I have a question about what the trigger lines mean
and hope you may be able to help us.

We have seen in the last few days, especially in a
few runs last night, that the DIRC timing peak looks
pretty bad in some cases, especially for one trigger
line. We have a plot which shows the time peak for
each trigger line in our PS files, for example in
http://www.slac.stanford.edu/BFROOT/www/Detector/DIRC/DircOperations/FastMonitoring/spew/spew-78259.ps
This is a bad looking time peak, as you can see on
page 9 which shows the peaks for all 12 sectors. The
time peak vs trigger line is shown on pages 17-19.
There you see that the peak looks OK for most of
the lines, that the overall bad distribution is
caused by line 1 which has no remnant of a Cherenkov
signal and seems to dominate the statistics.

For comparison, here is a run with healthy time peaks
from Sunday
http://www.slac.stanford.edu/BFROOT/www/Detector/DIRC/DircOperations/FastMonitoring/spew/spew-78204.ps
This one in fact has a Cherenkov peak in line 1.

And here is one with reasonably healthy time peaks
from today
http://www.slac.stanford.edu/BFROOT/www/Detector/DIRC/DircOperations/FastMonitoring/spew/spew-78270.ps
This one also has no Cherenkov peak in line 1.

The plots are filled and the trigger bit information
is retrieved in DrcOfm/DrcFM.cc.
What is the meaning of this line 1 and do you have
any model why our timing could be off for this line?
We have a suspicion that the bad peaks are caused by
bad trickle events which cause the DIRC to fire in
almost every PMT in some events, overwhelming the
Cherenkov signal in those trickle events. Knowing
what the lines mean would help very much.