Matrix Voice Hardware Review


This post was flagged by the community and is temporarily hidden.

New User Here With Matrix Voice

You’ve removed the ability of my account to post feedback: this is a curious way to initiate a “dialog”.

I am only interested in open dialog. If you are willing, resume my account, and we can talk.


Wow, just wow. :zipper_mouth_face::zipper_mouth_face::zipper_mouth_face:

Not cool you were censored.


Hello Dominique,

This user was automatically muted by the community system for having multiple posts flagged by members of the community as spam and for cross-posting this blog post. They have since been manually un-muted by our team. We always appreciate feedback and are open to discussion.

Thank you,


Thanks for resuming my account. Point-by-point:

  1. Congratulations on the launch. I’m not debating any of your success. All my concerns were with the immediate device.

  2. I’m not certain any of your steps work with Matrix Voice. I retried your steps and hit the very same errors on the Matrix Voice board (SPI errors). What next?

  3. I don’t think any other backers have run successfully through the guide with a Matrix Voice. With a Matrix Creator, maybe.

  4. My background and work are irrelevant. I participated in Matrix Voice in an individual capacity. I’m waiting for your guide on how to get the Google Assistant (and others) working. A person proving your hardware doesn’t work with the provided software would probably refuse your offer of ‘collaboration’.

I was eager to have the device do something useful, but at this time I’m skeptical at best. I have spent many hours figuring out the software and reading your forums.

You should be honest about your backers’ experience. Even a cursory glance at the forums (and elsewhere) shows incredible frustration by people who wanted to support you. Debating me with a pitchfork doesn’t improve your product.

(For community members who have skill but are still are struggling with the board, send me your resume. :slight_smile: )


Glad to have you back.

I too feel your frustration. This product was marketed as something for “everybody”. However with the complicated process of setup and or usage, it most certainly is not. I bought 2 of these and to be quite honest, I dont ever see me using them. In fact, I am thinking of putting them on ebay.

I had big plans for these. I had a home automation project I was going to use the Matrix for dictation and speech rec (No, not for Alexa).

I am simply not happy.


Thanks Dominique, and sorry for your poor experience. This is a difficult product to get working, a point that you are making more convincingly than me.

In fact, my review shows the packages provided above are insufficient. The malos binary is a pure proxy for relaying FPGA calls over SPI. It did not reprogram the FPGA based on the source from October 3. The -hal-dev package is a collection of header files that nothing reads from. No amount of heroism from backers can get this collection of software working. Using the Matrix Voice requires information about how the hardware is set up, what the audio processing in the FPGA is, …

We fail because we don’t have most of the software we need.

For folks joining late, my review is at:



Yesterday I posted videos showing how to install and use MATRIX Voice and MATRIX Creator. @t0t0r0, do You try it? Do you have any question or doubt about it? Please let me know, How I could help you to you start your MATRIX Voice.


Kevin Patino


Hi @t0t0r0,

You can check the latest binaries for the FPGA in (these files are included in the last debian package):

About HAL project activity

For RPi

For ESP32

The upcomming HAL in kernel space


fwiw, i got all the samples i tried working. it took effort but they work. at the moment i have the one running on the ESP itself, the one which lights all the LEDs red and changes their intensity based on sound level being received.


There’s nothing wrong about this critical post, at least i can understand @t0t0r0 frustration.

I love the MC, a great piece of hardware … but i was frustrated too when responsible people write that there will be proper Pulseaudio / PyAudio support ‘soon’. Just search for those tags here.

There are two possibilities:

  1. Don’t use the word ‘soon’ in this context anymore.
  2. Or redefine the word ‘soon’ in all encyclopedias of this world and replace it by ‘longterm’ or similar :slight_smile:

Frustrated people might leave the community soon :wink:

No offense here, but i think that community management and response times could be better. I hope there will be a turnaround for the sake of all.