64
votes
|
|
95
0
31
|
|
|
|
|
Propose your solution
Attachments
No attachments.
Duplicates
Comments
|
Auzy
wrote on the 17 Mar 08 at 11:31
|
|
|
I have no problems with sound quality in ALSA, however, in OSS, I had nothing but issues in the past (everyone needed sound servers, etc).
So whats changed, why switch to OSS again these days? Do you still need sound servers to play multiple sounds at once?
|
|
Auzy
wrote on the 17 Mar 08 at 11:33
|
|
|
Btw, I'll reserve my vote until after you've answered
|
|
tesla
wrote on the 17 Mar 08 at 11:45
|
|
|
If you read the post in the address i gave, the writer says that OSS was good enough in architect and was reliable, there was no need to write a new code. Also he says its easier to develop software and its crossplatform (OSS), and ALSA has lots of functions which are not used and slow. I read it and want to hear about these arguments.
I say if its open-sourced and has at least a few good parts, why can't we take these parts? Is it possible. If you say there is not any advantages of OSS any give it up, then OK.
I just want some sound quality improvements*. I don't know OSS, I just want to learn. Thank you for your answer.
* For example maybe an integrated enhancer in Ubuntu would be perfect.
|
|
Auzy
wrote on the 17 Mar 08 at 12:06
|
|
|
Ok, well one issue with OSS in the past I know, is there was no sound mixing.. Thats why you needed ARTS or esound at the time (both of which royally sucked). This is where the issue lies.. Basically, everything using arts could play a sound at the same time, everything using esound could play sound at the same time. But if you played sounds using both system at the same time, they were queued.
So, what this means, is if I go into doom3, and play, the sound card is totally tied up with doom3 sounds.. After quitting, every other sound the system has played is queued up (so if you are in pigin, that means sounds like clients connecting, disconnecting, etc, so it may be 30 sounds of sounds).
So yes, its fast, but speed is useless if it causes major issues.
And ALSA has low latency, and all the other stuff that muscians should love.
Can we have a muso to confirm this?
|
|
tesla
wrote on the 17 Mar 08 at 12:21
|
|
|
Ok, "well one issue with OSS in the past I know, is there was no sound mixing", but what about now? Isn't it corrected?
I also heard that Ubuntu 8.04 will use PulseAudio. Why will we use Pulse? Is it about mixing? If so, than maybe a corrected OSS a good alternative?
|
|
Auzy
wrote on the 17 Mar 08 at 12:33
|
|
|
Not sure if it does now have mixing.
The way I understand it, pulseaudio's purpose, like other sound servers, is to provide one audio interface, that can be used on any OS. Without it, every program would need audio code for Freebsd, windows,Alsa, OSS, BeOS.
With it? You just port Pulseaudio, and you save developers a lot of time. They still probably mix audio, but since ALSA can also mix, you don't get conflicts..
All I know, the day of OSS, was a headache, because if sound mixing on OSS was available, none of my distro's used it at that time..
|
|
dilomo
wrote on the 17 Mar 08 at 14:14
|
|
|
|
What about PulseAudio ?
|
|
Auzy
wrote on the 17 Mar 08 at 14:42
|
|
|
Pulseaudio as I said, is a sound server..
OSS and ALSA are interfaces to the sound card
|
|
|
|
Yeah, we need equalizer and professional sound system! :D
|
|
|
|
We need high-quality low-latency audio.
|
|
|
The only bad thing about alsa for me - is absence of resampling, that comes up when you try to play some games that output sound in other samplerate than your card's default. It CRACKLES!
Hope PulseAudio will solve that by it's fake alsa interface.
|
|
|
Ehm, i have some issues with these kind of brainstorms.
You seem to be second guessing the developpers. This has got nothing to do with usability. They, and pretty much everybody in linux land, is moving to pulseaudio. Why?
- its performant and low latency (like jack)
- it has active development and maintenance
- it is compatible with oss/alsa/esd
- it is designed with the modern desktop in mind, features like:
- controlling volume per application
- applying effects (like equalizer) per application
- have events based on sound (like turning music volume down when you accept a skype-call)
- unlike oss/alsa is it also suitable for professionals, because it is low-latency. (they are currently using jack)
- it is network transparent
So, in the future both the desktop-user, the server and the media professional can all use the same system.
|
|
Auzy
wrote on the 20 Mar 08 at 01:08
|
|
|
I should also add, that if OSS works in the way I think it does (like it used to), you will find that pulseaudio will conflict with the new KDE4 sound system (which I don't believe uses pulseaudio, and it used to be you couldn't have 2 different sound servers working properly at once). I'd just like to avoid having that situation again.
|
|
|
@auzy
No, it works fine with KDE4. They actually have an abstraction layer that does nothing itself, but just routes everything to whatever deamon the system sets up to take of it. It's called phonon.
It just exists so that programmeers of applications don't have to change their program if a user wants to use a different sound server or codec framework.
|
|
Z_God
wrote on the 23 Mar 08 at 01:45
|
|
|
Some OSS driver may be better than the ALSA equivalents. Anyhow this code could be used as an inspiration for ALSA developers.
Another thing:
Once ALSA applications are also fully compatible with these drivers with the appropriate library support in libaound, I don't see any reason to not include OSS/Linux in the repository.
|
|
|
The ALSA backend causes unacceptable latency. It is incapable of dealing with simultaneous output from multiple applications. Furthermore sound severs implemented to solve the aforementioned problem add yet more latency and overhead. Not to mention unnecessary complication.
The OSS4 backend provides very low latency and better sound quality. Due to its buit in mixer, simultaneous output from several applications is never a problem.
For this reason OSS4 is the only option for sound professionals and hobbyists wanting to work in linux.
The default configuration of many distros using ALSA and Pulseaudio is unable to cope with the needs of gamers. For example those wanting to play an FPS game with a VoiP client in the background have complications. Both the game and the VoiP client must be configured to output via Pulseaudio. The end result has so much latency that the game is now unplayable.
With OSS4 it just works. No Pulseaudio necessary.
|
Post your comment
|
