Is it me, or is sound really great on FreeBSD?

[tedbell]

Rip your vinyl records in flac24 and play that on OSS, not sndio. That lossless form will play a fuller spectrum and higher bitrate of sound from the computer than CD quality. It requires a 24 bit capable sound card/processor to hear the benefits of those sound files. Maybe you have done this already. Flac16 is CD quality. The last time I looked, sndio didn't have 24 bit sound on FreeBSD.

I know. LOL I have been ripping vinyl for about 15 years.
I usually rip them in Audacity with 32-bit FP and then convert to flac 24/48. I don't use sndio at all. I can't explain it but OSS sounds better to my ears than ALSA and DirectSound on Windows.

 

[userxbw]

I just listen to sound on my head phones and yes sometimes I cannot get it to turn up loud enough, that little extra boost that pulseaudio gets. I know too that some of it has to do with recording level that the movie or music was recourded at plays into it, but still pulseaudio seems to give it the push it needs. pulseaudio was developed for Linux/GNU so ... I live with what comes with FreeBSD.

 

[PMc]

I also favor OSS. I use my FreeBSD box for running DAWs,

Oha! So there is professional use on FreeBSD!

From scanning thru forums of professional people, it appeared to me they mostly use Windows and OSX, as that's what one gets software for, and then there is some Linux in the more "nerdy" section. So, one wouldn't even want to ask if anybody knows FreeBSD...

I for my part am just curious what the sound stuff in the computer can do - up to now I only used it to get youtube etc. make some sound on the speakers, disregarding the details, while I considered HiFi etc. as an entirely different (and analog) matter (as it formerly was). I am not practically involved in professional audio, but -many years ago- we used to organize small music festivals, so I should still know enough about the professional aera to somehow understand the topics (and one can learn a lot from these people).

Pondering about getting myself a whole nice new HiFi system, it appears to be possible to do the things one formerly did with a little mixer and an open-reel tape-recorder, to do these now entirely within the computer - given the proper software and maybe some hardware with appropriate connectors.

Also, there's native support for equalization, whereas in other systems you have to do a lot of shit to get it working proper. The bass and treble controls even show up on audio/wmsmixer with the correct icons:

These appear to be presented by the audio device driver and appear here, from where any graphical frontend can access them:

$ /usr/sbin/mixer
Mixer bass     is currently set to  55:55
Mixer treble   is currently set to  55:55

And then there's the before mentioned sysctl tweak hw.snd.vpc_0db, which acts pretty much like a built-in pre-amp.

Yeah, but how do we get these into some frontend where we could adjust it? And how would it be possible to e.g. route some output to different/multiple soundcards? Except by adjusting the sysctl's and then restarting the software?

 

[shkhln]

And how would it be possible to e.g. route some output to different/multiple soundcards? Except by adjusting the sysctl's and then restarting the software?

https://forums.freebsd.org/threads/switching-dsp-devices-on-the-fly.69773/#post-419290

 

[dieselriot]

PMc I don't really consider myself to be a professional, but I get your point. I guess the reason I use FreeBSD is because I'm a jack of all trades, so I'm into computers as well as audio and electronics and lots of other stuff. What I mean is, I know a professional musical producer, and although he's fiddled with *nix, he's usually too busy with audio stuff all the time to setup a comfortable environment or fix whatever breaks; plus it's harder to get support from software developers when you're not using the platform their product was designed for. So, as you've pointed out, most of these people just use OSX instead.

Yeah, but how do we get these into some frontend where we could adjust it? And how would it be possible to e.g. route some output to different/multiple soundcards? Except by adjusting the sysctl's and then restarting the software?

I have this ugly hack for adjusting hw.snd.vpc_0db. I use something similar to adjust the screen backlight.

~/.xbindkeysrc

"doas sysctl hw.snd.vpc_0db=50"
  Control+Mod4 + 1

"doas sysctl hw.snd.vpc_0db=45"
  Control+Mod4 + 2

"doas sysctl hw.snd.vpc_0db=40"
  Control+Mod4 + 3

... [omitted] ...

"doas sysctl hw.snd.vpc_0db=5"
  Control+Mod4 + 0

 

[oops]

hw.snd.vpc_0db controls default per-application volume. Why not adjust volume individually?

$ sysctl hw.snd.verbose=2
$ mkdir -p ~/bin
$ fetch -o ~/bin/ https://people.freebsd.org/~ariff/utils/appsmixer
$ chmod +x ~/bin/appsmixer
$ appsmixer
wine (dsp0.vp0):
        Mixer pcm      is currently set to  45:45
mpv (dsp0.vp1):
        Mixer pcm      is currently set to  45:45
firefox (dsp0.vp2):
        Mixer pcm      is currently set to  45:45
$ appsmixer firefox 100
firefox (dsp0.vp0):
        Setting the mixer pcm from 45:45 to 100:100.

 

[PMc]

I have this ugly hack for adjusting hw.snd.vpc_0db. I use something similar to adjust the screen backlight.

Ok, thats cool.
But, if I understood this scheme correctly, dynamically changing that value is not the intended use-case - the value appears to be a "construction parameter", i.e. the technical point on the fader (0..100) where 0db should be.

The actual fader does appear with (something like) xmmix -dev /dev/dsp0.vpN. This creates a window with a stereo volume fader, possibly for each of the respective virtual inputs, before bringing them together. And in my understanding, then, the hw.snd.vpc_0db parameter should just deisgnate the preset 0db point for all of these.
The ugly thing there is that one needs a separate instance of the graphical mixer program for each of them. The audio/dsbmixer port mentions that they could be brought into a single, tabbed application, but then, that port seems neither bugfree nor complete.

oops : That's an approach. Thanks!

 

[joplass]

I also favor OSS. I use my FreeBSD box for running DAWs, mostly FL Studio. The Included FL ASIO audio plugin works amazingly well, no need for JACK. By increasing the buffer size a bit I can run high-end plugins on an underpowered laptop at ~60% CPU usage. I'm pretty sure they choked a lot in Linux.

Also, there's native support for equalization, whereas in other systems you have to do a lot of shit to get it working proper. The bass and treble controls even show up on audio/wmsmixer with the correct icons:

View attachment 6914View attachment 6916

And then there's the before mentioned sysctl tweak hw.snd.vpc_0db, which acts pretty much like a built-in pre-amp.

It's because of things like these or moused() that I haven't booted into my Linux partition in months.

I don't get bass and treble. Is there something I am missing?

 

[PMc]

I don't get bass and treble. Is there something I am missing?

Maybe. I have this in /boot/device.hints
hint.pcm.0.eq="1"
hint.pcm.1.eq="1"

 

[joplass]

PMc, I got more than what I bargained for. Thanks for that.

 

[Deleted member 67862]

I don't know if it's me or not, but damn... I feel like sound is so much better on FreeBSD than it was on Windows 10...

Better than Linux too! I can actually listen to my music without popping or any of that other CRAP..

OSS is an unbeatable sound system, swapping from it for the mess that ALSA is was one of the worst decisions the Linux devs ever made.

 

[neel]

In terms of sound quality, FreeBSD isn't nearly as good as Windows and Linux IMHO on a 14" HP Spectre.

But in terms of out-of-the-box support audio, FreeBSD beats even Windows 11 on the 14" TigerLake Spectre. I also previously owned the 13" TigerLake which was a different story however.

Getting an audio driver on Windows 11 was hard, even harder when I installed Windows 11 "Enterprise" instead of "Home" on a my Spectre in dual-boot meant I couldn't automatically download audio drivers. Mainstream Linux distros need the sof-firmware package, but not FreeBSD.

Note: I work at Microsoft (not on Windows), but the notes to support TigerLake probably went to the wrong person at MSFT, so FreeBSD supports hardware the Windows 11 ISO lacks .

Elementary OS is probably the best Linux I tested in terms of audio on my machine, but EOS despite is dated package-wise, and I don't want something with old packages (unless I am inside a VM or on a old PC).

Going back, Windows 11 doesn't support the TigerLake NVMe or touchpad out of the box but FreeBSD 13.0 from 6 months earlier does. Microsoft can manage to kick out KabyLake and 1st Gen Ryzen and non-TPM2/UEFI PCs on Windows 11 but can't figure out how to support TigerLake on the Windows 11 ISO.

OpenBSD is great in many regards, even Wi-Fi is supported but TigerLake NVMe may not work on all PCs, like Dell and HP won't work with OBSD, but Lenovo and Asus will after disabling "VMD".

 

[Voltaire]

I have installed FreeBSD on a custom desktop, an Acer laptop, and a packard bell netbook. On these three systems the sound quality was very similar out-of-the-box, an extremely high quality sound. I've also noticed that musicpd has very high audio quality in FreeBSD and you can also increase FreeBSD's out-of-the-box sound quality by adjusting several settings.

I used Quod Libet on Ubuntu on the same laptop, and I had Quod Libet configured in bitperfect mode on Ubuntu. But I noticed that the sound of Ubuntu in bitperfect mode sounded much more distorted than FreeBSD's out-of-the-box sound on the exact same hardware.

Arch Linux, in my experience, has better sound out-of-the-box than Ubuntu and Clear Linux, but also less detail and less correct frequencies than FreeBSD. windows10 in my experience is close to Arch Linux in terms of out-of-the-box sound quality, it also has a distorted and less detailed sound compared to FreeBSD. macOS is the only operating system whose out-of-the-box sound is almost exactly as good as FreeBSD.

 

[ProphetOfDoom]

This is just speculation as I certainly don't have super-accurate hearing or anything... but could there be a psychological component to this? In digital music production, engineers are always advised to turn off their computer's screen when analysing the frequency content of audio, because it's known that what you see can greatly influence what you hear.
It may be that because the Windows desktop is a terrible confusing mess, any music one hears whilst looking at it will sound like a terrible confusing mess.
Anyway AFAIK there's still no consensus on whether vinyl sounds better than a CD, or 96khz audio sounds better than 48khz. The arguments just rumble on.

 

[wolffnx]

my formula is...oss with audacious patched with eq with 31 bands ONLY flac flies , ... mixer vol and pcm to 80%
from there to the TV via hdmi from my nvidia card, and the output from the TV to my home theater via analog (the volume of the TV is set to 40%),
and finally my home theater

 

[chrbr]

This is just speculation as I certainly don't have super-accurate hearing or anything... but could there be a psychological component to this? In digital music production, engineers are always advised to turn off their computer's screen when analysing the frequency content of audio, because it's known that what you see can greatly influence what you hear.

This is really true. I a previous job I have done tuning algorithms which should cancel any noise from multipath, adjacent channel interference and so on for car radio applications. A big step in the process has been a tool I have written which showed the difference in the spectrogram of the original audio with the disturbed on. With that tool it has been much easier to identify issues. When you see the difference in the GUI you listen more carefully.

 

[Tieks]

Anyway AFAIK there's still no consensus on whether vinyl sounds better than a CD

I know a valid reason why people think vinyl sounds better, although a CD is technically superior. The re-releases on CD from earlier released vinyl records often sounded worse. Especially the ones that featured a SPARS code ADD. Here they took the original analog recording, then used digital equipment to (re)edit and (re)mix. I don't know what equipment that was, but the result always sounded different and seldom good. Many greatest hits compilations suffer from that same problem.
Disclaimer: I don't play vinyl records any more.

 

[Izlude]

I'm finally hearing sound on FreeBSD. I was stuck with front headphone for a while and the sound is a bit bland as it doesnt have that gimmick where there's a sabre dac mixed with realtek, its only on rear jack. On Win10, it relies on creative's app, but has a few settings that kinda lowers the volume if there's other conflicting sounds. technically not "normalize" or "sound leveling", they call it Smart volume. I can "kinda" mitigate it by turning it to off, but it still will do it in some instances. On FreeBSD, it doesn't do it at all. It sounds nice, clear and loud, to the point my speakers get expected distortion if its too loud. I like it this way.

 

[cracauer@]

None of the other OSes should have audible nonlinear distortion either.

Of course on Windows there is always the possibility that the "driver" package (which is usually lots of userland hacks on top of the actual drivers) do more processing by default - and might screw it up.


As Tieks mentioned, the reason why some of us have vinyl is that in some cases no good CD exists. The first CD was screwed up (ever listened to the white noise on Toto IV?) and subsequent CDs were remastered, aka changed from the original producer or artist vision.

 

[Alexander88207]

Another thing is that FreeBSD uses the best rate by default that got regonized from the sound output device.

In Windows this setting is a bit hidden and not on the best by default if i remember correctly.

 

[Voltaire]

This is just speculation as I certainly don't have super-accurate hearing or anything... but could there be a psychological component to this? In digital music production, engineers are always advised to turn off their computer's screen when analysing the frequency content of audio, because it's known that what you see can greatly influence what you hear.
It may be that because the Windows desktop is a terrible confusing mess, any music one hears whilst looking at it will sound like a terrible confusing mess.
Anyway AFAIK there's still no consensus on whether vinyl sounds better than a CD, or 96khz audio sounds better than 48khz. The arguments just rumble on.

It may be that Linux's audio is 'good enough' for you personally. You should also know that there are people who can have better hearing and/or pay more attention to details.
In my case I can say that the difference in sound between FreeBSD in bitperfect mode + real-time sound settings + musicpd with a high 'nice' priority is a night and day difference with e.g. the out of the box sound of Clear Linux.
If you do some historical research on this topic you find that Linux users have been complaining about extremely poor audio since at least 2010, and the Linux community has never been able to solve this problem.
Based on many sources I can say that Linux audio was never on par with FreeBSD's sound since 2010:

OSS4 on Ubuntu (Lucid Lynx)

OSS4 on Ubuntu (Lucid Lynx) // levien.zonnetjes.net //

levien.zonnetjes.net,

levien.zonnetjes.net

After upgrading to Ubuntu 10.4 LTS, I was happy to notice that audio in all applications (including Skype) was finally working perfectly! However, I was less happy to notice that Pulseaudio was using quite a lot of CPU-time, and that the sound quality was absolutely awful... So I decided to give OSS4 a try. After some googling, installing a few packages and some minor configuration, OSS4 was up and running, and I must admit the improvement in sound quality is rather significant!

Why do people dislike PulseAudio?

Reddit - Dive into anything

www.reddit.com

Why is audio still so awful on linux?

Reddit - Dive into anything

www.reddit.com

So fast forward to 2007, when PulseAudio is actually unleashed upon the computers of everyone else except Lennart and his friends as it's adopted and enabled by default in Fedora 8. To put it mildly, nothing worked anymore. Very literally -- when we installed it at the crufty place where I held a part-time job there, it broke sound on every single one of the 10-15 different configurations we had, from laptops to desktops.

Never mind that Alsa offered dmix even back then, but you had to know to enable ... | Hacker News

news.ycombinator.com

Why OSS sound quality is superior vs ALSA

Why OSS sound quality is superior vs ALSA - Open Sound System

OSS

OSS

linuxreviews.org

Open Sound System (OSS4) superior to ALSA

Open Sound System (OSS4) superior to ALSA

Open Sound System (OSS4) superior to ALSA

www.audiocircle.com

To give an example of what immediately strikes me: someone complains about the latency in Pipewire (Linux) in combination with Wine and Ableton Live. Guess my experience on FreeBSD with Ableton Live? It has extremely good audio quality and sounded exactly the same as on macOS. So I mean the latency on FreeBSD and wine was excellent. I had fewer audio problems with Ableton on FreeBSD via wine than on windows and macOS.

Other examples of people experiencing serious problems with the Linux audio stack:

Reddit - Dive into anything

www.reddit.com

Reddit - Dive into anything

www.reddit.com

Reddit - Dive into anything

www.reddit.com

Reddit - Dive into anything

www.reddit.com

Reddit - Dive into anything

www.reddit.com

Reddit - Dive into anything

www.reddit.com

 

[cracauer@]

Well, pulseaudio is a really low bar to compare to when it comes to predictability and reliability.

Real audio applications use jack both on Linux and on FreeBSD. The major obstacle is Chrome, which really likes pulseaudio.

There shouldn't be any sound differences on the same hardware between Linux and FreeBSD both using jackd. I certainly don't observe any.

 

[Alain De Vos]

I used jack before. But i had after each reboot "rewire" the jack config which was painfull. Maybe i missed something.

 

[Voltaire]

I used jack before. But i had after each reboot "rewire" the jack config which was painfull. Maybe i missed something.

I installed Void Linux on an old laptop and wanted to use PipeWire in combination with Jack and mpv to play movies on the TV via HDMI.
I've spent hours trying to get this to work, but I haven't been able to.

PipeWire + Jack + mpv + HDMI doesn't seem to work on Void Linux, although Void is known as possibly the highest quality Linux distro.

 

[cracauer@]

I used jack before. But i had after each reboot "rewire" the jack config which was painfull. Maybe i missed something.

I use a shellscript that does the required connections/disconnects.

You can also use jack_plumbing if you always want the same thing.