search

respeaker / usb_4_mic_array

ReSpeaker 4 Mic Array with builtin VAD, DOA, AEC, Beamforming & NS
https://www.seeedstudio.com/ReSpeaker-Mic-Array-v2.0-p-3053.html
Apache License 2.0
142 stars 66 forks source link

Extract voice eliminate all sound #38

Open msanzn opened 4 years ago

msanzn commented 4 years ago

I Have purchased the ReSpeaker 4 Mic Array v2 recently, I am trying to obtain the voice commands while having music playing, as in the video demo of DOA and AEC (https://youtu.be/XivM-6PRgX8). I have installed the latest firmware, with the tutorial in the wiki http://wiki.seeedstudio.com/ReSpeaker_Mic_Array_v2.0/ the DOA data is obtained perfectly but when I try to perform the AEC, but while I obtain the data from the different microphones the processed data is only the mix of all the microphones with a higher volume, with the music. What I want to do is eliminate the maximun music possible. I have tried putting the microphone near the speaker and also at a similar distance as from the user.

I am connecting the Respeaker to a Raspberry pi 4 and obtain the data both from audacity and from the python code in the tutorial.

KillingJacky commented 4 years ago

Hi @msanzn , please check how many channels in the output (of the firmware you updated), we have one kind firmware output 6 channels. If you're using the 6 channels output firmware, please observe the 1st channel as the 1st channel is the one outputting AECed stream. Refer to http://wiki.seeedstudio.com/ReSpeaker_Mic_Array_v2.0/#update-firmware