Ancestors

Written by Dan Luu on 2024-08-19 at 06:48

Another surprise was that getting a halfway decent pair of speakers that can emit bass (but aren't particularly bass heavy) made a lot of audio sound worse without EQ.

The problem is that a lot of audio sources contain a huge amount of low frequency garbage, which most speakers don't reproduce. If you get speakers that actually reproduce the garbage, that audio actually sounds much worse. I don't remember this being an issue the last time I had speakers with bass, but

=> More informations about this toot | More toots from danluu@mastodon.social

Toot

Written by Dan Luu on 2024-08-19 at 07:03

there's a lot more low quality audio now. 15 years ago, I was mostly listening to was music on CDs.

Nowadays, it's YT, podcasts, etc., which tend to have bad audio, even when people have fancy setups.

E.g., I just listened to a podcast that has a full-time pro audio crew, with hosts and guests in the same room, in front of SM7Bs. 1 of 3 people had strong proximity effect; sounds terrible, and that person is way too loud. You're better off with 0 bass when listening.

=> More informations about this toot | More toots from danluu@mastodon.social

Descendants

Written by Dan Luu on 2024-08-19 at 07:07

Ironically, I bought these speakers on the theory that it would make people easier to understand on calls. I don't think this was prima facie stupid since a $40 pair of speakers/headphones makes it much easier to understand people who have really bad audio setups (for video calls or w/e).

But it turns out that a more expensive pair of speakers, on average, makes speech harder to understand (though, on the rare occasion I listen to good audio, it sounds much better).

=> More informations about this toot | More toots from danluu@mastodon.social

Written by Dan Luu on 2024-08-19 at 07:16

Also, a few random comments about audio quality on calls:

1. With a few exceptions, the best audio I get is from people who do their calls outdoors without wind. Nearly everyone else is in a too-reflective room.

2. The worst audio tends to be people using non-mac laptop mics, then airpods, and then fancy mics in bad rooms without proper setup. The macbook mic is better than all of the above.

BTW, some people I know A/B tested macbook vs. airpods and the macbook won handily.

=> More informations about this toot | More toots from danluu@mastodon.social

Written by Adam Vartanian on 2024-08-19 at 09:09

@danluu Airpods (or any Bluetooth mic/speaker combo) has to share the limited Bluetooth bandwidth between input and output, so both of them end up pretty awful.

You can make this really obvious if you're listening to music through them and then start anything that uses the microphone, the audio quality drops drastically (and the mic only gets that differential amount of bandwidth).

=> More informations about this toot | More toots from flooey@mastodon.flooey.org

Written by Ben Zanin on 2024-08-19 at 17:21

@flooey @danluu the audio quality drop is because simultaneous input/output requires shifting from high quality codecs like Opus, the Apt-X family, or a couple of others to A2DP. Bandwidth contention does factor in, but there is a lot of headroom before it becomes the limiting factor.

Bluetooth LE Audio offers a different structure for simultaneous input/output that should remove this particular problem, even if the LC3 codec is not used, but it's not universal yet.

=> More informations about this toot | More toots from gnomon@mastodon.social

Written by psilotum on 2024-08-19 at 15:41

@danluu how can we evaluate our own setup? I occasionally ask coworkers how I sound if I've changed something. Even that is subject to their setup plus network variance.

=> More informations about this toot | More toots from psilotum@c.im

Written by Dan Luu on 2024-08-20 at 00:05

@psilotum For audio issues for calls, I feel like just recording audio locally and playing it back is generally good enough. If you don't have speakers that can emit low bass, just eyeballing a frequency plot should be enough. On Mac, Garage Band is bundled and works well enough for this. I'm sure there's free Windows/*nix software as well, but I haven't used any of it.

=> More informations about this toot | More toots from danluu@mastodon.social

Written by John Kaniarz on 2024-08-20 at 00:11

@danluu MacBooks have a dual microphone and sophisticated signal processing to “triangulate” your voice.

AirPods can only record from one at a time, and aren’t located in a good spot for doing so.

=> More informations about this toot | More toots from jkaniarz@mastodon.gamedev.place

Written by Robin on 2024-08-19 at 07:31

@danluu

I discovered that a while ago by accident. My cheap Bluetooth shower speaker died, and I replaced it with a different brand which was a fair bit bassier. Podcasts all of a sudden were a whole lot harder to understand in this white-noisy environment, especially if one presenter had quite a low voice.

=> More informations about this toot | More toots from eythian@teh.entar.net

Written by Chuck on 2024-08-19 at 16:54

@danluu ironic isn't it? As I age my hearing is non-uniformly deteriorating and that annoyance is exacerbated by bad audio setups. I use VB virtual audio cable to run things through a nice equalizer setup to improve things but I have not figured out a way to do common presets for specific sets (like use setup A for this podcast, setup B for this zoom meeting, Etc.)

=> More informations about this toot | More toots from ChuckMcManis@chaos.social

Written by skybrian on 2024-08-19 at 17:27

@danluu Yes, what you need to do for voice recognition versus music are quite different. Hearing aids don’t do anything with bass. I switch to airpods for listening to music, and turn down my hearing aids a lot for playing it.

=> More informations about this toot | More toots from skybrian@mastodon.social

Written by Laurence Tratt on 2024-08-19 at 09:24

@danluu There have been some podcast episodes I've tried listening to, that I'd probably find interesting, but the audio is so horrendous (e.g. clipping, massive differences speakers' levels) that I've had to give up...

=> More informations about this toot | More toots from ltratt@mastodon.social

Written by Adam Charron on 2024-08-19 at 13:48

@danluu people talking (especially podcasts) could really do with a high pass filter around 100hz. The human voice doesn’t produce anything useful below this and it will just be muddying up the sound.

=> More informations about this toot | More toots from Charrondev@mastodon.social

Written by James Scholes on 2024-08-19 at 19:23

@danluu So much this. I started listening to a podcast recently in a room with speakers plus sub, and had to turn it off because of the bassy microphone popping. On some lightweight headphones or MacBook speakers, it probably sounded "fine".

=> More informations about this toot | More toots from jscholes@dragonscave.space