_vid = vid.set_audio(audio)
sd.play(new_vid.audio,sr)
#Trying to get the audio into the proper form to convert into a mp4
new_vid.audio = [[i] for i in new_vid.audio] #This makes the data into a list because the 'AudioArrayClip' Could not take in an array because an array is not 'iterable' this might not be the best solution but it has worked so far
new_vid.audio = AudioArrayClip(new_vid.audio,fps = new_vid.fps)
new_vid.write_videofile('37vids_balanced_noise/Baby_Babble.mp4')
也许有一种更简单的方法可以将数组转化为可行的可迭代形式。
解决方法
暂无找到可以解决该程序问题的有效方法，小编努力寻找整理中！
如果你已经找到好的解决方法，欢迎将解决方案带上本链接一起发送给小编。
小编邮箱:dio#foxmail.com (将#修改为@）