Review Committee Meeting Friday 10 March 2006 08:00 PST / 11:00 EST
Minutes: Friday 10 March 2006 08:00 PST / 11:00 EST
Agenda and Contact Info
- Discuss S5 untriggered search results [ HTML ]
AccuConference teleconferencing service:
Phone: 1-800-704-9896, participant code: 038621#
International callers ++1-404-920-6472 with same code
- Sensitivity curves: what is the way we should refer to the SRD goal
curve now? Need to ask Peter Saulson.
- CorrPower is not essential to the high-threshold analysis. But it needs
to be noted that it could be used to eliminate zero lag triggers but was not
used in efficiency determination.
- Need to stress that the S5 data is improving. This report deals with
"early S5" data.
- How were WB thresholds chosen? Low threshold: to achieve 20 uHz after
amplitude cut; high threshold: to achieve a certain number of triggers.
- Which S4 sensitivities should we compare to? Final results or the
results after only the amplitude cut? Use final S4 results. Stress that
this is comparing a high-threshold search (which is missing certain
components of the pipeline) with a low-threshold search. This way there
aren't multiple S4 sensitivities being presented.
- hrss vs. hrss plot: narrows for loud signals, but there is still a
broad halo. What is the origin? Probably due to high frequency signals.
- Bias in time reconstruction (see here and here). This occurs for low frequency MDC frames.
It seems to be a discrepancy between the logged time and the time
reconstructed from the MDC frame. Perhaps it has something to do with
calibration. Check with Keith Thorne. Also, Igor will try doing h(t)
- 38 events in high threshold search. Only two pass all cuts and these
were in time slides. Only one of the 38 was at zero lag but it failed the
amplitude consistency check.
- One of the surviving background events (event 10) has an associated
dq flag L1:OV_MC2. This flag has a high deadtime (10%) and safety is not
yet certain. It is not clear that the trigger should be vetoed by the dq
- Zero lag event q-scan and event display.
No evidence of an environmental cause. Data quality flags suggest photdiode
saturation in H1 (H1:OV_LSC, H1:PD_OVR). Clearly stuff is happening in
the auxiliary channels.
- What will be presented is just statistics of events. Number of events;
number surviving after each cut.
- Sensitivities: would be nice to see these superimposed on noise curve
as a sanity check.
- Suggest not stating these as S5 sensitivities but rather as
"early S5 sensitivities" because we may want to also have different
"mid S5" and "late S5" sensitivities that are somewhat different.
(All prelimnary of course).
$Id: minutes-2006-03-10.html,v 1.2 2006/03/10 17:34:17 jolien Exp $