• This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn more.

Using sndio in place of pulseaudio

free-and-bsd

Aspiring Daemon

Thanks: 66
Messages: 633

#27
Yes, it has a couple of advantages:

- It works, even when software drops OSS-Support
- License
- It has network features
- It doesn't mess up with multiple audio outputs

I am confused about your audio working though, as in there being no "go without" option if that works. As far as I know that should have stopped working recently, but maybe I am mistaken.
Nobody has explained to me yet how this is possible, but the fact is: I have www/firefox compiled with PA support (no ALSA), but I have neither audio/pulseaudio nor alsa-libs, nor pulseaudio-libs installed. Cleaned them all up for the sake of experiment a year ago, because I had problems making it work WITH audio/pulseaudio installed. Which I have posted here, too...
 

abishai

Aspiring Daemon

Thanks: 122
Messages: 511

#30
Patch is supplied to the port, so package already has it. That's why Firefox can work with OSS directly.

As there is no WebRTC solution [yet], port doesn't have OSS option officially.
 

free-and-bsd

Aspiring Daemon

Thanks: 66
Messages: 633

#32
tobiam OK, here is the news with newly built SNDIO:
Code:
[1/1] Installing sndio-1.2.0_2...
===> Creating groups.
Using existing group '_sndio'.
===> Creating users
Creating user '_sndio' with uid '702'.
pw: user '_sndio' disappeared during update
pkg: PRE-INSTALL script failed
EDIT: running vipw without changes to /etc/passwd helped.
 

free-and-bsd

Aspiring Daemon

Thanks: 66
Messages: 633

#33
Well, here's another funny thing about www/firefox with SNDIO option enabled. When I open a YouTube video there appears a second instance of firefox:
Code:
 PID USERNAME    THR PRI NICE   SIZE    RES STATE   C   TIME    WCPU COMMAND
55505 user       32  29    0  1733M   303M select  1   0:15   8.18% firefox
55477 user       51  20    0  1748M   270M select  0   0:07   3.15% firefox
58502 user        1  21    0 12442M 36840K select  1   0:17   1.86% Xorg
85679 user       13  52    0   574M 94324K sigwai  0   1:13   0.29% vlc
58522 user        1  20    0   115M 10816K select  0   0:04   0.09% urxvtd
I'm better satisfied with PA enabled version with only 1 instance running.
 

tobiam

Member

Thanks: 21
Messages: 88

#35
OK, looks like this last version (just updated) runs 2 instances regardless of whether it's SNDIO or PA.

Is it because of this?
Yes, that's most likely it. Press A in top to see the whole command. It's usually giving you a hint on what it is used for. If you somehow use Flash instead of HTML5 or any kind of plugins, those usually are also in a separate process. Should be pretty independent of sound related stuff.
 

free-and-bsd

Aspiring Daemon

Thanks: 66
Messages: 633

#36
Yes, that's most likely it. Press A in top to see the whole command. It's usually giving you a hint on what it is used for. If you somehow use Flash instead of HTML5 or any kind of plugins, those usually are also in a separate process. Should be pretty independent of sound related stuff.
Well I gave up on Flash plugin a while ago, and firefox does HTML5 quite well.

Is this here a good place to ask you a couple of questions about how you handle the sndiod? I mean, the man page is there, but I could find no HOTWO-like documentaion on sndio.

The reason why I'm asking is, for example, that I tried firefox with sndiod started, and the sound in a YouTube video was much lower than when sndiod was not started. Yet I could find no way to manipulate the output volume, aumix being no much help there.
 

tobiam

Member

Thanks: 21
Messages: 88

#37
Depending on what you want to achieve the -v flag might be your friend. :)

Sorry, I don't really know of any bigger tutorials and just rely on the man page.
 

free-and-bsd

Aspiring Daemon

Thanks: 66
Messages: 633

#38
Depending on what you want to achieve the -v flag might be your friend. :)

Sorry, I don't really know of any bigger tutorials and just rely on the man page.
Well, I don't quite get it from the man page how I can manipulate output volume for, say, HTML5 video in firefox. And there IS such need, because when I play it in firefox with sndiod started, the sound it certainly lower. Then there must be ways to adjust that, I figure.

Here is the OpenBSD manual. But it is based on the OpenBSD sound tools, which aren't ported to FreeBSD.
 

laurentis

New Member

Thanks: 7
Messages: 17

#40
Well, I don't quite get it from the man page how I can manipulate output volume for, say, HTML5 video in firefox. And there IS such need, because when I play it in firefox with sndiod started, the sound it certainly lower. Then there must be ways to adjust that, I figure.
There's a utility in ports to do just that: audio/aucatctl

In my case the master and per-application volume were already at maximum, so it didn't help.

However taking a hint from here, I figured that you could increase the volume by lowering the 0db relative level.

On my laptop, it's done with this sysctl:
sysctl hw.snd.vpc_0db = 10
 

antic

New Member


Messages: 1

#41
Why do we need sndio at all? OSS just works.
i assume the pkg firefox binary was complied with alsa and pulse set to "off"? i noticed the binary still uses them in the build but i don't seem to have any traces of alsa or pulse and OSS defaulted to handling firefox's sound output.
 

poorandunlucky

Well-Known Member

Thanks: 26
Messages: 359

#44
That's... informative. It shows me you know your stuff, and I feel compelled to ask: may I board the bandwagon of which you appear to be the driver, good sir? You strike me as a gentleman, and a scholar, and I can hardly contain my burning desire to, myself, bicker about the shortcomings of PulseAudio... which are, again? It's just The excitement, you understand... I'm not sure I would be able to phrase it just the same as you would! = D I wouldn't want to tarnish the graceful proverbial vehicle of our opinion, as it may, and so if you could help me put together the words so that I, too, may, perhaps, with time, a bit of luck, and your oversight, of course, milord, gain the favors and advantages of others support towards our blessed cause against the evils and shadows PulseAudio hast cast upon our kingdom! My leash, I implore you to radiate upon us the warmth of your wisdom and knowledge *bows*
 
Thanks: OJ

vermaden

Son of Beastie

Thanks: 959
Messages: 2,629

#45
@ poorandunlucky

I wanted to omit that as I probably stated that many times, and I consider it a waste of time, especially because its very 'negative'.

While FreeBSD has in kernel mixing of channels (up to 256 if I recall correctly) with OSS4, with low latency, in kernel space, Linux on the other hand some time ago BEFORE PulseAudio had ONE OSS channel (which often was blocked by WINE) and many ALSA channels. That worked until some app 'blocked' the OSS channel and then another app wanted to use OSS channel. In other words, a typical Linux mess.

Then PulseAudio from the great Lennart Poettering happened and it happened hard. I used Ubuntu for about a year because FreeBSD was not up to the task of supporting Intel G965 chipset with X3000 GMA Intel graphics adapter (only deadly slow vesa was possible), so I moved 'temporary' to Ubuntu to wait for its support in the FreeBSD land. Why Ubuntu and not CentOS? or why not Gentoo? I used Slackware and Gentoo in the past and these two bring me to the FreeBSD land, so I wanted to check a 'desktop' distro for the first time, and as Ubuntu was the most popular at the time, I decided to jump in.

This Ubuntu came with PulseAudio, not with the first version, PulseAudio was available for quite some time before that Ubuntu release. I did not used any 'strange' audio setup, just output to 2 speakers via typical mini jack output. As this was desktop I did not powered it off as I did not have such need, but PulseAudio forced me to periodic reboots. The whole audio on that Ubuntu Linux hanged hard after about a day of typical desktop usage, sometimes after two days, sometimes after half a day. No reload of services or even kernel modules (alsa) helped to gain the sound back, I was forced to hard reboot because of that PulseAudio shit. The strange thing is that the whole rest of the system just worked, I could do all other operations without a problem, but sound stopped with some jerky noice and thats it. As I used this Ubuntu for about a year I regularly applied updates with hope to fix this issue, but after more then a year it was not fixed. I came back to FreeBSD then and never looked back, but as I see more and more projects (Firefox/Chrome) force this shit (PulseAudio) into the FreeBSD ecosystem it drives mu nuts.

My opinion may be biased on 'this single system' but its not only my opinion. I am very disappointed that You did not seek for that yourself ... as everything that this moron Poettering does is just pure shit.

The PulseAudio basically did what Microsoft did with Windows Vista in audio subsystem, they moved it altogether into user space, which increases latency a lot and provides a lot of unneeded copying between kernel space and user space. Then as PulseAudio being merged into distros all the fault has been 'transferred' to 'buggy' ALSA drivers or other subsystems, it was NEVER the PulseAudio fault which is total bullshit because audio worked quite stable before PulseAudio and stopped working reliably after it has been merged.

How big the PulseAudio shit is is also visible on how big shit the other projects of Lennart are, like avahi or systemd. What is even more 'frustrating' that this moron without any sysadmin experience introduced this systemd shit because he wanted his laptop to boot faster (lack of words), here is a snipped from some old interview with him about that:

https://twitter.com/unixtippse/status/854740233077837824/photo/1



About the PulseAudio itself, just google for this: pulseaudio shit or pulseaudio problem

You will find tons of information why I typed just one word 'Everything.' ... I will just put some of them here because I do not want to type all this stuff here by hand while its already available elsewhere.

https://www.reddit.com/r/linux/comments/4icp1a/pulseaudio_is_a_toilet_full_of_roses/
https://voat.co/v/Linux/2056902
https://bbs.archlinux.org/viewtopic.php?id=183816
https://eksith.wordpress.com/tag/pulseaudio/
https://www.ibm.com/developerworks/...t_the_f_up_on_official_mailing_list51?lang=en
https://wyldeplayground.net/pulseaudio-cracking-sound-problem/
https://groups.google.com/forum/#!topic/microsoft.public.windows.vista.general/Y-v7VyHyySs
http://judecnelson.blogspot.com/2014/09/systemd-biggest-fallacies.html
https://ewontfix.com/15/

I added some systemd link because of someone does not learn on earlier projects (avahi / pulseaudio) and still creates shit instead of software (systemd) then this shows best how great PulseAudio really is.

Regards,
vermaden
 

sidetone

Aspiring Daemon

Thanks: 284
Best answers: 1
Messages: 853

#46
@ poorandunlucky

I wanted to omit that as I probably stated that many times, and I consider it a waste of time, especially because its very 'negative'.

While FreeBSD has in kernel mixing of channels (up to 256 if I recall correctly) with OSS4, with low latency, in kernel space, Linux on the other hand some time ago BEFORE PulseAudio had ONE OSS channel (which often was blocked by WINE) and many ALSA channels. That worked until some app 'blocked' the OSS channel and then another app wanted to use OSS channel. In other words, a typical Linux mess.

Then PulseAudio from the great Lennart Poettering happened and it happened hard. I used Ubuntu for about a year because FreeBSD was not up to the task of supporting Intel G965 chipset with X3000 GMA Intel graphics adapter (only deadly slow vesa was possible), so I moved 'temporary' to Ubuntu to wait for its support in the FreeBSD land. Why Ubuntu and not CentOS? or why not Gentoo? I used Slackware and Gentoo in the past and these two bring me to the FreeBSD land, so I wanted to check a 'desktop' distro for the first time, and as Ubuntu was the most popular at the time, I decided to jump in.

This Ubuntu came with PulseAudio, not with the first version, PulseAudio was available for quite some time before that Ubuntu release. I did not used any 'strange' audio setup, just output to 2 speakers via typical mini jack output. As this was desktop I did not powered it off as I did not have such need, but PulseAudio forced me to periodic reboots. The whole audio on that Ubuntu Linux hanged hard after about a day of typical desktop usage, sometimes after two days, sometimes after half a day. No reload of services or even kernel modules (alsa) helped to gain the sound back, I was forced to hard reboot because of that PulseAudio shit. The strange thing is that the whole rest of the system just worked, I could do all other operations without a problem, but sound stopped with some jerky noice and thats it. As I used this Ubuntu for about a year I regularly applied updates with hope to fix this issue, but after more then a year it was not fixed. I came back to FreeBSD then and never looked back, but as I see more and more projects (Firefox/Chrome) force this shit (PulseAudio) into the FreeBSD ecosystem it drives mu nuts.

My opinion may be biased on 'this single system' but its not only my opinion. I am very disappointed that You did not seek for that yourself ... as everything that this moron Poettering does is just pure shit.

The PulseAudio basically did what Microsoft did with Windows Vista in audio subsystem, they moved it altogether into user space, which increases latency a lot and provides a lot of unneeded copying between kernel space and user space. Then as PulseAudio being merged into distros all the fault has been 'transferred' to 'buggy' ALSA drivers or other subsystems, it was NEVER the PulseAudio fault which is total bullshit because audio worked quite stable before PulseAudio and stopped working reliably after it has been merged.

How big the PulseAudio shit is is also visible on how big shit the other projects of Lennart are, like avahi or systemd. What is even more 'frustrating' that this moron without any sysadmin experience introduced this systemd shit because he wanted his laptop to boot faster (lack of words)
A lot of GNU and Linux does that. GCC forced every dependency not related to compiling for a port, so it takes 18 hours to compile it. When if you can switch over to Clang, it can take 2 minutes to 2 hours to compile that same port. It is as if Linux tries to reinstall a whole operating system around a port, and instead of fixing, they just keep adding duplications of the same types of software. They fixed much of that problem for GCC on FreeBSD, but still GNU/Linuxisms on FreeBSD add duplicate software for no reason except to use a couple of libraries. I used to compile a program on an old computer for a Linux distribution, and now I wonder how much time was needlessly wasted, waiting, and hoping I chose the correct options to make a standardized install for one program every time. What a waste of time and electricity. Port maintainers' tasks would be so much easier without that useless gunk.

Forcing unneeded junk into a compile or subsystem does not make a program faster in the longrun, it makes it slower and buggy.

GNU/Linuxisms are great for top-level/userland programs, but for subsystems/infrastructure they are terrible.
 

poorandunlucky

Well-Known Member

Thanks: 26
Messages: 359

#47
@ poorandunlucky

I wanted to omit that as I probably stated that many times, and I consider it a waste of time, especially because its very 'negative'.

While FreeBSD has in kernel mixing of channels (up to 256 if I recall correctly) with OSS4, with low latency, in kernel space, Linux on the other hand some time ago BEFORE PulseAudio had ONE OSS channel (which often was blocked by WINE) and many ALSA channels. That worked until some app 'blocked' the OSS channel and then another app wanted to use OSS channel. In other words, a typical Linux mess.

Then PulseAudio from the great Lennart Poettering happened and it happened hard. I used Ubuntu for about a year because FreeBSD was not up to the task of supporting Intel G965 chipset with X3000 GMA Intel graphics adapter (only deadly slow vesa was possible), so I moved 'temporary' to Ubuntu to wait for its support in the FreeBSD land. Why Ubuntu and not CentOS? or why not Gentoo? I used Slackware and Gentoo in the past and these two bring me to the FreeBSD land, so I wanted to check a 'desktop' distro for the first time, and as Ubuntu was the most popular at the time, I decided to jump in.

This Ubuntu came with PulseAudio, not with the first version, PulseAudio was available for quite some time before that Ubuntu release. I did not used any 'strange' audio setup, just output to 2 speakers via typical mini jack output. As this was desktop I did not powered it off as I did not have such need, but PulseAudio forced me to periodic reboots. The whole audio on that Ubuntu Linux hanged hard after about a day of typical desktop usage, sometimes after two days, sometimes after half a day. No reload of services or even kernel modules (alsa) helped to gain the sound back, I was forced to hard reboot because of that PulseAudio shit. The strange thing is that the whole rest of the system just worked, I could do all other operations without a problem, but sound stopped with some jerky noice and thats it. As I used this Ubuntu for about a year I regularly applied updates with hope to fix this issue, but after more then a year it was not fixed. I came back to FreeBSD then and never looked back, but as I see more and more projects (Firefox/Chrome) force this shit (PulseAudio) into the FreeBSD ecosystem it drives mu nuts.

My opinion may be biased on 'this single system' but its not only my opinion. I am very disappointed that You did not seek for that yourself ... as everything that this moron Poettering does is just pure shit.

The PulseAudio basically did what Microsoft did with Windows Vista in audio subsystem, they moved it altogether into user space, which increases latency a lot and provides a lot of unneeded copying between kernel space and user space. Then as PulseAudio being merged into distros all the fault has been 'transferred' to 'buggy' ALSA drivers or other subsystems, it was NEVER the PulseAudio fault which is total bullshit because audio worked quite stable before PulseAudio and stopped working reliably after it has been merged.

How big the PulseAudio shit is is also visible on how big shit the other projects of Lennart are, like avahi or systemd. What is even more 'frustrating' that this moron without any sysadmin experience introduced this systemd shit because he wanted his laptop to boot faster (lack of words), here is a snipped from some old interview with him about that:

https://twitter.com/unixtippse/status/854740233077837824/photo/1



About the PulseAudio itself, just google for this: pulseaudio shit or pulseaudio problem

You will find tons of information why I typed just one word 'Everything.' ... I will just put some of them here because I do not want to type all this stuff here by hand while its already available elsewhere.

https://www.reddit.com/r/linux/comments/4icp1a/pulseaudio_is_a_toilet_full_of_roses/
https://voat.co/v/Linux/2056902
https://bbs.archlinux.org/viewtopic.php?id=183816
https://eksith.wordpress.com/tag/pulseaudio/
https://www.ibm.com/developerworks/...t_the_f_up_on_official_mailing_list51?lang=en
https://wyldeplayground.net/pulseaudio-cracking-sound-problem/
https://groups.google.com/forum/#!topic/microsoft.public.windows.vista.general/Y-v7VyHyySs
http://judecnelson.blogspot.com/2014/09/systemd-biggest-fallacies.html
https://ewontfix.com/15/

I added some systemd link because of someone does not learn on earlier projects (avahi / pulseaudio) and still creates shit instead of software (systemd) then this shows best how great PulseAudio really is.

Regards,
vermaden
Well, I can't see that it's any more or less quirky than everything else so far, or any less inclined to "just work" even if not properly or not in the best manner right out of the box (who's to say what's out of the box, or what the box is, really, though), but I'll keep an eye out... Thanks...
 

free-and-bsd

Aspiring Daemon

Thanks: 66
Messages: 633

#48
Well, I can't see that it's any more or less quirky than everything else so far, or any less inclined to "just work" even if not properly or not in the best manner right out of the box (who's to say what's out of the box, or what the box is, really, though), but I'll keep an eye out... Thanks...
Well, in Linux I personally didn't have any problems with pulseaudio.
But under FreeBSD it just doesn't work for me, while OSS does work out of the box. I have to build Firefox with pulseaudio (SNDIO not very good), but it doesn't give sound until I force remove pulseaudio. I don't see any need for it on FreeBSD.
 

free-and-bsd

Aspiring Daemon

Thanks: 66
Messages: 633

#49
@ poorandunlucky

I wanted to omit that as I probably stated that many times, and I consider it a waste of time, especially because its very 'negative'.

While FreeBSD has in kernel mixing of channels (up to 256 if I recall correctly) with OSS4, with low latency, in kernel space, Linux on the other hand some time ago BEFORE PulseAudio had ONE OSS channel (which often was blocked by WINE) and many ALSA channels. That worked until some app 'blocked' the OSS channel and then another app wanted to use OSS channel. In other words, a typical Linux mess.

Then PulseAudio from the great Lennart Poettering happened and it happened hard. I used Ubuntu for about a year because FreeBSD was not up to the task of supporting Intel G965 chipset with X3000 GMA Intel graphics adapter (only deadly slow vesa was possible), so I moved 'temporary' to Ubuntu to wait for its support in the FreeBSD land. Why Ubuntu and not CentOS? or why not Gentoo? I used Slackware and Gentoo in the past and these two bring me to the FreeBSD land, so I wanted to check a 'desktop' distro for the first time, and as Ubuntu was the most popular at the time, I decided to jump in.

This Ubuntu came with PulseAudio, not with the first version, PulseAudio was available for quite some time before that Ubuntu release. I did not used any 'strange' audio setup, just output to 2 speakers via typical mini jack output. As this was desktop I did not powered it off as I did not have such need, but PulseAudio forced me to periodic reboots. The whole audio on that Ubuntu Linux hanged hard after about a day of typical desktop usage, sometimes after two days, sometimes after half a day. No reload of services or even kernel modules (alsa) helped to gain the sound back, I was forced to hard reboot because of that PulseAudio shit. The strange thing is that the whole rest of the system just worked, I could do all other operations without a problem, but sound stopped with some jerky noice and thats it. As I used this Ubuntu for about a year I regularly applied updates with hope to fix this issue, but after more then a year it was not fixed. I came back to FreeBSD then and never looked back, but as I see more and more projects (Firefox/Chrome) force this shit (PulseAudio) into the FreeBSD ecosystem it drives mu nuts.

My opinion may be biased on 'this single system' but its not only my opinion. I am very disappointed that You did not seek for that yourself ... as everything that this moron Poettering does is just pure shit.

The PulseAudio basically did what Microsoft did with Windows Vista in audio subsystem, they moved it altogether into user space, which increases latency a lot and provides a lot of unneeded copying between kernel space and user space. Then as PulseAudio being merged into distros all the fault has been 'transferred' to 'buggy' ALSA drivers or other subsystems, it was NEVER the PulseAudio fault which is total bullshit because audio worked quite stable before PulseAudio and stopped working reliably after it has been merged.

How big the PulseAudio shit is is also visible on how big shit the other projects of Lennart are, like avahi or systemd. What is even more 'frustrating' that this moron without any sysadmin experience introduced this systemd shit because he wanted his laptop to boot faster (lack of words), here is a snipped from some old interview with him about that:

https://twitter.com/unixtippse/status/854740233077837824/photo/1



About the PulseAudio itself, just google for this: pulseaudio shit or pulseaudio problem

You will find tons of information why I typed just one word 'Everything.' ... I will just put some of them here because I do not want to type all this stuff here by hand while its already available elsewhere.

https://www.reddit.com/r/linux/comments/4icp1a/pulseaudio_is_a_toilet_full_of_roses/
https://voat.co/v/Linux/2056902
https://bbs.archlinux.org/viewtopic.php?id=183816
https://eksith.wordpress.com/tag/pulseaudio/
https://www.ibm.com/developerworks/...t_the_f_up_on_official_mailing_list51?lang=en
https://wyldeplayground.net/pulseaudio-cracking-sound-problem/
https://groups.google.com/forum/#!topic/microsoft.public.windows.vista.general/Y-v7VyHyySs
http://judecnelson.blogspot.com/2014/09/systemd-biggest-fallacies.html
https://ewontfix.com/15/

I added some systemd link because of someone does not learn on earlier projects (avahi / pulseaudio) and still creates shit instead of software (systemd) then this shows best how great PulseAudio really is.

Regards,
vermaden
Yea, it was systemd that finally made me jump ship, he-he. But I have hardly any reasons to complain, since I'm FAR better off with FreeBSD than I EVER was with Linux.

And, ironically, it was not even systemd or pulseaudio itself that made the last drop for me -- it was the fact, that with fresh Fedora kernel updates sound just stopped working on my system completely. I filed a bug report -- it wasn't even accepted under the pretext that it was "a duplicate of some solved bug" etc. But since I changed NOTHING on my system and since there was no other link than recent kernel update, I knew better what caused it. Why, a simple reboot with the previous kernel version made sound work again!

Nothing similar had EVER happend on my Linux systems before, in 10 years. This was the time I realized RedHat derived distros were not going to be useful for me any more. So... with that attitude in them it's hardly surprising that they let programmers into the project who break the system. The funny thing is, most distros accepted those changes. Given all this I can hardly blame that one programmer. Just let him enjoy his good luck as long as it lasts.
 

ILUXA

Well-Known Member

Thanks: 293
Messages: 495

#50
I also stopped my experiments with CentOS, after one day I executed 'yum update' and,
after reboot, systemd just stopped working... :D It showed some errors, that it cannot load bla-bla-bla
and the only one option was to boot server from bootable usb... And it was CentOS, one of the most "stable" GNU/Linux distros...
Thats why I like FreeBSD, because I'm 100% sure that my system will boot every time.
 
Top