Hacker News new | ask | show | jobs
by malfist 460 days ago
The report I read said it relied on bone conduction, so hearing protection wouldn't do a whole lot. Only things that can attenuate low frequency sounds before it gets to the ear. So muff style headphones might work, or mass
3 comments

It's not so much that LRAD relies on bone conduction to inflict pain (and sensorineural hearing loss!), but that the sound levels are so high that even if you block the air conduction route with earplugs, the bone conduction route (approximately 30 dB of attenuation compared with air) still might deliver enough sound to the inner ear to cause pain and hearing loss.

This kind of thing is a problem on aircraft carriers, where people working on the flight deck are so close to loud jets that no amount of conventional hearing protection will adequately conserve hearing. Creare has been working for the last decade and a half on special helmets for the US Navy to overcome this issue, resulting in the HGU-99/P Hearing Protection Helmet.

Ear muffs aren’t going to do much against 160dB
to put a finer point on it: decibels are logarithmic, e.g. a tenfold increase.

as in, an increase of 10 decibels corresponds to a tenfold increase in sound intensity.

this means 160dB is a lot. a loooooooooooot

Can you have a device which upon detecting the frequency emits some kind of counter vibration that cancels out th attack?
This is how active noise cancellation works in headphones. You stick little microphones on the outside of the headphones, then play back what’s picked up through the headphones themselves but with a very slight delay so all the peaks and troughs match up. The problem is that you need to put out sounds at least as loud, and that’s a pretty bad thing to get even slightly wrong if the energy levels are that high.
You do it with the waves inverted otherwise the delay would have to be dynamic and frequency dependent.
Not sure how all NC earpones work, but I'd say you still need the delay to properly process the sound and mix it with whatever is playing (unless you rawdog the input flipped and amplified directly in analog).

Normally sounds don't change in frequency that often so that's good enough (tm). I can hear myself typing now (short burst of sound), but the washing machine nearby, which is louder without earphones, is completely gone.

Sorry, you’re right. My understanding was that it was the same in practical terms to delay by half a wavelength, but that wasn’t correct.
It's not theoretically impossible but it is completely impractical to engineer such a thing - destructive interference has to be precisely matched to cancel out a sound, and if it's not you just get "beats" as the phases overlap.

And that match depends on matching frequency and distance - or having a very fast tuning system, and then you've got to do all this in a device that's not just another LRAD (at which point you're back to "the best defense is a good offense").