Stefan wrote:

> the CPU load of vttime and vtresample increases continuously to
> very high values,

Yes, vttime will be using most of a core, @raw is 8 channels and
it's having to do a high quality resampling on them all.  If you
only want 3 channels, use @raw:1,2,3 for input to vttime.

vtresample we can't do much about, it already defaults to the lowest
quality setting.

> strangely the process number of vtwrite is lower than vtresample,
> although vtresample has been started earlier.

This is often the case when commands are run with -B option to
put them into background.  Each command forks twice to detach
itself from the controlling terminal.  When a pipeline is starting
up, it's a race to claim process IDs as they're all forking.

> channel 1 seems to show some minor glitches, see attachment.

Looks bad.   Run the signal through vtstat -i to see
if channel one is clipping at all.

  vtread ... | vtstat -i

Maybe both channels are close to hitting 1.0 amplitude in
the stream and ch1 is slightly higher gain.  You need to
look out for clipping if using integer sample formats.

> had to call the vlfrx tools from /usr/local/bin into that
> script

As Jacek says.   cron runs its commands with quite a minimal
enviroment, not quite the same as a login shell. For example it
doesn't run .bashrc as would a login shell so any extensions
to PATH or aliases wont be available unless you put them
in the crontab or in the scripts run by cron.

--
Paul Nicholson
--