Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Nice, I've been looking to build something based around some ESP32 boards. But I can't seem to find a good echo cancellation algorithm. I guess the simple solution is to have a push to talk button on each end, but it would be nice to be able to do full duplex.


You might want to start here: https://github.com/pulseaudio/pulseaudio/tree/master/src/mod...

It would be very interesting to port that code. Another option, obviously, is to use a Raspberry Pi instead with pulseaudio and module-echo-cancel enabled.


I'll take a look - really useful - thanks!


do you think it has enough computer power to handle the audio signal?


Definitely! It's a dual core 240MHz processor. I've been running a bunch of tensorflow code on it for wake word detection.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: