使用pyav、ffmpeg、python解码mp3文件后如何使用音频帧

如何解决使用pyav、ffmpeg、python解码mp3文件后如何使用音频帧

我正在使用 python 和 pyav、ffmpeg 来解码内存中的 mp3。我知道还有其他一些方法可以做到这一点，比如 pipe ffmpeg 命令。但是，我想探索 pyav 和 ffmpeg API。所以我有以下代码。它可以工作，但声音非常嘈杂，虽然可以听到：

import numpy as np
import av # to convert mp3 to wav using ffmpeg
import pyaudio # to play music

mp3_path = 'D:/MyProg/python/SauTimThiepHong.mp3'

def decodeStream(mp3_path):
  # Run NOT OK
  
  container = av.open(mp3_path)
  stream = next(s for s in container.streams if s.type == 'audio')
  frame_count = 0
  data = bytearray()
  for packet in container.demux(stream):
    # <class 'av.packet.Packet'>
    # We need to skip the "flushing" packets that `demux` generates.
    #if frame_count == 5000 : break         
    if packet.dts is None:
        continue
    for frame in packet.decode():   
        #<av.AudioFrame 0,pts=1843200,1152 samples at 44100Hz,stereo,fltp at 0x1d074bd2e48>
        # type(frame) : <class 'av.audio.frame.AudioFrame'>
        #frame.samples = 1152 : 1152 diem du lieu : Number of audio samples (per channel)
        # moi frame co size = 1152 (diem) * 2 (channels) * 4 (bytes / diem) = 9216 bytes
        # 11021 frames
        #arr = frame.to_ndarray() # arr.nbytes = 9216

        #channels = []  
        channels = frame.to_ndarray().astype("float16")
        #for plane in frame.planes:
            #channels.append(plane.to_bytes()) #plane has 4 bytes / sample,but audio has only 2 bytes
        #    channels.append(np.frombuffer(plane,dtype=np.single).astype("float16"))
            #channels.append(np.frombuffer(plane,dtype=np.single)) # kieu np.single co 4 bytes
        if not frame.is_corrupt:
            #data.extend(np.frombuffer(frame.planes[0],dtype=np.single).astype("float16")) # 1 channel: noisy
            # type(planes) : <class 'av.audio.plane.AudioPlane'>
            frame_count += 1
            #print( '>>>> %04d' % frame_count,frame)   
            #if frame_count == 5000 : break     
            # mix channels:
            for i in range(frame.samples):                
                for ch in channels: # dec_ctx->channels
                    data.extend(ch[i]) #noisy
                    #fwrite(frame->data[ch] + data_size*i,1,data_size,outfile)
  return bytes(data)

我使用管道 ffmpeg 来获取解码数据进行比较并发现它们不同：

def RunFFMPEG(mp3_path,target_fs = "44100"):
    # Run OK
    import subprocess
    # init command
    ffmpeg_command = ["ffmpeg","-i",mp3_path,"-ab","128k","-acodec","pcm_s16le","-ac","0","-ar",target_fs,"-map","0:a","-map_metadata","-1","-sn","-vn","-y","-f","wav","pipe:1"]
    # excute ffmpeg command
    pipe = subprocess.run(ffmpeg_command,stdout=subprocess.PIPE,stderr=subprocess.PIPE,bufsize= 10**8)
    # debug
    #print(pipe.stdout,pipe.stderr)
    # read signal as numpy array and assign sampling rate
    #audio_np = np.frombuffer(buffer=pipe.stdout,dtype=np.uint16,offset=44)
    #audio_np = np.frombuffer(buffer=pipe.stdout,dtype=np.uint16)
    #sig,fs  = audio_np,target_fs
    #return audio_np
    return pipe.stdout[78:]     

然后我用pyaudio播放数据，发现很嘈杂

p = pyaudio.PyAudio()
streamOut = p.open(format=pyaudio.paInt16,channels=2,rate= 44100,output=True)
#streamOut = p.open(format=pyaudio.paInt16,channels=1,output=True)

mydata = decodeStream(mp3_path)
print("bytes of mydata = ",len(mydata))
#print("bytes of mydata = ",mydata.nbytes)

ffMpegdata = RunFFMPEG(mp3_path)
print("bytes of ffMpegdata = ",len(ffMpegdata)) 
#print("bytes of ffMpegdata = ",ffMpegdata.nbytes)

minlen = min(len(mydata),len(ffMpegdata))
print("mydata == ffMpegdata",mydata[:minlen] == ffMpegdata[:minlen]) # ffMpegdata.tobytes()[:minlen] )

#bytes of mydata =  50784768
#bytes of ffMpegdata =  50784768
#mydata == ffMpegdata False

streamOut.write(mydata)
streamOut.write(ffMpegdata)
streamOut.stop_stream()
streamOut.close()
p.terminate()

请帮助我理解 pyav api 的解码帧（在 for frame in packet.decode() 之后：）。 应该处理更多吗？还是我有一些错误？

这让我疯狂了 3 天。我猜不出要去哪里。

非常感谢。