Prior to the locking trials...
scripts/LSC/LSCoffset script had behaved peculiarly:
This script spawns LSC/offset3 in order to remove the dark offset from the channels.
How ever the offsets had been nulled every other PDs
(i.e. The offsets REFL11 I&Q were nulled.
The offsets REFL33 I&Q had been left untouched
The offset REFL55 I&Q had been nulled
and so on.)
I found that the script run many instances of "offset3" scripts in background.
It seemed that tdsavg did not like too many averaging channels at once.
So the "&"s in the LSCoffsets were removed and now the script runs much more slowly,
but works for all of the PDs listed.
I think I have never seen the offsets in REFL33 and REFL165 nulled down to this level before.