Skip to content
HearingTracker
Log in

MEMS directional microphones with outstanding +67,5 SNR (yes, sixty seven)?

8 replies329 viewsLast reply

Link to this topic
Reply

#audioengineering #mems #microphones #immersiveaudio #smarthome #iot #chatgpt...5

Soundskrit kills the noise without killing the signal. Traditional mic arrays rely on heavy-handed beamforming that butchers sound quality with processing artifacts. These digital glitches don't just sound bad—they actively confuse translation...

Screen and copied picture:

image

image

FYI: hearing aids in or have two omnidirectional microphones. Directionality is achieved through digital analysing and processing based on time and spatial differences. I’m not sure what about form factor like Virto Infinio.

4 people in this conversation2 min readJump to latest

But are they gonna be able to make them small form factor, how small compact can they go I wonder.

So they fit two or more microphones into a single package and also do the signal processing on-board?

How else could this be done?

Notice that traditional MEMs microphones have a single opening in their metal cover. This one has many openings…

I know the guy responsible for the idea that made the current generation of MEMS devices work. There’s several trade offs between package size, mic sensitivity, mic noise, available voltage.

I had a closer look at their website. It seems that they achieve directionally by building it with a certain geometry.

When they say ‘configurable in several ways’, that likely means ‘configurable at manufacturing time’.

I think if this ever makes it into , it would have to be as an additional microphone. You don’t want super strong directionality all the time.

In reply to @Um_bongo

I still wait for energy efficient xMEMS receivers. Providing basics like broadband gain up to 20 kHz has potential to be quite a thing. We will see.

Even in 6-10 kHz range providing much gain, more than with balanced armature will be quite an improvement.

I’m not all over this, but the main issue wasn’t (iirc) the power consumption, but the mic floor noise due to the low mass of the silicone wafer diaphragm. The idea of using multiple sensors (electron pumps) on one or more surfaces with a comparative circuit to allow the removal of randomised (Brownian motion) noise is already in play to improve performance.

Putting more mic sensors in an array obviously allows for greater improvement through second order comparison for both time delay/direction and relative noise in the signal.

No, because air particles don’t make the receiver diaphragm bounce given that it’s much higher mass and attached to a drive rod/balanced armature reed. Also it wouldn’t matter if they did bounce it, the signal has already been amplified and the receiver is actively driving the output, the distortion effect would probably be well below measurable levels.

Related discussions