Pages: [1]
|
 |
|
Author
|
Topic: abnormally poor sound quality (Read 3067 times)
|
lmmars
Zebranky food

Offline
Posts: 3
|
UQM 0.4 with 3do music and voice sounded great on my system while it operated with Fedora 3. I freshly installed Fedora 4 and now everything sounds fine except UQM. UQM sounds crackly and muffled. The music files (extracted from the .uqm packages) sound as great as before on other applications. How does UQM play sound? Understanding the mechanism would give me some clues to fix this. Can someone suggest how to fix this? Thanks!
|
|
|
Logged
|
|
|
|
Novus
Enlightened
    
Offline
Gender: 
Posts: 1938

Fot or not?
|
UQM uses SDL to play sound. SDL in turn can output through aRts, ALSA and OSS at least; it attempts to auto-detect the most suitable sound output. UQM prints to standard error information on which sound output it is running.
SDL audio subsystem initialized. Opening SDL audio device. using alsa at 44100 Hz 16 bit stereo, 4096 samples audio buffer
You can override this by setting the environment variable SDL_AUDIODRIVER to e.g. "alsa", "arts" or "dsp".
|
|
|
Logged
|
|
|
|
|
lmmars
Zebranky food

Offline
Posts: 3
|
Dear meep-eep,
I thought it may have been openAL too, but openAL doesn't seem to be on my system. Thanks for your suggestions.
Dear Novus,
Setting the SDL_AUDIODRIVER variable solved the issue. I think SDL was forwarding the audio to ESounD (but I'm not sure -- is there some way to check?). Setting the variable to ALSA made everything sound normal again, but disables the audio-stream mixing ESounD provides. I'm going to check whether my ESounD is setup correctly.
Thanks for your help, folks!
Question: Do you know a permanent way to change the SDL audio driver?
|
|
|
Logged
|
|
|
|
|
Novus
Enlightened
    
Offline
Gender: 
Posts: 1938

Fot or not?
|
As I said, UQM will report the sound device SDL is using on stderr. Look for the line that says something like:
using alsa at 44100 Hz 16 bit stereo, 4096 samples audio buffer
Here, "alsa" is the sound system being used (ALSA).
You can automatically set environment variables on login by adding them to a script that is run at login. Try looking at ~/.profile and ~/.bashrc. Adding "EXPORT SDL_AUDIODRIVER=alsa" to ~/.profile should do the trick. See also this article on shell initialisation.
|
|
|
Logged
|
|
|
|
lmmars
Zebranky food

Offline
Posts: 3
|
SDL was definitely using ESounD when the audio sounded garbled.
meep-eep and Novus: Your suggestions work. Thanks!
Novus: Sorry to not read your first message more carefully. I have a bad tendancy of not noticing some key words while looking directly at them. It's strange.
|
|
|
Logged
|
|
|
|
Pages: [1]
|
|
|
|
|