@Arnomane

Yes, you shouldn’t, but I mean if the user compiles PA and installs it, then he has to do the right configuration or use the default which is ok. Or if the user download PA through the package manager of the distro, then the distro has to config PA correctly. It is like enabling DMix by default, it was a distro choose.

“1) NO per application volume control per default. All apps that do audio playback should control ONE global PCM control (the one you can find in your mixer applet as well).”

Why not? It is a great idea, with that you could decrees the volume of music or a video if a call comes.
The other two points is PA willing to resolve.
They (PA team) are not fixing a no existent bug or reinventing the wheel. It is true it has worked ok for many people, but it can work better, and with PA this can be true. We need better sound tools, like a modern OS, take a look at Mac OS X or Win vista, they have better sound tools than Alsa can offer wight now.

I remember not too far back they were considering dropping gstreamer support from phonon but since I think this has been worked on quite a bit. On my arch install phonon-gstreamer is the default and haven’t come across any problems.

- I have a 5.1 speaker set
- The 5.1 speaker set is fed through an USB soundcard
- The USB soundcard is attached to my fileserver
- I want to be able to play sound on my 5.1 set from every computer in my network

So I need a solid solution which is able to route the complete sound stream from every client to the fileserver.
PulseAudio was the only one which provided me so far something which worked at least basically (not to mention the stuttering sound via wireless connections, strangely disappearing connections to the soundserver, missing control from KDE applications, PA related bugs when playing music via Amarok, ….).

IMHO PA could really need a little more love in KDE/Phonon.

Anyways I simply didn’t feel like bothering with PulseAudio’s nasty extra volume settings and horrible playback delays. I just wiped of any piece of PulseAudio and afterwards applications using libxine/ALSA just worked like a charm as on my other computers (ok I installed Packman’s libxine in order to have the proprietary codecs…).

As you are refering “nice sound”.

“Nice sound” would be:
1) NO per application volume control per default. All apps that do audio playback should control ONE global PCM control (the one you can find in your mixer applet as well).
2) A nice mixer applet that shows to the people which volume controls are combined with a logical AND.
3) Never create a situation where more than two volume controls are combined until the sound comes out of the speaker.

So really the most important thing anyone could do is make phonon-gst suck less.

You could try Fedora or Mandriva (I think openSUSE also), PA works very nice with those distros, and now seems that Ubuntu has a working PA configuration.

The in GNOME based distros the have now very nice sound tools, at the level of Mac OS X. We need that in KDE!

http://developer.pardus.org.tr/people/tekman/blog/?p=115

Except the one Pardus student who developed finger print login support for KDM, no one of the other blogged or were not aggregated at Planet KDE. So I have absolutely no idea what happened to the PA mixer.

For KDE apps I only configured Phonon to use PA.
For GNOME apps, with configuring gstreamer and/or ESD it works.
For the rest of the apps this http://www.pulseaudio.org/wiki/PerfectSetup should work

Everything works, and it is ease to configure the 5.1 surround speakers.

What KDE needs is GUIs for PA, like the GTK+ ones, could be plasma widgets or something in KMix. Until then I have to use the GTK+ apps like pavucontrol or pavumeter.

PA has some great features, like per-application volume control and better sound quality than the Alsa mixer.

http://en.wikipedia.org/wiki/PulseAudio#Features

For the one interested, there is a brainstorm idea to integrate PA with KDE, which would benefit all distros and users. http://brainstorm.ubuntu.com/idea/20083/

Audio hardware is greatly supported by ALSA (including basic sound mixing) and there are several good audio/video codec frameworks. If you want some platform independence within your application (e.g. on plattforms that don’t have ALSA and/or another codec framework) you can either use something like Phonon or hope that Gstreamer exists for every plattform you want.

If you want to do advanced sound mixing and editing etc. you will quickly come to Jack. You certainly don’t want to use PulseAudio for that.

If you want to have network transparent sound you should use an appropriate tool for audio (codec!) streaming (maybe IceCast or VLC) and shouldn’t transmit raw PCM data streams (PCM streams over networks are like VNC compared to NX). Ok if you want a framework for controlling and feeding remote audio devices as well there is NMM (see http://www.networkmultimedia.org ) which existed years before PulseAudio.

Ok the PulseAudio folks just wanted to do their own thing and wanted to reinvent the wheel again. Ok thats fine, Free Software lives from alternatives and maybe they have some better ideas (there is a reason that there are so many text editors but a sound server who redirects access to my ALSA audio devices and that harms perfectly working competitive audio frameworks running in parallel on that system is an unfriendly uncooperative piece of software that should have never made into serious Linux distributions.