Sound waves enhance deep sleep and memory | Hacker News Reader

Sound waves enhance deep sleep and memory

news.northwestern.edu

127 pointslxm46 comments

>13 participants 60 and older...

Oh we can just throw this shit right out. 13 participants? 60 and older in the US?

How about some controls for Diet, Exercise, and other general, you know, basic science shit? No? Well then allow me to just go ahead and call this study "statistically insignificant to the point of being potentially misleading" because I have pretty high expectations when it comes to rigor in academic study. This is what Northwestern chooses to highlight?

Point being, I grew up sleeping with Bach, John Jarvis, and Yngwie Malmsteen music for 20+ years. I think shoving a few cutesy-noise waves into the ears during rest is probably nice, like having a good pillow or decent sheets, but otherwise highly over-stated. This 13 person old fogey study means jack-shit in the big picture.

Want to have some real fun?

Let's get the Neural Network "Music" that Google is Researching to play during the sleep of 60 year olds and find out how that goes over. All these tech-approaches to very primal emotional pathways are disturbing to be on a general principle, mostly because not a single one of them has shown anything other than Placebo-esque results, and when placed against a control group, fail. Bonus hypothesis: The Google Machine Learning Music actually drives the Elderly Insane.

"Statistically insignificant" isn't a loose term you can throw around just because the sample size is small.

Here's the actual study, FYI: http://journal.frontiersin.org/article/10.3389/fnhum.2017.00...

13 people participated in two nights of sleep. I didn't see anything about a control group, or any control for the day preceding the night of sleep.

I tend to agree with this study, because of my personal experience with sound and its affect on my sleep, but I don't think this is anything but a reason to do a more in-depth study.

Ok. And your take is that in the original example they show sufficient arguments behind its statistical significance?

(honest question, not trying to sound snarky or anything.)

(PD: I meant "in the original paper".)

Also not apparently controlled-for is whether these people live in urban vs. rural (noisy vs. quiet) environments, and therefore can ordinarily expect sleep to be interrupted vs. not interrupted, by outside noise that the experiment's pink noise might be camouflaging.

People like you are the reason I invest my time on HN. Thank you!

A bit ironic how this is also a data point for why not to.

...as noted in a sentence ending in a preposition, which most pedantic editors of written English would regard as a habit not to emulate.

> Let's get the Neural Network "Music" that Google is Researching to play during the sleep of 60 year olds and find out how that goes over.

I'd be interested in using a feedback loop and an MRI scan (or whatever scan is needed) to selectively stimulate areas in the brain. I.e., computer generates music, looks at MRI, changes music accordingly, looks at MRI again, etc.

This bothers me:

>> Giovanni Santostasi, a study coauthor, developed an algorithm that delivers the sound during the rising portion of slow wave oscillations. This stimulation enhances synchronization of the neurons’ activity.

They did not actually measure "synchronization of the neurons". This is just speculation. In psychology they insist that you need to propose a mechanism for change. In this case it may be true, but they state it as fact without any actual evidence.

I suspect this is sloppy editing replacing a "may enhance" where Santostasi said "enhances". That kind of stuff happens all the time when scientists are interviewed.

My point is that no mechanism is even required if an effect can be documented. I let go of an object and it falls toward the ground - easy to replicate. I do more tests and find a constant acceleration - easy to replicate for dense objects (not feathers). And so on. For basic stuff I think showing an effect that can be replicated is way way more important than offering any explanation. IMHO even offering a tentative explanation gives people the impression that things are more understood than they really are.

In the summer months, I usually sleep with a fan (on a low setting) pointed at the bed. This is primarily to keep mosquitoes away (it works perfectly for obvious reasons), but the sound of it, I believe, really aids restful sleep and I often have it on even when I don't need it.

I have clinical insomnia and have relied on white noise from a fan for many years to sleep. On cold nights, I point it away from me, but always fall asleep more easily with it on the highest (loudest) setting. When I travel, I turn the bathroom fan on in hotels. The constant sound helps keep my thoughts from running rampant.

On the rare occasions when I can't do either of these things, I improvise with earbuds, but the effectiveness is diminished.

I also sleep with a fan on low, and yes, it does work.

One possible mechanism is the (well-attested) phenomenon of perceptual set. Ever buy a new refrigerator? For a week or so, you'll notice the sound every time it turns on. After that, you'll never notice it again. A similar thing is going on when you notice someone saying your name at a crowded party. Your brain "tunes out" the background noise from the multiple conversations, but when your name is mentioned, you're suddenly listening to that specific conversation.

At one time I lived near busy railroad tracks. For a week or two, I was getting blasted awake several times a night. After that? Didn't even hear 'em.

What may be going on is that the fan noise raises the background sound level enough that minor, unpredictable sounds (wind, bugs hitting the window, whatever) that would normally trigger alertness are enough below the ambient sound threshold that your brain doesn't pay any attention to them.

I've found audio books where the narrator has a repetitious rhythm that doesn't vary too much to also be as good as music or white noise, for the same reason.

Low frequencies are such a zen inducing thing. Same goes for planes (minus the AC). I even add a sample of my laptop fan through a headset mic, the sound was shitty but it felt like plane sound, and I played for the calming/focus effects

Really? I find that the noise bothers me. And even if I do manage to fall asleep with a fan on, it'll wake me up later at night and I'll have to get up to turn it off anyway.

In the past week, I have been using Mitch Altman's Brain Machine, which I bought from him during SHA2017 in the Netherlands, and it seems it has improveen my sleep (which was already quite good). It already has happened several times that I felt asleep during the stimulation process (blinking LEDs and binural beats on head phones).

Last week, I have felt much more relaxed in my thoughts, being able to concentrate for long periods without feeling the urge to visit social media or check my email. Maybe the fact that I returned to work from a frustrating vacation during which I made some important resolutions with respect to social media, also played some role. But being able to break some habits that I have been struggling with for decennia is nevertheless a significant event.

That's odd, I couldn't for the life of it fall asleep with a headphone near me. I bought one of the Mitch's machines(the neurodreamer) in the past, but I can't possible imagine how someone can fall asleep with a blue blinking LED in their face. Maybe I'm just a little to sensitive but I figured maybe something a little more subtle might make sense.

> I can't possible imagine how someone can fall asleep with a blue blinking LED in their face.

They appear to be red... Sounds pedantic but there is a big difference if we are talking about sleep.

For those looking for noise gen options, sox can be used to get a variety of fairly complex noise sounds. I tried a bunch of things based on an old blog post [0] with a bunch of helpful comments and the Reddit thread linked from there [1]. I like:

play -n -n --combine merge synth brownnoise synth pinknoise mix band -n 1200 1800 tremolo 50 10 tremolo 0.14 70 tremolo 0.2 50 flanger 4 2 95 50 fade h 2

There is also this clever Star Trek style noise (from the Reddit post):

play -n -c 2 synth whitenoise lowpass -1 120 lowpass -1 120 lowpass -1 120 gain +14 fade h 1

[0] http://unreasonable.org/white_noise_generator_with_sox_for_L...

[1] https://www.reddit.com/r/scifi/comments/n7q5x/want_to_preten...

Other people have complained about the number of participants, but that isn't the issue (it was run as a crossover study, so 13 participants should probably be fine).

The issue is that despite purporting to test fancy pink noise synchronized with brain activity, the control treatment was no noise rather than unsynchronized pink noise. That means that it's impossible to tell whether the synchronized noise was more effective than normal pink noise would have been (or even possibly just earplugs, although it looks like the earphones they used do a decent job of blocking noise: https://www.amazon.com/dp/B0046H8ZHS ).

Also worth noting: as someone with a serious sleep disorder (non-24 hour circadian rhythm), I've found that there are a lot of things that affect sleep quite strongly the first night and much less shortly after. The effect often disappears entirely quite quickly.

Here is the original article if anyone wants to take a look: Acoustic Enhancement of Sleep Slow Oscillations and Concomitant Memory Improvement in Older Adults: https://www.researchgate.net/publication/314295073_Acoustic_...

«enhance», the article won't say how much though.

the article won't say how much though.

Did you read TFA? "However, the average improvement was three times larger after pink-noise stimulation."

Three times larger than "a few percent". So, somewhere between three and six percent, I guess. Meaning that if the test set consisted of 20 items, they remembered on average one more.

And the study had 13 participants. They did two runs one with the snake-oil noise and one without. Can you even call it a study? With such a small numbers the study can show any random thing you want, from reducing feet odour to increasing yawn count. Even if it worked it's poorly designed because it assumes that the benefits fade away, ie. are not present on the 2nd day. Would be probably better to split them half and give stimulation then sham and for other half other way around (tip. don't split one senior in half!).

The study was done on 13 people for one night in the lab. It's an interesting result, no doubt, but to make any practical conclusions I'd wait for someone to replicate it for thousands of people after months or even years of at home use.

Interesting. From my own experiments a pair of soft earplugs and night blinders on eyes does the trick for me. Sometimes when I am under massive stress, adding a winter cap helps a lot to get wonderful quality sleep.

A coincidence for me seeing this article...I'm staying at an Airbnb next to a somewhat busy street, so I slept with a pink noise app on last night and I had some of the most vivid dreams I can remember

Hey me too! I used an app called Pzizz

I used "pink noise" on Android, with a logo somewhat like a human tongue. Super simple but it avoided a subtle clicking sound that repeated on a tone generator app I downloaded. I'll check out pzizz as well, thanks!

will this work with any pink noise apps ? or is this a new thing where "synchronization" that is needed?

The study had people wear electrodes to determine/drive the noise oscillation frequency. The noise oscillator was being controlled by the sleeping human (they used a PLL, which to me implies a bit of "catching up" lag if the human changed frequencies).

However, who's to say that the human has to be in control? You could replicate the study, but have the computer determine the oscillation frequency (or maybe just keep it fixed) and maybe find similar results.

The biggest technical challenge for a DIYer would be "comfortable enough to sleep in" electrodes.

Yes, it should work but make sure the app is genuine. Really though, even a table fan works.

What do you mean by genuine? Any recommendations on Android?

I don't know what he meant by legit, but some white noise videos on youtube feature jumpscares after a while. These are specially bad on people with cardiovascular issues and/or elderly. Always be careful before any kind of meditation-like video.

By legit I mean that they don't pass off white noise or some other noise as pink noise. Try and see what works for you. There is also the YouTube Music app you probably know of that you can use to play something overnight.

This reminded me of https://rain.today

for self test: https://www.youtube.com/watch?v=WJ9Go1PnAVA&list=RDWJ9Go1PnA...

what is the that lady using in the video?

If articles like this make it to HN... maybe it's time to start selling homeopathic snake oil...

I have such déjà vu about this article. The weirdest part is that it feels like the reason it is familiar is because I had a dream about it once.

Reply on news.ycombinator.com