Matrix Products. Bad situation with Voice and Creator

#1

Hi Matrix guys! I am the one from some people who have a big expereince in mic arrays. I am engineer with patents contains my own mic of arrays, the methods for localization and some scientific articles. I am also the buisness partner for some startups which creates smart speakers.
My attention is for this one: https://newatlas.com/matrix-voice-raspberry-pi-voice-recognition/53034/
This is a repost that your Matrix Voice board is:
It’s 3.14 inches in diameter (that’s Pi to two decimal places of course) and has far-field voice capture, beamforming, acoustic source isolation, noise suppression, de-reverberation and acoustic echo cancellation among its feature set, too.
So after introduction to github I see that you have not supplied your board with smth like you described. It is not a far field because you didn’t create the reverb code so the max distance is near 3 meters during the lack of noise. In noisy rooms it is not working at least from 1 to 3 meters.
You code provide the ALSA only raw PCM samples.
I am partner of company which aquired at least 6 boards which are creator or voice. But we are disappotinted with that you lied using your own sources of news. It is not good.
You use the incorrect way of public informing causing the sales to be growth but your customers are maybe disappoitend and then give your products back.
I totally agree that now the Matrix projects is smth like a fool!
I have a big amount of another working products with dereverb, AEC, AGC, beamforming and localization algos. I made a test to compete between synaptics, intel, allwinner, microsemi and matrix. So after my partner check the results (which are bad) and then they give me a matrix boards to check why did they aquire the bad result. And it was a big news for me to understand that you, Matrix guys, just didn’t do nothing in acoustic algos. It is a simple data logger which is multichannel with pdm to pcm transform in FPGA. For this purpose the using of FPGA is a really funny decision. The PDM to PCM transfrom can be made in many ARM devices in software only. It use the DMA and it is enough to at least 80 Mhz of cpu like cortex M3. And it can be introduced in cpu like Cortex A using the DMA, the SPI interface and the correct C code in kernel module.
So my opinion is that you must aquire you boards back, return money and make a real product.
The another way is to give the community the proper code. If it will be not free than I agree it will be more good then nothing. The another incorrect way is that you didn’t support the hardware sources for the hardware product which now the Matrix is. So in such situation the Matrix as a product is just a waste of money and time.
And please, there is a fact that the Matrix voice and creator with Pulseaudio crashes the kernel module make the system hangs. I don’t understand how did you make a decision to sell the board which has a lot of strange engineers features (errors)

#2

Right, maybe you should contact them via their website and not post a rant on the community forum.

Please delete this silly post.

#3

No. My post will be here till then company give what the customers want and what do they really can do.
What do you mean about silly? My post contains a lot of info which can cause the customer to real understand about what the Matrix is. I am also get the letters from another customers which sent it me right today. So I think the problem is really huge!

#4

I find your post silly because you ask for something completely impossble (aquire all the board back & return money) and post a link to a website not of the company itself, but a news site.

If you had checked this site: https://www.matrix.one/products/voicem, you would have read something different.
So calling Matrix Labs liars while not checking the correct source is not a good way to start off in my opinion.
And after that you state that there seems to be a critical defect in the kernel modules without any further specification.
My Voice runs fine using the kernel, so without any furter comments, your “fact” is pointless

This is al beside the fact that the majority of users here on this forum are not advanced users as you say you are, but are using the creator/voice a a microphone array and use the ledring or other features not even remotely touching the issue you try to address.

There are open issues for sure, but your claim that Matrix Labs should take a the board back is really silly. Even more so because you only focus on one part, the microphones, but totally ignore are the other sensors and features like Z-wave / ledring / esp32 and all the rest.

So the “problem” you think is really huge, might actually be really small to almost non existant.

In my opinion that is.

1 Like
#5

Since you know FPGA and the algos, would you start an open source project to redo the FPGA firmware? That would be awesome.

#6

Hi Delphi,
I 100% agree with you.
far-field voice capture, beamforming, acoustic source isolation, noise suppression, de-reverberation and acoustic echo cancellation among its feature set, too.
This features were mentioned prominently in the campaign for the Matrix Voice, and is a major reason for my decision to buy it, yet I haven’t seen any of the examples demonstrating this functionality.
Unfortunately, Matrix Voice seems to be simply a lie.

#7

My plan was to improve my voice interaction project consisting of Raspberry Pi and a ReSpeaker 2-Mics Pi HAT with a powerfull microphone array with advanded features.

I was stupid enough to fall for the empty promises of matrix-voice advertised with the following features: Leverage fast audio processing and advanced algorithms for voice. Beamforming, AEC, de-reverberation and noise cancellation. Performance can be customized and improved over time by your own customization or by downloading our updates.

Statement from https://www.matrix.one/products/voice

The fact is that the Matrix-voice ($55.00) does not even come close to the performance of the ReSpeaker 2-Mics Pi HAT ($9.90)

You can’t even use the audio output of Matrix-Voice usefully, because it only support 44100 16 bit audio (unbelievable).

#8

Like I said, there are still open issues. And if you purely want to use only the Mic Array you might be better off with the other products indeed.
However, the fact that it is running on esp32, with a ledring to go with it makes it more than just a Mic array. You can leverage the pinouts as well, so the board itself is packed with features.
I am not using all the far field, beamforming and other features so to me this product is far from a bad choice.

The Respeaker can not run standalone, does not have a ledring (only 3 leds) and not even audio out without a Pi, so comparing a price between the two is not fair, you should add in the price for a Pi (zero) and a Neopixel ring.
The ReSpeaker Mic Array v2.0 has an audio out as well, but no esp32. and only 12 leds and a much higher price than the Pi Hat.

I agree is it strange that only 44100 can be used right now, but this is a firmware issue which can be solved.

So in my opinion, when you buy a Matrix Voice you get a nice dev board with a lot of nice features and it is certainly not a fraud.

1 Like
#9

Hello @delphi & @milode,

Thank you for your posts. We appreciate your feedback and concerns, and would like to take some time to address them.

Let us begin by saying that we are a development platform whose goal is to make it easier for anyone to create with hardware. Our engineers and community members continue to develop features for our MATRIX devices to elevate the platform to its full potential. Our method for choosing what to dedicate our engineering resources to has always revolved around the feedback we have received from the community. That is why after having met our goal for the MATRIX Voice Indiegogo campaign, we upgraded the board, at no additional cost to our backers, with a more powerful FPGA, audio outputs with a 3W audio amplifier, more RAM, and an 8th microphone.

We cannot deny that in our Indiegogo campaign we promised to develop algorithms for far-field voice capture, noise suppression, beam-forming, de-reverberation, acoustic source localization, and acoustic echo cancellation. These are all extremely complicated audio processing algorithms/techniques whose complexity has thus far greatly outweighed the requests for it. This is not to say that we do not want to or intend on developing them in the future. However, it has not been possible for us to develop those features while at the same time work on features that have been requested/used by an overwhelming portion of our community.

Regardless, we completely understand your frustration with this given that those are your primary features of interest, and we apologize for the inconvenience you have experienced. We have never tried to deceive our community, but we have had to carefully choose what we dedicate our limited resources to in order to best serve our community.

A big feature of our platform, which our advanced users have been very happy with, has been our commitment to open source. Up to this point we have open sourced every line of code we have written with the exception of the DSP algorithms within the FPGA. Over the last few months, we have been working on how best to make even the DSP available to enable anyone to freely reprogram the FPGA, and maintain microphone functionality. We will be making a community announcement about this later today, but in an effort to support our advanced users with audio & FPGA programming experience, we have decided to open up our DSP algorithms so that all of our code, including the FPGA Verilog, is open-source. With this, our community, and skilled engineers, such as yourselves, will have full autonomy to implement advanced audio processing on their devices.

You can check out the GitHub repos with our entire FPGA code (including the DSP) for the MATRIX Creator here and the MATRIX Voice here.

We hope this will open up some doors for you, and we once again apologize for your inconvenience.

Sincerely,
Team MATRIX

2 Likes
#10

I’m sorry, but I can’t agree with you.

(@Romkabouter if you haven’t figured it out yet, I’m talking about Matrix voice standard.)

On your current website (https://www.matrix.one/products/voice) the Matrix-voice is advertised as a finished product:

What’s the MATRIX Voice?

MATRIX Voice is a development board for building sound driven behaviors and interfaces. MATRIX Voice was built with a mission to give every maker , tinkerer, and developer around the world a complete , affordable, and user-friendly tool for simple to complex Internet of Things (IoT) voice app creation .


All that you say, has nothing to do how you advertice your „product“ which realy bears the expression „voice“ in its name!!!

You should really consider telling the truth to the people who plan to buy your great product so that they are not as disappointed as many other customers besides me.
Do you really have no idea how many members of your community have exactly the same problem?

#11

I’m glad you brought this up, give a heads up for people like me haven’t got that far. I bought the Creator still have a lot of work to do before even connect to it, and FPGA programming/algo is beyond my expertise.

However, I think Matrix Creator/Voice could be a pretty good platform to build those things up, especially after they opened the whole FPGA repo (maybe we should thank you for part of this. I found they already had the instructions posted on hackster.io in Feburary, but you definitely pushed them made the announcement yesterday) .

I hope you could do something on the platform, at least the hardware looks pretty attractive to me, I haven’t found anything comparable to the Creator in the market. Of course maybe you don’t think it’s a good idea to give your work free to promote the products from a company making profit, but again, none of those Linux or open source stuffs are running on Salvation Army’s hardware, at the end of the day, Intel or AMD or some ARM variations or whatever made by Apple or Foxconn or ASUS or some Chinese no brands, they are all commercial products, I mean, it’s OK, as long as you reach your own goals through the process.

I hope to see the progress after they opened the FPGA source code.

EDIT: I thought you were the author of the original post, sorry :slight_smile: I mean @delphi in the context