Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Higher quality audio makes people sound smarter

tips.ariyh.com

Post reply on HN
Print view View original

451–460 of 510 posts

Re: Higher quality audio makes people sound smarter

#451

Post by hippira »

nvarsj wrote: post #229
benkuhn wrote: post #13

It's also really easy to have high quality audio! The author recommends a "podcasting" microphone, but a $35 standalone headset mic[1] is almost as good and much easier to use. If you want to hear a comparison, I got kind of obsessed with this problem at one point and took some comparison recordings here[2]. (You need a standalone mic since most headsets, even really nice ones, have really bad mics because most heads…

The ModMic is also excellent, and you can attach it to existing headphones [1]. I use this at home with my prized Sennheisers. It baffles me that some people don't seem to care about their audio quality on calls. The most obnoxious are those who use speakers and you get echo on all your talking, and despite telling them, they still never bother to get a decent mic. Another common offender are the Bose QC35s: they hav…

ModMic(tm) is quite expensive.

Re: Higher quality audio makes people sound smarter

#452

Post by gwern »

azinman2 wrote: post #275

While the results seem believable, what was the sample size? The p values? None of this is cited either on the blog or the paper abstract, and I don’t have an account to be able to pull up the actual paper.

https://www.gwern.net/docs/design/2018-newman.pdf

Re: Higher quality audio makes people sound smarter

#453

Post by gwern »

Strilanc wrote: post #210

I looked at the paper's abstract [1], and this is a conclusion made from just two talks/interviews by the authors of the study? Should probably replicate it before telling millions of people to buy hundreds of dollars of audio equipment... I wasn't able to find an open access copy of the study, though I suppose there's always sci-hub. 1: https://journals.sagepub.com/doi/10.1177/1075547018759345

Yes, Sci-Hub works fine here. Since you apparently don't deign to use it, I've mirrored it here: https://www.gwern.net/docs/design/2018-newman.pdf

Re: Higher quality audio makes people sound smarter

#454

Post by u678u »

ZoomZoomZoom wrote: post #423
u678u wrote: post #359

Earlier quoted context omitted.

Count me in as someone that wanted not to use headset but couldn't face talking into a mic. I looked at shotgun mics but it got too expensive quickly and seemed a risk as it wasn't common. Anyone tried?

You probably don't want a shotgun mic, as a regular condenser set at a distance of 30-70cm will work just fine in most situations. When talking about condenser microphones for voice recording most people think about large-diaphragm condenser type, but actually it's absolutely possible to get great results with a small-diaphragm condenser with a tighter polar pattern ("pencil"). You can always play with the angles to…

Thanks. Most of the small condensers seemed to be designed for instruments which is why I thought shotgun would be better. I really could live without a 12 inch long mic so I'll take a look.

Re: Higher quality audio makes people sound smarter

#455

Post by Serow225 »

justincormack wrote: post #354
BurningFrog wrote: post #320

Earlier quoted context omitted.

I care, but not enough to ask people to QA my setup. I don't know of a way to check how I sound without bothering anyone.

Using the Zoom "record in the cloud" feature should roughly correspond to how people hear you BUT it does not let you know if eg your setup echoes someone else's voice. Bother someone, find a friend, ask your manager, geek out about audio, something.

There's a way to launch a "test meeting" where you can hear yourself as others would: https://zoom.us/test

Re: Higher quality audio makes people sound smarter

#456

Post by owenversteeg »

httpsterio wrote: post #449
owenversteeg wrote: post #422

Earlier quoted context omitted.

In my experience AirPods have excellent mics for what they are. They're definitely a million times better than the built in mic on the various (high-end) phones and laptops I've used in recent years. I wonder how they compare to a standalone mic or a decent headset mic (or that ModMic you mentioned.)

Airpods can't compete with a decently priced boom microphone that actually comes close to your mouth. The distance from your mouth to your ear (where the mic resides) is quite long especially considering how little space they have to throw in a capsule into. So, either you'll get a lot of ambient noise, the signal is quiet or Apple will do some algorithmic trickery that tries to approximate some kind of echo cancelat…

Hmm. I just measured the distance between my Airpods and the corner of my mouth at 3" or around 7.5cm. Very approximate measurement, but it seems to be not far off of recommendations for where to place headset boom mics (google says 1-3 inches from the mouth.)

Also, it's worth noting what the goal is here. The aim is not to capture the most accurate sound period. I've had calls with people who clearly have very expensive setups, but I end up hearing pen clicks, keyboard sounds, breathing, swallowing etc. The Airpods seem to do a great job of making my voice sound good in general. I've gotten compliments on my audio (so it can't be that bad) and the Airpods don't seem to pick up my breathing, typing, etc so I'm happy.

Re: Higher quality audio makes people sound smarter

#457

Post by jamiek88 »

noodlesUK wrote: post #435
KeepFlying wrote: post #326

Earlier quoted context omitted.

I assume it also depends on if they are using the laptop speakers or some standalone ones. I'm guessing the cancelation tech is tuned for the onboard speakers

This depends. On my (dell) laptop, the mic is basically right between the two speakers, below the lip of the laptop. It’s possibly the worst placement you could come up with for a microphone, because it barely picks up voice, and picks up all the typing, desk noises and speaker echo in the world. But I suppose that’s not surprising from the company that thought that a webcam beneath the laptop display would be a good…

I really thought they had some clever software or leasing to make the picture appear as if you were looking into it because of the placement but nope...just a nose cam.

Re: Higher quality audio makes people sound smarter

#458

Post by quadrifoliate »

FirstLvR wrote: post #147
nicbou wrote: post #138

Earlier quoted context omitted.

This was an excellent writeup. There's only one thing I would add: put the camera closer to where people's faces are. It feels like you're looking directly at them, and it makes a big difference. I made a habit of looking directly into the camera now.

the camera thing is really an issue in my company, we mostly work on software development so we share our screen constantly in our meetings... no one cares on turning on the camera and this has become regular behavior the problem is that you dont know if the other people are actually paying attention and human interactions need that feedback

> the problem is that you dont know if the other people are actually paying attention and human interactions need that feedback

This is going into the realm of the kind of monitoring software that tracks your eye movement to make sure you're concentrating.

If people are not paying attention to whatever you're presenting in your meeting, maybe the meeting is not relevant for those people. Consider cancelling it.

Re: Higher quality audio makes people sound smarter

#459

Post by alsetmusic »

benkuhn wrote: post #13

It's also really easy to have high quality audio! The author recommends a "podcasting" microphone, but a $35 standalone headset mic[1] is almost as good and much easier to use. If you want to hear a comparison, I got kind of obsessed with this problem at one point and took some comparison recordings here[2]. (You need a standalone mic since most headsets, even really nice ones, have really bad mics because most heads…

Step 1. Get a quality mic. Step 2. Control the sound in your environment. The best mic in the world won’t help if you sound like you’re recording in the middle of your kitchen.

One of the worst aspects of listening to a great interview is when the guest is in a space with tons of audio reflections. You want the sound of your voice, not the room.

Many podcast hosts climbed into closets with sound dampening clothes on hangers during the pandemic. It worked out reasonably well.

If you’re doing audio professionally, consider treating the recording space. If you don’t want to put panels on the walls, get free-standing panels that can be stored when not in use:

https://auralex.com/

Get a pop filter:

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

Control sibilance:

https://urm.academy/death-to-sibilance/

No use having a great mic if you don’t control the things you don’t want it to capture.

Re: Higher quality audio makes people sound smarter

#460

Post by Strilanc »

gwern wrote: post #453
Strilanc wrote: post #210

I looked at the paper's abstract [1], and this is a conclusion made from just two talks/interviews by the authors of the study? Should probably replicate it before telling millions of people to buy hundreds of dollars of audio equipment... I wasn't able to find an open access copy of the study, though I suppose there's always sci-hub. 1: https://journals.sagepub.com/doi/10.1177/1075547018759345

Yes, Sci-Hub works fine here. Since you apparently don't deign to use it, I've mirrored it here: https://www.gwern.net/docs/design/2018-newman.pdf

Oh no I love sci-hub but I second-guess linking it. Thanks for the mirror!

So, reading the paper proper, I was wrong about it being the authors' talks. They picked existing engineering talks and interviews. It'd be nice to be able to actually see the clips to get a sense of how much they changed the quality. They say they put the clips on youtube for the test but don't take the opportunity to link the clips.

In the Science Friday graph all the error bars overlap so meh. For the combined technical talks the error bars don't overlap and they're all in the same direction, which is way more encouraging, though I really wish they'd tried it with a pool of more talks to remove that systemic effect. I also worry about the fact that random mechanical turks may differ from the usual audience for a technical talk in important ways (e.g. not knowing the subject and so having nothing to go on besides audio quality). They say in a note at the end of the paper that they tried filtering by correctly answering a question at the end, and that the effect was still there, but they don't show the data for it or say if it affected the strength.

Overall I feel it's a reasonable methodology and certainly suggestive results, if a bit lacking in the exact methodological details and the number of talks they tried it on. And it's not like I have a prior that audio quality doesn't matter...

Post reply on HN