A Better Way to Cancel Noise | Hacker News Reader

A Better Way to Cancel Noise

nautil.us

80 pointsprostoalex55 comments

Here's an intro video [0] and here's the actual paper [1].

[0]: https://www.youtube.com/watch?v=VxcKkhmqXP0

[1]: https://synrg.csl.illinois.edu/papers/mute-sigcomm18.pdf

That aside, I found that construction earmuff is quite good for blocking sound and gain focus — with just about 5% price of top noise canceling headphone in the market. It took a bit getting used to though(and took quite some times to 'stretch' the earmuff in order to make it not clamp too tight to your head, but it works quite well and very cheap alternative.

It looks derpy though so I can't wear it in the office — where I want to use it the most. :(

You can sometimes pair a decent in-ear pair of headphones with shooting / construction earmuffs, depending on the clearances. That’s the best of both worlds IMO. Also, don’t worry about looking derpy; I give you blanket permission to not care :-)

You mean having construction muffs over earphones? Professional gamers have been doing that for years in 1v1 matches.

A great alternative to your setup is a set of plugfones [0].

They're an earbud that's an earplug first and a speaker second, which means they have an NRR rating and are generally pretty good earplugs.

[0] https://www.plugfones.com/

I've tried these for motorcycling and have found that the foam style is enormous and won't fit in my normal-sized ears :(

(They also have a baffled style that won't stay put while I put on my helmet but of course that won't affect you in an office)

It looks derpy though so I can't wear it in the office

You'll be fine. A guy in our office wears some, it's no big deal.

Heh, most people think mine are headphones. The derp factor only occurs once they figure out that they're not. :)

You can get earmuffs that are also headphones. I know Sennheiser collaborated with Peltor on some, a couple of years ago.

Wearing it in the office - if you can get away with it - would also be a good way to make a point that the office environment is too damn loud to think.

Don't ruin your hearing.

You can't listen to music in headphones all day without damaging your hearing.

The open pit coding fad is causing an epidemic of hearing damage.

the problem with these is that the tension on the spring (necessarily) is very high and they're painful to wear after a couple of hours.

I stretch it out quite a lot since I bought it. It's much better than the one that not stretched, though I admittedly have to take it off after a couple of hours like you said. My work focus is lost by then anyway, so it's a good time for a little break.

I found the tradeoff for the silence you got and its uncomfortable-ness is okay for its small pricetag though. :)

This sounds great - can you wear glasses with it?

I can, but it's certainly better not wearing glasses, because at the end of the day, it's an earmuff that designed to clamp your head quite tightly.

Also, the cushion on these earmuffs isn't the most comfortable cushion in the world, compared to high-end headphone. Well, it's built for construction worker after all...

The MUTE solution sounds interesting, and is certainly worth some research, but I see it having some problems. One obvious one is that the receiver might stay in that guy's window, and you might be in the apartment, say 15 feet away, so within that distance you can overcome the latency issue that was mentioned (because radio waves travel faster than sound waves) but the fact remains you're 15 feet away from the microphone in the receiver, and by the time the sound reaches you, it has changed dramatically. It has reflected off of 1000 surfaces by then and looks a lot less like it did at the receiver. So creating a waveform that's 180 degrees out of phase would be a gigantic challenge. I suppose you could do some work with Impulse Responses to solve this. You could take an IR of the room where the listener is and probably account for a lot of the changes and use some DSP to modify the waveform to make it more correct. But then if you move somewhere else in the room, everything changes and needs to be recalculated. Interesting idea though.

The solution is to use adaptive filtering.

The farther the microphone gets from your ears the harder the problem gets to solve. It's an interesting idea, but I suspect this won't see prime time in any meaningful way.

The description was rather vague, but seemingly equivalent to a pair of bose headphones with the microphone extended out a few feet in the direction of noise to allow more time to calculate the antiwave?

Yeah, except of course there are reflections from all over the place, so just extending it out in one direction wouldn't solve the problem. I think this is what the reference to IOT was all about — if you had a few sensors in the room, then you could do a better job of this.

I've wondered if a HomePod or two could do this — I understand they are currently able to discern a very quiet "hey Siri" (even when playing loud music), due to their spatial awareness. Of course, the HomePod would also need to know where the listener(s) is located in order to cancel the noise.

I live in NYC and I'm a longtime user of the Bose QC20s. They are probably the best commercial noise cancellation devices. I even sleep with them. The article promises some significant leaps in the technology. No doubt as processing gets faster and smaller we'll have even smarter noise cancellation. Looking forward to that...

Can't they stop human voice, music and children crying?

These three are literally the only things I want to filter out while I'm focusing.

Noise cancellation is designed to cut out the ambient noise - which is mostly white - of which there is a lot in a big city. It does this with a phase cancellation of what is probably a fairly predictable waveform. Human voices I can imagine are not that predictable. Maybe advances in machine learnings and processing may be able to provide better cancellation techniques.

> I even sleep with them

I used to do that with my QC15 first and then after it broke twice, with my QC25 replacement. Sleeping with them on is really tough on them, though. After the third replacement device, I decided I couldn't keep on doing it no matter how much I actually enjoyed putting it on at night and hear that soft static. The slight pressure actually felt pleasant.

Quitting using them for sleep was hard, though. I certainly felt the urge to buy a new pair but now I sleep with a pillow on my head if I'm sleeping on my side, or a puffier pillow put sideways under my head positioned in a way that it will cover my ears if I am sleeping on my back. It is no perfect replacement but I think the reason why I do that is because I am still seeking that QC20 feel.

They make a pair for sleeping with if what you want is white noise: https://www.google.com/search?q=bose+sleepbuds

It isn't exactly white noise that I am looking for. It is the residual waveform QC25 has after it has done its best effort at noise cancellation.

With that being said, after looking at here [0] it actually looks pretty promising although it seems to not have active noise cancellation.

[0] https://www.bose.com/en_us/products/wellness/noise_masking_s...

I'm assuming you've tried basic earplugs with white noise generator at full power? When I used them to block out building construction noise, they worked well. Maybe even too well - I was wondering if I'd wake up if a fire alarm went off. Active noise cancellation doesn't work well with noises like ambulance sirens.

Are you using earplugs then? I tried to sleep with headphones but it never worked for me and finding a comfortable but effective earplug is kinda hard.

I completely agree. My QC20s are probably hands down the best electronics purchase I have made. A gadget that has significantly increased my quality of life, and they last a long time.

I was on the fence about them. I had the Sennheiser BTNC 4.5 shortlisted since they're cheaper. Is Bose's noise canceling significantly better than Sennheiser's?

I can't rate Bose vs Sennheiser, but I can tell you that my $300 Bose QC35 II is far, far more effective than either my $100 Creative HN-900 or my SO's $100 Audiotechnica QuietPoint ATH-ANC7. I have a friend with both the Bose QC35 II and the $300 Sony WH-1000xM2 and he rates them about on par for noise cancelling, but puts the audio quality on the Sonys a little higher.

The two cheaper models are garbage by comparison, to the point that I'd say to just skip the noise cancelling altogether if it's only as good as the cheaper ones.

There is a new Sony coming out now they is supposed to be the best until Bose release their next I guess.

https://www.google.ch/search?q=sony+noise+cancelling+headpho...

Those Sony headphones had a defective design, the plastic would crack consistently in the same spot. You can get a sense by going through the products web page on Sony’s website

I have not used the Sony's, but the Bose QCs are absolutely fantastic. I think they a very safe bet.

Someone needs to invent an expanding foam that you spray in your ear and it solidifies in a few moments to form a perfect seal. It needs to somehow not go too deep and cause problems, and be easy to pull out and discard afterwards, ie, you can sell it forever like razor blades. Bonus points if it's biodegradable and can go in the compost.

Why would you want to take it out?

Perfect at work.

Perfect at the bar on the way home, to protect against the pub bore.

Perfect at home for the wife & babies, and a few years later ...

Perfect immunity from the acoustically-lethal menopausal wife & teens combo.

I have some of the best noise cancelling headphones (Sony MDR-1000X) and the thing few people realize is that active noise cancellation actually increases perceived pressure on your ear. For me it's like being underwater. I'm only really using them while flying.

I had both the MDR-1000XM2 and the Bose QC25, and the QC25 actually feel more like underpressure. I don’t think feeling overpressure is inherent to noise cancelling. I suggest you try out the QC25/35 in a store and see if they feel different to your ears.

I have the QC35 and it feels like overpressure to me. Might be an individual thing.

Yep, same headphones and I get the same sensation. It doesn’t bother me too much but when it’s not too noisy I leave the noise canceling turned off for that reason.

To me canceling headphones are only useful if I don't have a migraine. If I do have one and I try to use noise cancelling headphones, it feels like my head is spinning in circles and is about to blow up. Removing sound will for me always be better than combating noise.

It seems like this would require an array of microphones, plus knowledge of the user's location, mushed together into a model of the environment. Like MIMO for sound.

Completely agree. With only one microphone placed somewhere around the cubicle it would not be able to detect the direction the sound is coming from. In worst case the sound wave reach you before the microphone... Also in order to create the perfect anti wave the distance between you and the microphone has to be known really well. I think this is actually one of the huge advantages of wearable noise canceling headphones.

If this device truly did exist in the marketplace, I would pay a very good price for it.

i dream of cars with an array of speakers around the wheels to cancel the road noise.

The multi-channel case of using many anti-noise generators is much more complex than a simple single-channel.

I once proved a nice theorem about the multi-channel case, using Lagrangian multipliers and the theory of Gershgorin disks, which are what you get if someone upsets a Tower of Hanoi game all over your complex plane:

https://en.wikipedia.org/wiki/Gershgorin_circle_theorem

Assuming road noise is nowadays mostly low frequency noise - BOSE's car suspension [0] might do the trick: "It does for motion what noise-cancelling does for noise" [1].

[0] https://youtu.be/3KPYIaks1UY

[1] https://www.clearmotion.com

So... There's a concept, but no actual product. Got it.

They built a prototype to verify the idea. No product yet.

But a cool idea

I'm with the GP on this.

It isn't that novel either because of a failed product trying a similar concept:

https://www.youtube.com/watch?v=MCW5HUkrr-o

No, these two are very different. The one in your YouTube link is absolutely a scam, because no one can do "remote" noise cancellation without feedback signals around human ears. But this work makes very much sense to me (as an audio engineer).

I hate white noise and It bugs me a lot in current office settings (theory is that my brain is trying to make sense of what's thrown at it).

Isn't anti-noise signal, in itself, a noise? Whether or not it perfectly matches the noise signal or lags behind. Wouldn't that be more patterns that the brain would try to make sense of.

It would not be more signal.

https://en.m.wikipedia.org/wiki/Wave_interference

The goal of noise canceling is to fully use destructive wave interference to create a 0 amplitude "signal", which is no signal at all.

The point of antinoise is that the pattern is erased. The air is still.

Reply on news.ycombinator.com