Re: The "correct" way to turn on esd.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, 2003-07-09 at 14:49, Alejandro González Hernández - Imoq wrote:
> On Wed, 2003-07-09 at 13:43, Florin Andrei wrote:
> > On Wed, 2003-07-09 at 10:29, Alejandro González Hernández - Imoq wrote:
> > > 
> > > At home I do have esd running, but here at work, by default esd doesn't
> > > run on every reboot.  What is the correct way to enable it?
> > 
> > It should be started automatically. If it doesn't (or not reliably at
> > least), you must figure out why.
> 
> Anybody else know about this?

I'm not a big authority in esd, by no stretch of imagination. However...
I believe esd is started not when you power up your system, but when you
login to your graphical environment (Gnome).
In the Gnome menus, go to Preferences / Sound and take a look at the
checkbox named "Enable sound server startup". That one should govern, if
i'm not mistaken, the launch of esd.

If that checkbox is selected yet esd fails to start, then you have some
other issues. Look in the log files and stuff like that. <shrug>

> > Myself, i disable esd on my workstations. Never felt the need to run it.
> 
> You say you disable it. How do you do it? Maybe the contrary process of
> what you do would enable it.

Yeah, i just unselect that checkbox. After that esd is never launched on
my systems.

> Also, what is the best way to enable OSS and/or ALSA? I already have

OSS is already enabled. It is the default sound drivers collection of
the current Linux kernel.
OSS is exclusive. If you use it, your apps must "speak" OSS if they want
to bypass the sound daemon.
ALSA knows it's own API, but also "speaks" OSS, so those apps that only
know OSS can continue to function.

> some ALSA-related stuff installed in my system (thanks to Matthias Sou's
> RPMs) but I haven't figured out how to enable it yet.

If you really want to do it (that is, if the current situation is a
major pain in the back), here's a recipe:

Read this document:

http://freshrpms.net/docs/alsa/

Highlights in the document (i just repeat the most important parts of
it, but you must read and follow the original):

You need to install the following packages:

	Mandatory:
kernel-module-alsa
alsa-driver
alsa-lib
alsa-utils
	Optional:
alsa-lib-devel
gnome-alsamixer
xmms-alsa

The first package in this list must match your CPU architecture (i386,
i586, i686, athlon) and your kernel version _exactly_. Take a good look
at Freshrpms and choose the right package version.
The other packages do not depend on architecture and/or kernel version.

You must configure /etc/modules.conf appropriately. Go to this page...

http://www.alsa-project.org/alsa-doc/

...and choose your card. The instructions are in the Details link.

You must comment out any existing references (if any) to sound modules
in modules.conf (the ones that were configured when you installed Red
Hat).

These are the most important things. Just follow the document on
Freshrpms and it will be ok.

After making all changes, reboot. Your system should come up with the
ALSA driver enabled. Play with either gnome-alsamixer (GUI application)
or alsamixer (text-mode application, you must run it in xterm) and
unmute your sound channels, adjust the levels, etc.
When you're happy with the sound settings, run this command as root...

alsactl store

...and your mixer settings will be saved.

> > esd sorely needs a major overhaul (or replacement :-D).
> 
> I am willing to replace it anytime :).

The problem is, there aren't many replacements for it.
arts might be a good idea for average situations (Joe Blow listening to
his MP3 collection while hacking open source code :-D), but the
relations between KDE and Gnome are not exactly extraordinarely warm and
friendly. :-)
JACK is powerful and complex, but it's more oriented towards the
professional arena, and there aren't many applications to support it
anyway. On top of that, it requires all sorts of low-latency and
real-time black magic in the kernel to reveal its full potential.

See?
Bottom-most, at the kernel level, you've got the sound driver, which
could be OSS or ALSA.
On top of that, there's the sound daemon, which could be esd, arts,
JACK, etc. It's like a dispatcher, or "traffic controller for the sound"
between the kernel and the sound apps, or between apps, or between apps
and the network (in the case of the evolved sound daemons).
On top of that, there are the sound apps, which may choose to access the
sound daemon, or bypass it and go directly to the low-level sound
driver. The latter method is A WHOLE LOT easier if the sound driver is
ALSA. The former method is preferred if you have a good sound daemon (it
should be the preferred way in most situations, but it's not because the
sound daemons on Linux are not exactly perfect).

> I just wanted it enabled because
> I got a gaim's plugin, "festival" that allowed me to convert text to
> speech and it is very cool, but if I chose the "play" method, it won't
> work if I am listening to music using XMMS, that's why I ran esd,
> configured XMMS to output to esd and also configured gaim to do it, and
> now it is working.

Yeah, that happens quite often when using OSS as a driver. I remember
when sound used to be painful. :-)

There's a good chance that ALSA may enable your applications to share
the soundcard resources without conflicts or without blocking each
other, unless your card is really really crummy (or the ALSA driver for
your card is buggy).

Since ALSA has an OSS emulation layer, you may just leave your apps
configured to use OSS and they will work automagically.
Some apps, however, have specialized ALSA output plugins. On Freshrpms
you can find an ALSA plugin for XMMS. It works well, however for me it
doesn't make any difference whether i use the ALSA or the OSS output in
XMMS, as long as my system sound driver is ALSA.

As far as i'm concerned:
- I use ALSA directly (with the applications configured with their ALSA
or OSS outputs - XMMS, Xine, etc.) for simple tasks such as playing
multimedia content.
- I use JACK for heavy-duty multimedia processing (digital sound
synthesis/processing, MIDI and DAW stuff, etc.); all this "serious"
stuff relies in a large degree upon a kernel patched for low-latency,
preempt, capabilities, ALSA RTC, etc.
- JACK is not launched by default on my systems, but only manually.
It's been a couple of years since i stopped using esd and the native OSS
drivers, never missed'em.

Here's a good sound drivers roundup, comparing OSS and ALSA:

http://www.linuxhardware.org/features/01/03/06/179255.shtml

> Thanks again for the input.

If you continue to have issues, give a shout.

-- 
Florin Andrei

"Never send a human to do a machine's job." - Agent Smith




[Index of Archives]     [Fedora Users]     [Centos Users]     [Kernel Development]     [Red Hat Install]     [Red Hat Watch]     [Red Hat Development]     [Red Hat Phoebe Beta]     [Yosemite Forum]     [Fedora Discussion]     [Gimp]     [Stuff]     [Yosemite News]

  Powered by Linux