Status of mics with PyAudio / Portaudio support (kernel modules) and MALOS Python everloop examples

Hi guys, any update on this?
And please do not answer “soon” :wink: Or maybe explain what is the issue?
I’m just asking as my project to get Kitt_ai working is on hold, pending this.

Hello @OPK

The support for Pyaudio and PortAudio are still in development.

  • Features in development:

    • Playback Support (Audio Output)
    • 96KHz Microphone Sampling rate.
  • Bugs:

    • PortAudio: The config parameter doesn’t set the kernel config. Now it’s necessary to set up in the asound.rc. Unfortunately, it generates other issues. (We are working on this issue along with support ALEXA SDK) PD: We just get it works! but with some workarounds! (High Priority)

    • PyAudio: It seems that the communication with the kernel module is fine. We need to test it more carefully

    • HAL Abstraction: The kernel modules and HAL layer are incompatible. We need to port all of the HAL and MALOS to use Kernel modules.

Now, You could use it and try to use with Kytty AI. If you get some issue or trouble let us know.

I can not set a date for the full support for PortAudio. We are working on it now and I will keep posting updates here in the community about its status.

Thank you for your valuable feedbacks.

Best Regards,

Kevin Patino

Hi Kevin,
Thanks for the details. As you said it would work now:

I just finished another attempt from scratch to get it working, all but to get the same result as before (Kitt_ai wakeword engine still not working on Matrix VOICE), the wakeword is not triggering anything (not kitt_ai nor sensory).

Recording is ok and tne ./micarray_recorder demo works fine without noise.
There are no “pipes” in /tmp as malos is not compatible and not installed

When I tap “Tap to speak” I hear a ding but Alexa does not hear me…

Do you have any valuable feedback?

Greetings I’m also part of the community that works with mycroftDocumentation. We are very interested in getting matrix voice to work with the platform.

Utilizing python-alsaaudio and python-pyalsa is stable. Pulse audio when utilized with mycroft utilizing the matrix kernel modules crashes the kernel. Let me know the best place to provide feedback.

Hi,

Please share the crash message.

– Andrés Calderón

Your latest matrixio-kernel-modules seems to have resolved the issue with pulse audio.

2 Likes

Do you have a tutorial for the setup with Matrix Voice and Mycroft AI?

No guide yet for Mycroft @SpaceGlider. Have you check this Mycroft AI Partnership

-Yoel