How to play audio from a mp4 video?

[Akascape]

I created a video player for tkinter using PIL and pyav, and the video is playing properly without any issue. But it is lacking audio.
I want to know how I can add audio playback by reading the audio frames and play it, maybe using sounddevice or pyaudio...
The source code of the video player is here: https://github.com/Akascape/tkVideoPlayer
(Summary: I simply used the frame = next(self._container.decode(video=0)).to_image() method to read the frames and show them in a label image using pillow)

Here is the progress:

out.mp4

[seeones]

Here is an example of playing audio using PyAudio. How to control the synchronization of audio and video during playback needs to be controlled in the main program.

class AudioRender():
    def play(self, path):
        import av
        import pyaudio
        container: av.container.InputContainer | av.container.Container = av.open(path)
        audio_stream: av.audio.AudioFrame = container.streams.audio[0]

        samplerate = audio_stream.rate
        channels = audio_stream.channels

        p = pyaudio.PyAudio()
        stream = p.open(format=pyaudio.paFloat32,
                        channels=channels,
                        rate=samplerate,
                        output=True)

        for frame in container.decode(audio_stream):
            audio_data = frame.to_ndarray().astype('float32')
            interleaved_data = audio_data.T.flatten().tobytes()
            stream.write(interleaved_data)

        stream.stop_stream()
        stream.close()
        p.terminate()

[Akascape]

@seeones Thanks, it is working, but there is a problem with sync, when I decode using next(self._container.decode(video=0, audio=0)), it generates video and audio frames unevenly with some video files which causes stuttering.

Example:
This type of decoding works good with sync:

video_frame

audio_frame
audio_frame

video_frame

audio_frame
audio_frame

video_frame
...

But this decoding doesn't sync properly:

video_frame
video_frame
video_frame

audio_frame 
audio_frame
audio_frame
audio_frame
audio_frame
audio_frame

video_frame
video_frame
video_frame
...

Is there any way to evenly decode both frame types?

This is the exact code:

frame = next(self._container.decode(video=0, audio=0)) # simultaneously decode both audio and video

if 'Video' in repr(frame):
         # show the video frame
         self._current_img = frame.to_image()
         self._frame_number += 1               
else:
        # play the audio frame
         audio_data = frame.to_ndarray().astype('float32')
         interleaved_data = audio_data.T.flatten().tobytes()
         audio_stream.write(interleaved_data)