Hacker News Reader

Comment by carlmr | Hacker News Reader

How do you know the frequencies they're at? Did you Match it with sounds from your computer?

What constitutes high quality microphones for you? Where these microphones that were made for this low frequency range?

EDIT: Especially on the microphones maybe the actual noise isn't 28Hz, but the 28Hz are harmonics of a 14Hz, 7Hz, 3.5Hz,... vibration. These harmonics may get produced inside some structures near your ear, explaining why you hear it and others don't.

You'd need special equipment to measure these low frequency vibrations.

View on HN

Yes matching sounds on the computer. I can actually hear a beat when they are close but not exact.

Mics were a variety of xlr and usb from rode, sennheiser and yeti over the past few years. Definitely in the frequency range but only moderate sensitivity at that freq and low to moderate self noise. Not exactly scientific but I would expect to see a bump of energy in the spectrum and get nothing.

I'd be really intrigued whether you can find something at the harmonics using an infrasound microphone (some can measure down to 0.1Hz), but the price is probably prohibitive for an experiment.

That's the next step, just have some other projects to get out of the way first. There's a fairly low cost infrasound recording option using loudspeakers[1] that I think is worth a try. I'm also interested in building a seismograph, which might be able to pick it up on the high end of its range. If the sound is anywhere as loud as I hear it, there isn't much need for anything ultra-sensitive.

[1] https://www.youtube.com/watch?v=mekmsgjSqR4

I actually plan on recording sounds from my head, too. Could be weird spasms or something causing it, the lack of any variation in the two pitches makes that seem unlikely but possible.

Saw some interesting research on using infrasound to locate elephants. they managed to come up with something that was about $9 in parts including microphone. (Plus a SBC for analysis).

A bit of searching lead to some DIY stuff using subwoofers (speakers) as the microphone input then some low-pass filtering and then fed to an op-amp.

Here's the link to the research paper on the elephant detector https://www.researchgate.net/publication/320596116_Eloc_Loca...

Yeah you'd need PCB mics with reference level sensitivity. They cost over 800$ and the equipment that measures the noise costs in thousands.

They might be able to do it with an Earthworks M30 measurement mic, which claims accuracy down to 3 Hz, and an appropriate digitizer. It's expensive but not as blisteringly expensive as lab mics like Brüel & Kjær's gear.

I wonder if fans with (not a power of 2) blades could also generate surprising tonal (i.e., not transient) frequencies.

I could imagine this logic also applying to other pumps in the house: AC pumps, baseboard heat circulators, sump pumps, etc.

Reply on news.ycombinator.com