3 Sep
2015
3 Sep
'15
11:42 p.m.
On Wed, 2 Sep 2015 23:51:36 +0200
Adrian Knoth <adi(a)drcomp.erfurt.thur.de> wrote:
On 09/02/15 21:35, Joakim Hernberg wrote:
IIRC the problem Fons uncovered had to do with the effective latency
changing by one period when clients connected/disconnected. This
doesn't appear to be related. And AFAIK it only happens with USB
devices.
--
Joakim
Well yes, I'm aware of that and it was this
extra latency that I
wanted to ascertain. What puzzles me is that each time the jack
server is invoked, it will need a different amount of latency
compensation even when no other parameters have changed.
Didn't Fons mention a similar problem lately? (think two weeks ago).
He's currently travelling, but maybe you want to check out his patches
(yet to be posted).
Note that Fons' patches are only about client execution order in
jackd1. If you also get different values in jackd2, then it's either
unrelated or needs fixing as well.