Stand-Alone Matrix Voice


#1

Hi, I’m getting confused in choosing a Matrix Voice on indiegogo.
Some basic questions:

  1. Does matrix voice have to be used with a raspberry pi? If yet, is it assumed that the $55 and $65 versions require obtaining a raspberry pi from some other source to function?
  2. The Wi/Fi BT functionality of the $65 version with ESP32, does it mean that it can be used without a Rpi? Why might we need a microcontroller on the voice if a raspberry pi is attached?
  3. Only the $105 version includes a power supply: does it mean I will have to buy a power supply independent of the purchase if I go for the cheaper versions?
  4. My use case: voice control in the living room for my smart home. I plan to do the processing on a powerful fast computer that is in a different room. What is the best set of items to buy to achieve this? If having a small computer in the living room would improve latency significantly, I might go for that.

Thank you!


#2

Hi @drexocom, how are you? So about your questions:

  1. Not necessarily, the power supply comes from the RPi connection but you can use a micro-usb power connector also. So you could use the MATRIX Voice without the Rpi. No, the $55 and $65 versions doesn’t come with the RPi.
  2. No, the ESP32 model it’s designed to be a stand alone wifi/microphone communication. As said before the MATRIX power configuration comes from the RPi connection (or the micro-usb).
  3. Yes, the cheaper versions doesn’t come with the RPi so you will have to buy the power supply for it.
  4. Humm, my opinion, i think the second option would be sufficiente for your use case, and you won’t need to buy anymore itens. But if you want to go to the first option using a computer in a different room, i asked my colleagues about it and you could stream the audio to your computer with a cloud round trip, but we don’t give that support.

You can look on our docs too and our github repository:

If you want anymore information just ask, have a nice day :slight_smile:


#3

Thanks for the reply!

One more question regarding having a computer to process voice be in a different room:
Why did you say the audio must go through a cloud? Can’t I just use the built-in wi-fi to stream in on my local network, without it leaving my house? I understand this would require advanced coding, etc, but does it really have to go through a cloud?


#4

Yes, you can use it your own built-in wifi. The cloud was just a option. But I think you can go without it. :slight_smile:


#5

I have a follow-up to the original question #2 regarding the standalone version with the onboard ESP32. Does having the ESP32 impact how the board interfaces with the Pi? My planned use is to connect the Voice to a Pi3 for use with Snips.ai, in which case I don’t need the ESP32, but I may want to use it for something else in the future. In that application, can the ESP32 be temporarily disabled (or simply not used)? I see some interest in this product on the Snips.ai and Mycroft forums and I’d like to not have to worry about whether a particular implementation will not work because of the presence of the ESP32.


#6

@dbenhart,
As you mentioned the ESP32 presence does not limit the functionalities of the MATRIX Voice. So, if you think you would need the ESP32 in the future feel confident on getting the ESP32 version.

Yoel.