Your project is very creative

[lixy123]

Hello, I saw your software project and tried to transplant the voice function of the project to ESP32 hardware, but failed. The windfall is that ESP32 is connected to a microphone and turned into a remote microphone, which can be connected arbitrarily by other hardware such as raspberry pie. I hope to discuss video and voice projects together. This is the address of my project. Welcome to join us to help improve it. https://github.com/lixy123/ESP32_Remote_MIC

[paranerd]

Hey, glad you like the project. Yours sounds very interesting, too! Not sure if I can be of any help here, but let me know if you have any specific questions. Maybe I came across a solution during my development. Best regards. Am Di., 8. Okt. 2019 um 14:49 Uhr schrieb lixy123 <notifications@github.com

…

: Hello, I saw your software project and tried to transplant the voice function of the project to ESP32 hardware, but failed. The windfall is that ESP32 is connected to a microphone and turned into a remote microphone, which can be connected arbitrarily by other hardware such as raspberry pie. I hope to discuss video and voice projects together. This is the address of my project. Welcome to join us to help improve it. https://github.com/lixy123/ESP32_Remote_MIC — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1?email_source=notifications&email_token=AA2UWATXUPLM5UJJN7NLXRTQNR6XFA5CNFSM4I6RJSLKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HQKZQ7Q>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AA2UWAT4Y2WGS6TLYLSLEHDQNR6XFANCNFSM4I6RJSLA> .

[lixy123]

   Thank you for your reply. It costs about $4 to prepare an ESP32 with a camera and microphone in China. It's very small. It can achieve all the functions you can achieve with raspberry. It's a challenging project. I borrowed your main project code, read microphone information through ESP32, and transmit wav voice through HTML5. It's almost successful, but the voice are unclear and noisy. Yin, I'm ready to give up trying. I'm waiting for you to rescue this project.
   By the way, I have implemented a raspberry project of voice monitoring, which uses snowboy and voice recognition function , can transcribe the wav sound file into text and record, but the program structure is not as good as yours. Welcome to refer to it.    https://github.com/lixy123/Listener

[paranerd]

To be honest, I don't know anything about audio-/video processing apart from what you can find in my code.
I started this project, then went away from it for some months because I was stuck. Picked it back up, dug my way through, got it to work and moved away to pursue other projects.

I remember that synchronizing recorder and player has been a huge pain. If you already get output that is just unclear and noisy, it seems that you're almost there! Try playing around with some parameters on both ends, maybe you find the "lucky combination" ;-D

Sorry I couldn't be more helpful on this. Wish you all the best with your project!

[lixy123]

Okay, thank you for the helpful clues! I agree that you mentioned the sound frequency, the input of the number of voice digits is incorrect, need to call out the magic combination, thank you!