seeed studio ReSpeaker USB Mic Array for Voice Pick-up, Home/Office Automation, in-car Voice, Healthcare Device.

The speaker output is the only good thing about this. I have been trying to change settings and firmware for months and NOTHING makes a difference with mic volume. The incoming volume is way to low for anything useful unless its about 1ft or less away from the speaker.

Haven't played with hardware (yet), mostly used it out-of-box (except changing firmware to next-up gain, I think?). Worked straight away (shows up as regular USB audio device, used on MacOs and Linux/ChromeOS) and, compared to "regular" USB condenser boundary mic, it's night vs. day! Person speaking can stand pretty much *anywhere* in a15x15ft room, and ReSpeaker picks up loud and clear, with no feedback from speakers. Beamformer does distort vocal tone a little bit, but this isn't meant to be a studio mic anyway (at least not with "default" hardware). For conversation, it's fantastic, and I haven't even tried to optimize placement (e.g. it's close to wall on just one side, and about 1ft away from speaker, yet it doesn't seem bothered at all by any of that).In short, quite impressed so far!Edit: I should add that, for the echo cancellation component to work properly, it's important to *also* route audio *output* via ReSpeaker (and then route from the ReSpeaker's line-out to, eg, the final external speakers). Obviously, having that reference signal allows ReSpeaker's DSP to accurately estimate which part of the recorded audio is actually coming from the speakers.If that routing is not possible, then you can still use ReSpeaker as an excellent beamformer, but consider running echo cancellation elsewhere--which may or may not work as well. Eg, in my experience, Google Meet does generally fine, but Zoom totally sucks for some reason (perhaps because I didn't *actually* turn off the echo canceller on the ReSpeaker, and Zoom's echo canceller can't deal with the lag -- obviously, any echo canceller introduces a bit of lag, however tiny, which is also one of the reasons you generally shouldn't have more than one in your audio chain). Anyway, a starting guess (in case it helps someone), but I didn't investigate further because.. I didn't need to for my setup. :)

I am currently working on an AI home automation project (basically JARVIS), and I needed something that could detect my voice at long distances similarly to Alexa and Google Home. I tried a few microphones, and they were not cutting it. This seemed a bit expensive, but I took the plunge. I am happy I did.Word of warning, I was not able to get the array to work on Ubuntu. However, after installing another Linux distro (Manjaro), I got it to show up as an audio device. I walked across my apartment living room, and it picks what I say perfectly! The Speech To Text engine I'm using is able to understand me as well. Don't waste your money for room-wide voice recognition on just a conference microphone like I did. This is sleeker and works much better.

Mic array works well, but it doesn't function as a speaker. Unknown devices in Windows 10. Contacted tech support at the company, and they wouldn't give me any, so I returned it. If you won't support your products, then neither will I.

Initially when I set this up I had some sort of weird echo when I talked. My theory, since I was using headphones, is that it was picking up echoes of my voice and the auto gain was amplifying the echoes.I installed the latest firmware, which was quite easy, and after that it seems to be working great! I haven't tried adjusting any of the tuning so far. I've tried standing 20 feet away in another room, with a doorway between me and the mic, and it still sounds quite good.I also installed NoiseTorch to get rid of typing noise, and I have a set of speakers I connect to the ReSpeaker if I don't want to use my headphones.