Playing a file with PyAudio about pyogg

[DieTapete]

Right, sorry that was just that one file i used for testing.. i thought it was similar to some issue that i had beforehand when it tried to "play" the header of the wave file. So everything is fine. Closing this issue now.

[mattgwwalker]

Hi, Thanks for your interest in PyOgg. There are a fair few examples in the library already (although they don’t use PyAudio). Have you tried those examples yet? We’re you able to get any audio to play? Cheers, Matthew

…

On Sun, 18 Jul 2021 at 11:01, DieTapete ***@***.***> wrote: Hi there, I am having trouble getting this library to work with pyaudio. I just want to decode an opus file and play it back but no combination of options seems to work. Could you provide an example? — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#87>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA653XAV62YHHFKDWFOIA4LTYIDVTANCNFSM5ARQCETQ> .

[DieTapete]

Yes, I tried those examples and managed to succeed to play a sound file. However none of the examples actually use streaming to play a file..

[DieTapete]

I managed to get it kind of working but it sounds weird (a bit too fast and like with a phaser effect):

import pyaudio
from pyogg import OpusFile, OpusFileStream

def play(filename: str):
    p = pyaudio.PyAudio()
    stream = OpusFileStream(filename)
    out_stream = p.open(format=p.get_format_from_width(width=2),
                         channels=stream.channels,
                         rate=stream.frequency,
                         output=True,
                         output_device_index=1)
    
    print("stream.bytes_per_sample {} channels: {}, frequency:{}".format(stream.bytes_per_sample, stream.channels, stream.frequency))
    
    while True:
        # Read the next part of the stream
        buf = stream.get_buffer_as_array()        
        # Check if we've reached the end of the stream
        if buf is None:
            break
        out_stream.write(buf)
        
    out_stream.stop_stream()
    out_stream.close()
    p.terminate()

[Zuzu-Typ]

Hey there,

well, you've basically already found the solution yourself.
The only thing you need to change is how you pass on the buffer.

out_stream.write() seems to be expecting 8-bit unsigned integer values (i.e. bytes) by default and thus, when you pass the audio data array containing 16-bit signed integers, it tries to interpret those as 8-bit, resulting in double speed and clipping.

There's two ways of changing this. Either you use the get_buffer() function instead of the get_buffer_as_array() function (which is called internally anyways, so it's not like you'd lose anything), or you add another argument to the out_stream.write:

out_stream.write(buf, len(buf))

Both seem to work.

By the way, you can use stream.bytes_per_sample instead of the constant width of 2 in the get_format_from_width() function.

Cheers
--Zuzu_Typ--

[DieTapete]

Thank you for that quick reply @Zuzu-Typ! I got it working now. :)

[DieTapete]

BTW: I still get a cracking sound in the beginning of the file. I avoid it for now by just skipping the first buffer.. is that the right way to go?

[Zuzu-Typ]

Thank you for that quick reply @Zuzu-Typ! I got it working now. :)

You're welcome (:

BTW: I still get a cracking sound in the beginning of the file. I avoid it for now by just skipping the first buffer.. is that the right way to go?

Hm.. I've tried two different files, but didn't get any cracking, at least I didn't notice any.
Could you please try using a different file to check if it's doing the same thing?
This is one of the two files I've tested: https://github.com/TeamPyOgg/PyOgg/blob/master/examples/left-right-demo-5s.opus