[Mpg123-devel] Problem with mpg123

[Johannes K. <joh...@da...>]

Hi,
Here's some more info about the situtation, found some new info...

I have a soundblaster live card installed in the system. I'm using alsa for generating 4 mono channels, one for each language.
I have a script that fires up four mpg123:s

root@mediapc01:/play# cat start
mpg123-alsa -m -q -Z -a pcm.fin  /play/audio/ska.mp3  &
mpg123-alsa -m -q -Z -a pcm.swe /play/audio/junior.mp3 &
mpg123-alsa -m -q -Z -a pcm.eng /play/audio/devil.mp3 &
mpg123-alsa -m -q -Z -a pcm.rus /play/audio/capleton.mp3 &

If I run the first line from my bash shell with the &-operator, cpuload stays normal.
But if I run it inside a  script, cpuload jumps up to 80%.
If I run the  script without &-operator, cpuload stays normal.
If I run the first line from a script changing alsa device to "-a front" (default front device) with the &-operator, cpuload is normal.

Ok, changing the alsa device has an influence.
Running the script as shown, cpuload jumps up for each process. But if I run them separately from a screen session (one process per screen)
everything works with normal 1,7% cpuload.

This is really strange, how can the process behave different when run inside a simple shell script?

This is my asound.conf:
pcm.fin {
    type share
    slave { pcm front:0; channels 2 }
    bindings {
    0 1
    0 0
    }
}


pcm.swe {
    type share
    slave { pcm front:0; channels 2 }
    bindings {
    0 0
    0 1
    }
}

pcm.eng {
    type share
    slave { pcm rear:0; channels 2 }
    bindings {
    0 1
    0 0
    }
}

pcm.rus {
    type share
    slave { pcm rear:0; channels 2 }
    bindings {
    0 0
    0 1
    }
}

Is there any better way to split the front and rear channels into 4 discrete mono outputs?

TIA:
Johannes Kullberg
Data Group Kotka
Konekotka Ky
joh...@da...<mailto:joh...@da...>