On Wednesday, November 13, 2024 6:33:38 AM Central Standard Time Holger
Dehnhardt wrote:
I didn't have these problems under jack2, but then
the RME was also
connected to the Scarlett via TOSLink, so there was only one interface
visible to the system.
That is the difference. When you have two interfaces either they must have the
sample clocks synchronized using hardware (e.g. word clock connection between
the devices, or possibly Toslink if the device can be configured that way), or
you must use software resampling.
Errors during the software resampling is likely the cause of the noises you
hear.
I therefore suspect that it is a problem with the
hard disk usage, as I can't think of any other reason.
That is the least likely problem. Pipewire does not use the hard disk, and if
there were a problem with hard disk access by the audio application it would
result in underrun warnings.
not found a way to optimize hard disk access and I
don't even know if this
is done directly by Pipewire or if one of the usual libraries is used for
this.
Pipewire does not access the hardisk so not related at all to pipewire
libraries.
Have you seen this phenomenon before or do you have
any ideas what
else I can try?
How have you configured the two devices in Pipewire?
Are you using the pipewire-jack interface to access the pipewire server as a
JACK server? If so you could try configuring pipewire so that only the RME is
accessed by pipewire, the Scarlett is ignored by pipewire, and then use
zita_a2j and zita_j2a to access the Scarlett as a resampled device.
Was there some reason you did not leave the Scarlett connected using Toslink?
That is probably the easiest way to get the two devices working together.
--
Chris Caudle