This bothers the Linux user

...

Other urls found in this thread:

src.infinitewave.ca/
levlaz.org/an-ode-to-linux-desktop-users-everywhere/
youtube.com/watch?v=qxt9ik2QKOQ
twitter.com/SFWRedditImages

Show how to do it under Linux.

Bonus points if you know what you're actually doing...

16-bit 44.1KHz is ideal.
24-bit 48KHz is pushing it.
Anything higher is audiophile-tier, placebo, and wasted CPU cycles.

Forgot to mention, pretty much 99.99% of the content you will play will never be higher than 24-bit 48KHz anyway, and 75% of that is at 16-bit 44.1KHz or lower. So anything higher is pretty damn useless.

24/96 is the max my hardware can do and I usually keep input set there,

Output I will vary but really the Apple audio engine is so easy on the hardware that I've been experimenting with 32 float/96 output and whatever magic rate conversion / dithering they're using, probably Dolby as they've been a Dolby licensee since forever.

You're missing my point though. Read the image title.

>32-bit float
It's useless if your hardware can only output 24-bit values.

>You're missing my point though. Read the image title.
So? Linux can do that too, it's just not as fancy. See pic.

LOL that's not your hardware dawg.

>audiophile delusions
>OS X in image
>bashes linux because of PEBCAK
10/10 troll post

either that or in disguise trying to figure out set a sampling rate in his ubuntu GUI-tier babby linux install

>24-bit is pushing it
16 bit noise is readily audible if you use software volume, and even more so because most 16-bit audio chips are really more like 14 bits in practice. (And of course, all 24-bit audio chips are really more like 19-20 bits in practice)

I wouldn't say 24-bit audio is pushing it in terms of driving your output (DAC). But 24-bit audio is pointless for music storage, of course.

To elaborate, you're just changing Pulse's internal bit rate / sample depth. Its onboard mixing, sample rate and bit depth, and dithering algorithms are, unfortunately, for shit.

What I'd like you to show me is how to set your output to 16 bit, 44.1 KHz, as if you're going to play back an image of a red book CD, and make sure your entire audio pipeline is sample accurate without dithering or shaping or sample rate and / or bit depth conversions.

Hint: if you're running a recent Linux you probably can't even do it.

You can't change your hardware audio format you retard, what you're changing there is software controlled.

>listening to music on your 'pooter
lol

>You're missing my point though. Read the image title.
wew lad, are you trying to insinuate that linux can't use your meme audiophile placebo sampling rates?

Lincucks unironically defend this.

What if you wanted to record at 16 bit, 44.1 kHz and simultaneously monitor your input through your headphones, sample-accurate?

You actually can't even do this with Linux...

>To elaborate, you're just changing Pulse's internal bit rate / sample depth. Its onboard mixing, sample rate and bit depth, and dithering algorithms are, unfortunately, for shit.
resample-method = soxr-vhq

PEBCAK to the max

in b4 your golden ears can hear the difference between SoX (widely regarded as one of the best resamplers available) and whatever generic garbage apple uses (in b4 speex lmao)

>and make sure your entire audio pipeline is sample accurate without dithering or shaping or sample rate and / or bit depth conversions.
Holy shit the level of audiophile placebo is real

Most modern audio hardware supports multiple sample rates and bit depths, and have onboard sample rate converters as well. OS X gives you a choice of which you wish to use (hardware or software). You just can't control this from (modern) Linux, even if you think you are, you probably aren't in reality.

>Most modern audio hardware supports multiple sample rates and bit depths, and have onboard sample rate converters as well. OS X gives you a choice of which you wish to use (hardware or software). You just can't control this from (modern) Linux, even if you think you are, you probably aren't in reality.
this is wrong

There is no ports available to get a 32bit digital signal to a dac. If your dac is built-in, and you have pure anolog transferred from a 32bit digital signal on your card, you have a machine from the future that doesn't exist. What those 32bit options are for is taking a higher resolution recording for mixing purposes, you will never in your lifetime actually listen to a crystal clear 32-bit digital signal.

What is ALSA?

You seem to think Pulse is "the" Linux audio system...

You sure told him freetard

WHY IS PULSEAUDIO SO SHIT

funny coming from a guy confusing the pulse settings with ALSA

>you have a machine from the future that doesn't exist
>what is DSD
wew lad, been in a cave on mars for a while?

Or you could just, unplug it.

well it's simply wrong

I don't know what else to say. ALSA can drive your device at any rate it supports. see You can see that it's currently operating in 48 kHz mode

You're the one who's confused if you think your little games with the pulse config files are doing anything hardware-wise...

>What is ALSA?
User-unfriendly once you have multiple DACs and ADCs

It's not. It allows individual per application and per stream volumes, system wide equalizing, network audio streaming, creation of virtual audio devices and so much more.

Look, I have an in-depth understanding of pulseaudio and ALSA; I'm familiar with the API of both, I've dug around in PulseAudio internals, and I've performed countless measurements of things like effective latency throughout the audio stack

The only reason you seem to think I'm confusing pulse with ALSA is because you set it up that way, by asking a loaded question full of insinuations about pulseaudio, and then suddenly moving the goalposts to ALSA when I told you how to fix your pulseaudio “problems”.

DSD only can stream up at 192khz 24bit dumbass.

>integrated camera
>unplug
its still not an excuse to not have a basic GUI device manager...

ALSA in fact only works in 16 bit, 48 KHz these days (on modern Linux), and decimates / dithers all audio with truly antique, poor quality algorithms.

Everything in and out of your sound card gets munged by this shitware user.

PEBCAK

I've used ALSA with multiple DACs and ADCs just fine, although I switched to pulseaudio because dmix is shit

>ALSA in fact only works in 16 bit, 48 KHz these days (on modern Linux),
well this is factually wrong

ALSA will operate at any sample spec and rate you open the device with. ALSA doesn't even do any resampling/dithering etc. user, it's bit-perfect.

Maybe you're confusing ALSA with PulseAudio?

>DSD only can stream up at 192khz 24bit dumbass.
I'm afraid you're incorrect user, what you're talking about are pcm renderings of dsd streams...

Write one if you think it's important.


>PEBCAK
>I've used ALSA with multiple DACs and ADCs just fine
I didn't say it doesn't work. I said it's not user-friendly.

And yet it still can't handle Bluetooth as good as Windows

wow look at all these awesome features... that are supported by literally every OS since at least 1995

>I didn't say it doesn't work. I said it's not user-friendly.
I found it perfectly user friendly. Writing an asoundrc to suit your needs is not exactly rocket science

If you thought ALSA is not user friendly, then you aren't the target user.

>ALSA will operate at any sample spec and rate you open the device with.
not true

your point?

>Write one if you think it's important
this kills linux desktop

Feel free to try it?

λ speaker-test -c 2 -r 96000


while this is running do a cat /proc/asound//pcm*/sub*/hw_params

It will tell you the exact specification the device is operating at

Alsa forces the output to match the sound file, which is technically the best because there is no forced resampling of an played file.
Also you can force it with "rate" setting in an ~/.asoundrc file.
Why would you use pulseaudio for professional recording is beyond me, especially since alsa has a mixer build in for years now, and since most audio software is probably written with alsa or jack in mind.

The best DSD has ever done is DXD which was 384khz 24bit, I can't find any references to it being able to handle 32bit.

>this kills linux desktop
you have the right to use windows
you also have the right to browse Sup Forums

How did you stumble your way onto a technology board again?

>especially since alsa has a mixer build in for years now
dmix is a complete pile of horseshit

That said, if you're doing pro audio you'd be using JACK either way

This doesn't prove anything user.

>dmix is a complete pile of horseshit
pic related

>This doesn't prove anything user.
It proves the hardware is running at the sample rate you request. Linux is physically sending it samples at that rate, over whatever interface they're communicating at.

I don't know what more you want. I would bet 100€ on the fact that you can't ABX any of these settings any way.

pure ALSA for comparison

PulseAudio for comparison (it does gaussian latency smoothing)

DSD's operating at 1-bit, 2.8224 MHz or even 1-bit 5.6448 MHz now user.

What alternate timeline are you fucking in?

...

goddamn Linux users are so easy to troll

all you need to do is spout blatant misinformation in the most uninspired way possible and you will get dozens of threads

cya nerds

>It proves the hardware is running at the sample rate you request.
>the hardware
you have no idea where rate conversion is even taking place

also you have no idea what program, or bit of hardware, is even doing rate conversion

linux BTFO

>you have no idea where rate conversion is even taking place
In your hardware (DAC) chip

ALSA is a bit-perfect API. Maybe you're confusing ALSA with dmix? Even so, nobody says you need to use dmix, and when you're using dmix you also get to set the sampling rate

I seriously don't know what more you could possibly want. But, ha ha, “Linux BTFO!!1”.

>In your hardware (DAC) chip
nope!
>ALSA is a bit-perfect API.
nope, not in any modern Linux! used to be...

he probably wants a gui menu with list of possible outputs that lets him choose the biggest numbers just "to be sure". He probably converts mp3 to flac for best audio quality too.

Okay, so you're just confusing ALSA with dmix or perhaps pulseaudio

Perhaps you should work on clearing up your misunderstandings first, user

>MP3 to FLAC
nah, he converts FLAC to WAV

after all, the bitrate is higher therefore it sounds better

>so you're just confusing ALSA with dmix or perhaps pulseaudio
Also wrong...

>hear the difference between SoX (widely regarded as one of the best tries at an open source resampler available) and whatever highly tuned excellent software apple uses (proprietary NSA-grade Sorenson magic)

Well you're more than welcome to read the source code to clear up any source of confusion?

In fact if you read the source code you'll see the point. ALSA's been crippled internally since the 3.x days.

Oh hey, we don't even have to guess; src.infinitewave.ca/ has measurements of Apple's resampler

come and watch apple's proprietary NSA-grade sorensen magic get BTFO'd by free software

[citation needed]

>(((measurements)))
oh user, you're falling for fake news again

yes, yes - don't trust the measurements; buy our magic audio pebbles instead, stupid goyim!

You can look at the many discussions online about this, the alsa mailing list is positively jammed with complaints and has been for years, for example, but jack users seem to have taken notice as well.

It's just not a suitable platform for audio any more, other than watching youtube videos and shit.

>You can look at the many discussions online about this, the alsa mailing list is positively jammed with complaints and has been for years, for example, but jack users seem to have taken notice as well.
Feel free to link one

Nah, I'd rather let you pretend my decision not to link anything is proof that your audio's sample accurate even though it's not and a quick search would show you this.

These days, there's no excuse for not knowing how to search for yourself, and in the process you're going to learn something.

Maybe this is your issue - you'd prefer to remain ignorant of the many issues Linux audio faces, and the dire state of open source audio solutions in general.

I gave it a quick google search and could not find anything relevant under "alsa not bit perfect" or "alsa mailing list bit perfect".

Burden of proof is on the one making insane accusations like that ALSA would somehow benefit from doing something other than simply passing on the samples to the hardware.

>Maybe this is your issue - you'd prefer to remain ignorant of the many issues Linux audio faces, and the dire state of open source audio solutions in general.
Well I'm not an audiophile so I honestly don't give a shit about “bit perfect” playback gold HDMI cable memes. I run PulseAudio because turns out mixing and volume control are important features

levlaz.org/an-ode-to-linux-desktop-users-everywhere/

Wtf

I knew that NSA purposely made it harder to disable cameras on Linux

it's literally one checkbox?

wtf i love linux now

>GNU/Linux
>"""technology"""
a bottle opener is more technologically advanced than that shit.

>work for free
Fuck off commie shitstain

you're right, Sup Forums should be a board dedicated exclusively to GPU wars and gaming battlestation dick measuring contests

>gave it a quick google search
see that's odd, my quick DDG search led directly to relevant results

>Gentoo
now I see your problem

I don't believe you. Wanna know why? Because DDG is shit at keyword search

or you're just shit at picking keywords

here, this can help you user

youtube.com/watch?v=qxt9ik2QKOQ

>dawg

Why are you lying?

I just checked and it's true, ALSA is hard-locked to 16/48 and dithers everything!

Look on the Ubuntu forums, there are tons of complaints there.

Actually I checked myself and it's locked at 32 float and 48KHz. It's 32 so that you have a lot of space to digitally lower the volume without reducing the bit rate bellow 16 (since most audio is recorded at 16)

>Ubuntu forums

>since most audio is recorded at 16
Actually, that is wrong. Most audio is _mastered_ at 16 bits.
16 bits isn't enough dynamic range for a lot of instruments.

Yes, you're correct. Thank you for the clarification.

>locked at 32 float
that seems to be the mixer and not your hardware user, it's doing at least some dithering and conversion

>16 bits isn't enough dynamic range for a lot of instruments.
16 bits is the difference between a mosquito somewhere in the room and a jackhammer at a foot away

what instruments exceed this dynamic range?

>not your hardware
Never said it was. Windows does the same thing

>mixer
Are you referring to dmix? You don't need to use dmix if you don't want to. Just use the hardware device directly

i.e.
pcm.!default {
type hw
pcm "ODACrevB"
}


ALSA will only do bitrate etc. conversions if you insert a conversion filter on top of this, e.g. ‘type plug’. Most distro's default asoundrc uses ‘plug’ out of the box, which is possibly why you're getting so confused about all this stuff

Keep in mind though that ‘plug’ will only insert conversions where necessary, i.e. if you try driving the device at a rate not supported natively by the device

>what instruments exceed this dynamic range?
I'm pretty sure drums need more than 16 bits. You want to be able to record the drummer absolutely hitting the crap out of it, but also be able to record the very quiet fade-out of the cymbals.

the OPs pic looks like a mac thing,

this is my audio mixer

again, jackhammer vs mosquito. I play the drums and no drumset I've ever encountered hits 120 dB

Even then, I'm sure most audio engineer make absolutely sure they don't clip their audio, and record at a much higher bit depth than they might actually need.

>audio engineer
audio engineers will*

>I play the drums and no drumset I've ever encountered hits 120 dB
No room or audio system's quiet enough for you to have had 120 dB headroom either user...