Page 1 of 1
Posted: Wed Aug 01, 2018 5:26 pm
I Really Want To Use The AEC especially in living Room because I usually have music playing so it Would Really improve The quality for The Voice recognition. However The Audio Output of The board Is definetly a Lot worse than My speaker setup. So The question Is wether there Is a easy way To Input The audio output To The board so it can still Use AEC (I Use The respeaker mic Array v2 with a raspberry Pi 3b+)
Thank You for The help
Posted: Fri Aug 03, 2018 12:51 am
I have a similar issue. This is related to the mic array telling the system it is only capable of handling 16 kHz audio output. AEC will probably not work using the audio output of your raspberry as the xmos chip on the mic array can't process the audio output... If this would be possible it would awesome
Posted: Fri Aug 03, 2018 9:10 am
Hi, I talked with the software engineer and the 16Khz is related with algorithm. Current we do not have plan to update the algorithm. thanks for understanding.
Posted: Mon Aug 06, 2018 4:59 pm
carlos it does not work at all when I have a higher rate? because my current plan was building a easy mixer connecting the audio outputs of all my audio devices in and mixing them so they are equally loud. Then connect the resulting audio to the mic array for AEC. The real playback I would not do with the array so it has a better quality while still getting filtered quite well.
so bill the question of mine is not wether you will give a better output quality, but wether it works when I connect audio with higher sample rates, because my understanding was when my input is for example 48kHz it still works, just the quality is not really awesome.
Thank you for the help
Posted: Tue Aug 07, 2018 3:43 pm
And I could use a general explanation:
Whats Aux_sel for?
what are Aux_R and Aux_L for? can I use the to input audio that I want the respeaker to play? ( so it gets used with AEC)
I need to convert all my audio to 16khz before inputting it? (in case the above is even possible)
if it the two aux pins are audio out, then how does the reapeaker core sneds the mic array the audio to play?
At least I could not really find the explanations in the respeaker mic array v2 documentation
Posted: Thu Aug 09, 2018 4:35 pm
Aux_sel is designed for respeaker core v1.0(MT7688) to switch the aux output only.
For AEC, please try https://github.com/voice-engine/ec
We still does not suggest to use the mic array v2.0 to play music. thanks.
Posted: Sun Aug 12, 2018 5:06 pm
I'm using Pulseaudio and have configured it to use a combined output.
module-combine-sink - https://www.freedesktop.org/wiki/Softwa ... #index11h3
At least for me the output audio quality is than on the capabilities of the audio output itself (16 KHz Limit only for the MicArray).
So MicArray and 2nd Audio Device are receiving the audio output.
My 2nd Audio Device has not the 16 KHz Limit.
However the AEC is significally worse, but still working (maybe because my 2nd Audio Device is faster for the audio output in comparison with the ReSpeaker MicArray v2 audio output).