Hacker News new | past | comments | ask | show | jobs | submit login
White noise improves learning by modulating activity in midbrain regions (2014) (nih.gov)
219 points by Friday_ on March 8, 2022 | hide | past | favorite | 103 comments



From Norman Doidge's book "The Brain That Changes Itself", I learned that:

Long term white noise exposure might be associated with hearing loss, especially age-related. The theory was that "nerves that fire together, wire together", too much white noise compared to clear sounds will cause the neural mapping for sounds to gradually become fuzzier, making distinctions between sounds will become harder.


"However, a recent study in the JAMA Journal of Otolaryngology–Head & Neck Surgery states that white noise’s lack of structure can worsen tinnitus symptoms. The study researchers also suggest that white noise could possibly “accelerate the ageing of the brain” and increase the risk for dementia"

Study: https://jamanetwork.com/journals/jamaotolaryngology/article-...


Anecdotal, but white noise sends my tinnitus wild. If the white noise is loud enough, you won't necessarily hear the change until it's turned off, but after that my ears are positively screaming.

Same is true of other kinds of background noise, such as computer fans, air conditioners or driving in a car. A while ago, based on some self study, I tried to 'fix' my tinnitus by targeting it with specific frequencies rather than noise, and that provided temporary relief by actually reducing the perception of the tinnitus briefly (like 5 minutes). But white noise has always increased it.


Hmmm, I wonder what implications that has for those of us who like to sleep with a fan.


I'm skeptical about generalizing those results to more "natural" sounds like fan noise. As others have mentioned true white noise is very unnatural due to its uniform power spectrum. Pink and brown noise though seem more natural due to the dampening at higher frequencies, and I'd bet things like fan noise, rainfall, waterfall, etc. lie somewhere along the 1/f^alpha for alpha > 1 (in between pink and brown).

I'm sure that we've evolved to handle things like noisy nights (thunder, rainfall) and living near rivers. You probably don't want to go beyond 50db though.

But to be on the safe side I'm going to turn by fan one notch down (40db -> 30db) to be on the safe side.


Ya, I would think there would have been some correlations drawn by now between hearing loss and living by the ocean or something.


Nature sounds like rainfall & ocean waves vary a lot I imagine, with lots of breaks and pauses, the lack of a continuous sound probably helps.

A fan that's rotating back and forth might produce neurologically better sounds, as the breeze bounces off different parts of the room than one that is fixed in one direction?


Actually hm I just tried a hearing test at https://hearingtest.online (very well done site, read the sidebar notes) and I seem to have a characteristic 4khz dip. I don't listen to music with headphones and have never gone to concerts, perhaps the years of fan noise at night did indeed contribute to this.


Or walking by the sea (though that's more akin to pink noise, I think)


Interesting given the studies that show noise pollution from roads etc lead to similar conclusions[1], even when taking air pollution into account.

Similar effect at hand?

[1]: https://www.theguardian.com/society/2021/sep/09/transport-no...


On the subject of hearing loss, can I plug my idea again that all devices should have a "constantly gradually decrease volume" option? It's way to easy to turn up a device for a temporary reason and then an hour later you realize the sound has been way too loud for a long time.

I would rather push a button to turn up the volume twice an hour than realize I've been damaging my hearing for the last hour for something I'm barely paying attention too. Especially for music, if I notice the volume decreasing, I can turn it up, and if I don't notice, then just let a fade to nothing since I'm not paying attention apparently.


I found myself particularly fond of this idea after reading your comment. Buttons840 I like that idea, you have my vote.


Must every article on HN about some positive discovery around noise, hearing, and tinnitus always be followed with a top comment about how everything we think about these things is wrong, and everything we’re doing is damaging our hearing and/or making our tinnitus worse?

I’ve been sleeping with Marpac sound machines for years, partly due to tinnitus but also partly just to help with sleep, only to find out it’s probably damaging my hearing and/or making my tinnitus worse.

Can someone please post some positive news about hearing/tinnitus?

This is seriously the second link in like 3 days.

Maybe I should just never click any HN article hearing related.


It can take a long time for conflicting theories to resolve in neuroscience. For example, the existence of adult neurogenesis in humans is still hotly debated despite at least a decade or two of papers on both sides. Meanwhile, it has already been accepted as a fact by the news media, and has begun to appear in self help books as evidence of the ability to 're-wire the brain'.

So, enjoy your sound machine. I'm sure bad sleep is probably implicated in hearing loss, too. Or at least it will be once the right grad student reads this comment.


"nerves that fire together, wire together" - ha, that's quite a profound statement. I've been thinking recently how we manage to hear harmonics as one sound without preexisting algorithms in our brain to detect them, and this is probably the answer. Edit. This also means we hear sounds differently. Someone who listens much to classical music hears harmonics really well, but someone who listens to sounds with different aharmonic structure, will learn to hear those aharmonic patterns as highly musical.


Harmonics or harmonies? Your edit is nonsense either way.


So there are 16,000 neurons in the ear that perform something like FFT with 16k bins over a 100ms sliding time frame. Now imagine there is a 2nd layer where each neuron is watching the corresponding neuron in the 1st layer and when that one fires, it looks up what other neurons are firing at the same time. This correlation event is recorded somewhere inside the neuron. After listening to lots of sounds, the 2nd layer neuron will notice a simple pattern: when its frequency fires, neurons corresponding to 2x, 3x, etc. frequencies also fire. As a result, the 2nd layer neuron fires upon seeing a harmonic pattern.


I wouldn't be so quick to call nonsense. Spending time listening to certain things most definitely affects your ability to pick out specific details.

All heavy metal music sounds about the same to me. But I'm sure that is not the case for fans of heavy metal.


Why does someone that listens to classical hear harmonics (I'm still not sure whether you mean tonal harmonies or harmonics aka timbre) better than someone who doesn't?

Sure, maybe they'd be better at picking out a bassoon in the mix than a metal fan, and a metal fan might be able to tell the difference between a Jackson guitar and a ESP guitar than the classical fan, but that's just due to familiarity, not their overall sense of hearing.

I don't think any serious musician would agree with you whether you're talking harmony or harmonics. However, listening to music in general will definitely allow you to hear different harmonic intervals better than if you didn't, but this is again just due to familiarity/ear training and doesn't overlap into your overall ability to hear.

We just use mnemonics. For example, Amen = 4th twinkle twinkle = 5th NBC = 6th etc.


Hearing is a process that extends from your ear all the way through your brain. The hard line you think exists between “hear” and “familiarity” or “ear training” doesn’t exist. The last few years I’ve been spending around Vietnamese, and while I still speak relatively little of it, I’ve been able to slowly adjust to be able to separate out sounds and words that previously sounded like one big mess.


And that's something completely different. Language and music use similar parts of the brain, but you're just mashing audio perception together in a way that just isn't related and is not what the article is talking about.


You’re being overly strict in defining a general word, in a way that worsens understanding.

We aren’t computers with parts. There isn’t a microphone connected to a DSP, there’s a neural net on every level from signal processing up to whole experience.

When someone says “hear” they mean almost all of that pipeline, not just the mic quality.

Picking out a bassoon in the mix is hearing better, at the level that matters.


White noise is insidious. There was this one team that sat in an area with lots of white noise. They seemed to like it - said it helped them focus. Now they all have hearing deficits.


Not necessarily a fan of white noise but I’d take white noise over unsubstantiated claims from a Popsci book any day. Those things are chokefull of un reveiwed conjecture that has no place in a book or anywhere else except the authors brain where all they should be is hypotheses.


I got into it because of tinnitus, and used it to sleep, but eventually started using it when working. Years ago in menlo park, with headphones on and white/brown noise drowning out the world. I likened it sensory deprivation, as nothing rises above the noise floor and it can fade away from your attention.


I found that earplugs + headphones with brown/grey noise absolutely made the universe go away. I don't use it now because my work has changed and I work in a single-person office, but I found this super effective at reducing distractions.


"we tested the hypothesis that auditory white noise, when presented during the encoding of scene images, enhances subsequent recognition memory performance and modulates activity within the dopaminergic midbrain"

Important highlight is that primary task (of remembering images) did not use (or need to use) auditory sense.


Andrew Huberman saying in a podcast that it's not the case for verbal recall:

https://youtu.be/Ze2pc6NwsHQ?t=3350

Interesting podcast, btw. got it from HN few days ago.


I got it from HN as well and it is excellent! Kudos for both Andrew and whoever commented a link to it :)


He's pretty good for sure, but he's a little too excitable on some topics and makes some leaps. I know he's done this on some of his exercise science episodes, for instance, with his discussion of cold therapy.


I've always been confused by people that enjoy white noise. It's awful.

Here is a sample from the wikipedia page on white noise :

https://commons.wikimedia.org/w/index.php?title=File%3AWhite...

Does anyone genuinely enjoy listening to that?

On the other hand, even though I don't, I could imagine why someone might like Brown noise (sounds almost like the ocean) :

https://commons.wikimedia.org/w/index.php?title=File%3ABrown...


I'd take white noise over distracting background noises (i.e., roommates, coffee shops) in a heartbeat. The point isn't to enjoy the sound per se, it's to block distractions.

But yes, I do prefer brown noise over white noise by a wide margin.


But then why choose such a horrible sound to drown out the background? Why not something easier on the ear? Personally I find the noise so jarring and intolerable I'd be more distracted by it (even on a low volume) than whatever irritating noises I'm hearing in the background.


Well that's your feeling and that's fine. But I'm okay with both white and brown noise, don't feel it's jarring. In fact, any other kind if noise distracts me


Not so much a fan of white noise either, but I like brown noise so much I wrote my own generator: https://justbrownnoise.com


This is really nice. It sounds a lot like a waterfall or whitewater. Very soothing.


"Like" is the wrong word. Silence is best, but if there are lots of auditory distractions in your environment then white noise drowns them out pretty well and just fades into a background hum for me.


The issue that I suffer from is that there is no such thing as silence. Even the sound of my breathing can be triggering, and the quietest of quiet keyboards will still set me off.


When I code I use white noise, or pink noise, but it's run through a low pass filter to trim off the harsh high ends.

I'll also add a slight comb filter then modulate the frequency with a slow LFO (around .05 hz) to give it some movement. Some extra processing then happens like a compressor to even it out a bit and add warmth.

It's quite effective, especially for noisy environments.


Do you have a sample you could provide?


At least for me, it's not listened to with the same volume you'd normally use for music or speech; white noise at that volume would definitely be awful. It's meant as barely perceptible background noise.


I use a non-looping white/brown noise machine to help me sleep at night and it works wonders in the city.

One thing I've wondered is the health effects of listening to that noise at ~ 40db all night over all these years. OSHA says that anything under ~ 80db is safe for 8h, but I do wonder if there are any longer term impacts to hearing or audio processing for softer sounds played over a lifetime. I could easily see the brain adapting and 'ignoring' noise in that spectrum. For now, I simply consider the extra sleep I get to be worth the price.


Interesting, could it be possible to elaborate which listening device you are using?


Not OP but I use one called LectroFan. They also have a small travel-sized one that doubles as a bluetooth speaker. It has revolutionized my ability to sleep while traveling.


I have the same make. I really like the lower range noise for blocking out stuff. I didn't notice any sort of bluthooth speaker ability. I might have to RTFM. That'd be useful.


Headline in 30 years: The long term effect of white noise caused a decline in IQ by 5%.

We would do better not to manipulate our brain when we do not know the mechanics.


Voluntary white noise is the bottom of the adverse environmental factors list I'm worried about.


When are we going to start crossing things off that list.


Possibly the same reason why people like to listen to music while working? The essential feature common to white noise and music you're familiar with could be the predictability. I know for me a good song (uptempo, not too distracting) can help me block out the outside world and make me feel very productive.


Music with words distracts me as I like to 'understand' songs and just enjoy the message.

I tried electronic music and as a European it really makes me want to dance, hah.

By far the best music for me is classical music. It literally makes tedious things like writing emails, documenting and coding much more enjoyable.


Another option is to listen to music that's in a different language. I have a playlist that I affectionately call "French Bistro".


I listen to the “radio” soundtrack from The Sims for the same reason. The songs have lyrics but they’re too indecipherable to be distracting.


I do the same with Scandavian music like Wadruna and then whatever youtube throws at me. No idea what they are saying but it's a nice melodic music that doesn't distract me and no worries about the song getting stuck in my head.


There are many genres of electronic music and only some are dance-oriented. If you want to check them out, try non-vocal playlists on https://www.di.fm/playlists or on YouTube or Spotify.


Same with me, the music I normally like distracts me while I work. To fix it I open up a youtube tab with a "cafe sounds" loop running (while also playing my music). Then it sounds like there's some nice music playing somewhere in the distance but I can't make it out. I also open up a "train sound" loop, which really dials in my focus.

It's a bit of a cacophony, but it got me through grad school!


The lofi hiphop (aka lofi girl) radio station on YouTube is another example of non-distracting music. It's there, but not something that draws attention.


Try ambient music.

Space music (a sub genre) is my favourite.


White noise: https://en.wikipedia.org/wiki/White_noise

https://news.ycombinator.com/item?id=28402424 :

> https://en.wikipedia.org/wiki/Neural_oscillation#Overview says brainwaves are 1-150 Hz? IIRC compassion is acheivable on a bass guitar.

Doodling improves memory retention / learning, too. IDK how much difference the content of a doodle makes? Hypothesis: Additional "cognitive landmarky" content in the doodle or received waveforms would increase retention up to a limit.


" IDK how much difference the content of a doodle makes?"

I'm a doodler. I also make art. And you are right, the cognitive load of the doodle makes a difference. My 'learning doodles' are really a mindless endeavor, and they only realistically look arty because I have been making art for decades now. The actual content of the doodle matters little. The real point is more keeping the hands busy and the mind lightly engaged with something else - I suppose it is akin to listening to a podcast while walking or driving.

I can get something similar from taking notes, but even that is better if I have doodle space since I don't need to write everything down.


Regarding doodling, it seems that any active productive study method is at least somewhat beneficial. I've found this in my own personal experiments. For example, testing myself with cloze deletions, creating Anki cards, generating mind maps, Cornell notes, inline annotation/marginalia, doodling, generating questions, generating mnemonics, mind palaces, etc.

An interesting one is reading and reciting out loud: https://www.tandfonline.com/doi/full/10.1080/09658211.2017.1...

Another non-intuitive method that is helping me a lot is pacing around my house slowly while I read. It goes to show that cognition is an embodied phenomenon. It's unintuitive when intelligence is viewed from the traditional split mind/body paradigm but just take a look at an image of our nervous systems. Those wires to and from our brain and guts wrap around every part of us.


Memory and retention in learning > Methods of improving memory and retention https://en.wikipedia.org/wiki/Memory_and_retention_in_learni...

Mnemosyne/Anki, (FreeMind, yEd, Gephi, AtomSpace as-moses, RDF bnodes, ONNX,) mind maps, RestructuredRext, MyST-Markdown, todo.txt (TaskWarrior,), StructuredProcrastination, ActivityWatch, Dogsheep, *-to-sqlite, awesome-quantified-self, https://github.com/woop/awesome-quantified-self

But metrics for actual memory retention? I've heard of TinCan xAPI w/ a LRS. nbgrader, Khan Academy exercises, OpenBadges to demonstrate proficiency,


Awesome. Thanks for the resources. Coincidentally, I stumbled on this research today on the effect of exercise + meditation on word recall: https://hpp.tbzmed.ac.ir/PDF/hpp-9-314.pdf


"Episodic memory enhancement versus impairment is determined by contextual similarity across events" (2021) https://www.pnas.org/doi/full/10.1073/pnas.2101509118

From "How Our Environment Affects What We Remember" https://neurosciencenews.com/environment-memory-20165/amp/


Makes sense, since Nikola Tesla said, "If you want to find the secrets of the universe, think in terms of energy, frequency and vibration".

Probably sound vibrations, create an orchestra which is pleasant for the mind, althogh this is pure speculation.

I would love top scientist to do more research in this area.


On a MacOS machine with Homebrew, run "brew install sox" and then use this to generate brown noise (or white or pink)

play -n synth brownnoise gain -25



This is genuinely creepy sounding. It might be less creepy if they didn't compress the time so much


I've used these sound machines for years to both sleep and work: https://yogasleep.com/collections/sound-machines/products/do...

For me personally, white-ish noise has always been a calming factor, which allows me to relax and thus concentrate, if I need to.


I prefer soft 8D audio for concentration/work-mode on alexa speaker.

If I'm anxious - more upbeat sounds and pop/rock songs like Linkin Park or Imagine Dragons, w/ headphones.

Closing your eyes is needed and just follow the music as it makes you feel a little dizzy. I never understood 'ASMR' until I found this music thing, really feels trippy. Helps a lot w/ my autism/ADHD.


There are several instances that I know about of machine learning / signal processing methods ( for example empirical mode decomposition) that purposely inject noise into the algorithm to improve accuracy / fidelity / independence. I'm sure that others here can provide better examples than I.


I prefer pink (and occasionally punk) noise


When I lay down for bed I play a track called Pink Noise for Sleeping on a loop and set a 3 hour alarm that shuts off the music when it runs out. https://music.apple.com/us/album/pink-noise-for-sleeping/111...

When I'm awake I play Lightning Bolt, Oozing Wound and Discordance Axis on repeat. I guess they're so loud they're almost like noise lol


brown noise for me. brown and grey, mixed.


Noise in general helps in several area in my life:

  - Working out, pump up music for motivation. 
  - Cleaning or chores, enjoyable music or podcasts for motivation or filling space.
  - Light, shallow work, lo-fi to upbeat chill without vocals for motivation and masking environmental noise.
  - Deep, concentrated work, white or pink noise for masking environmental noise. Depending on the task, I will opt for silence.
To further explain white or pink noise use benefits in my case, they provide subtle stimulation and are minor distractions if I concentrate on them. However, the utility gained from masking environmental noise exceeds the utility loss in the rare moments that I shift focus to the white or pink noise, so I think it's a net positive for me.


I find brain.fm or movie soundtracks work great for study, focus or deep work.

I've tried white noise and can, like others, attest that brown or pink are better.


I enjoy some noise, like fan, AC, but get auditory hallucinations from most ‘noise generators’ (apps). I start to hear voices, orchestras tuning etc. Very strange and distracting. I’m thinking it’s my mind trying to make sense of all the random frequencies? Any tips to counter this? ( I could record my fan and play that... in a loooooong loop ;)


This happens to me when listening to a white noise app / machine that has a short loop. You can buy machines that are explicitly non-looping and it's so much better.


Why would they loop? Creating white noise is one line of bash.


Most of these that I've encountered are webapps and they loop/crossfade a white noise sample rather than synthesize it. I don't think looping is the issue here though. I get the same voice-like auditory hallucinations even when playing with an analog white noise source in my analog synthesizers. A static square wave without any other sounds also causes them for me.


Most of these machines are 'multi-function' machines that play crickets / waterfalls, etc in addition to white noise. That sound is best handled by capturing a snippet and looping it over and over again. The same applies for white noise as well unfortunately.

Also it's usually a good idea to buy a little machine for this that uses like 5W vs running your computer all night.


It's crazy they loop, it just takes one transistor in avalanche mode to create white noise. Maybe I should mod my Bose Sleep buds that have ridiculously bad sound quality and short loops (and battery life).


It's easy electronically but if you're writing an app for a device/OS that doesn't implement that capability and expose it to apps you're left with either pseudorandom numbers or a sample. A very long losslessly compressed sample can be good but would require a large download, so in practice these apps generally have short, lossily compressed samples. These samples (and the alternative, pseudorandom numbers) sound absolutely awful. A human mind on the verge of sleep is an insanely powerful signal processing device and recogniser/producer/completer of patterns.

Raspberry Pis have hardware random number generators. I use one to generate smoothed brown noise for sleep. It's blissful.

I've been meaning to get round to solving this problem on phones, making an app that can, without a ridiculously large download, generate really high quality sleep noise. I have some ideas but it's not an easy problem.


> A human mind on the verge of sleep is an insanely powerful signal processing device and recogniser/producer/completer of patterns.

This resonates with my feelings. Hardware pseudo random noise is something i'd like to try, but I'm not sure my partner is so keen on it, how to deliver the sound in a discrete and non-intrusive way?


White noise is fine and all, but nothing beats a good old 10h Server-Room Humming video.


The Star Trek: TNG engine room ambiance is up there.


Came to hear to say that was one of my favorite noises :)


I play brown noise when I work. It really helps.

I just go “Alexa play brown noise”. It helps my brain filter out noises.

The only thing it doesn’t work for is barking dogs. For that I have to put on my headphones and play lofi on top of Alexa’s brown noise.


White noise is horrible to listen to. Brown noise is where it's at. Exponentially smoothed for the connoiseurs, get that soft rumble going.

White noise is an equal mix of all frequencies. It's what you get from radio static. It's hissy and nasty.

Brown noise is the sound of things being randomly bumped around (Brownian motion). It's what you get from waterfalls and thunder. It's smooth and delicious.


Not to be confused with the ‘brown note’. Uug

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


I like this site, that lets you tune it:

https://mynoise.net/NoiseMachines/greyNoiseGenerator.php

But yes, pink or brown noise is better than white.


Harsh Noise Wall is where it's at.


Merzbow is my preferred "focus time" noise


Kudos, I take it you're focusing on your good taste in music.


I'm a fan of pink noise.


Amusingly, I initially thought this was deep learning related, where if you augment your training set by adding white noise to each sample, then your trained NN will be more robust.


can anyone comment on binaural beats and the claim that they can be used to manipulate brain activity- such as promoting focus or relaxation?


This is the old scientific American article (from 1973) from where a lot of the claims originate, to my knowledge:

Gerald Oster, "Auditory beats in the brain", Scientific American, 229 (4): pages 94-102, October 1973.

https://www.scientificamerican.com/article/auditory-beats-in...

From what I understand, later controlled studies do not seem to back the claim that specific frequency differences lead the brain to an alpha state, or theta state etc. as claimed. But there may be potential in reducing anxiety:

R Padmanabhan 1, A J Hildreth, D Laws. "A prospective, randomised, controlled study examining binaural beat audio and pre-operative anxiety in patients undergoing general anaesthesia for day case surgery", Anaesthesia, 2005.

https://pubmed.ncbi.nlm.nih.gov/16115248/


Sure. Thinking that frequencies picked up by your ears has any sort of 1:1 correspondence or influence with the irrelevant overall firing frequency distribution of your brain is like thinking that adjusting the speed of your washing machine to an integer harmonic of 440hz helps keep your piano in tune.


> dopaminergic midbrain

Could white noise also help with addiction?


I don't think so.

Any music can stimulate the release of dopamine and I would argue that white noise would be on the lower end of that comparison. Plus, the amount of dopamine released from listening to music or white noise would be trivial compared to drugs, alcohol, binge eating, nicotine, video games, or binge-watching TV. And, the tricky thing about addictions is that they're easier to replace than to simply stop, so it's important to find a healthier replacement and then try to taper the newer behavior. I don't see white noise helping directly with addictions because of how imbalanced they are as stimulation sources.


Ok, but perhaps you could listen longer to it, e.g. while sleeping.


What's the effect size?


i find pink noise more relaxing than white noise when i want to focus.




Consider applying for YC's Summer 2025 batch! Applications are open till May 13

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: