S5 Qpipeline High-Frequency Review Telecon 20th of May 2008 11:00 Eastern
Agenda and minutes of the S5 Qpipeline High-Frequency Review Telecon for Tuesday the 20th of May at 11:00 Eastern time.
Agenda and Contact Info
- Review of HFS draft
InterCall telecon service: USA (toll-free): 1-866-616-1738 UK (toll-free): 0800 073 8914 Italy (toll-free): 800-906-494 Germany (toll-free): 0800-1014-907 participant code: 251 288 9495#
AttendanceBrennan Hughley, Jonnah Kanner (minutes), Ik Siong Heng
Jonah: may combine eqns 1 and 2
Jonah: check coefficients in (3) and (4)
Jonah: where did 20 ms come from?
Brennan: smallest we felt was reasonable - especially for more exotic signals where you are not matching exactly the same thing.
Siong: we should quantify the difference between 20 ms and 10 ms.
Joanh: i think the burden is to show that 20 ms is a REASONABLE choice, not the best choice.
Siong: ok. but we also have to be prepared to defend why it is different from the choice in the low-freq. search.
Siong: bottom of pg. 5, justify why corr-power instead of cWB, etc.
Brennan: We don't know normalization of SN waveforms yet, working on it.
Brennan: Crappiness of GA0d5 might be due to low freq content of signal.
Siong: Maybe we should include a table of hrss 50%, 90%
Jonah: Also, maybe quote a characteristic number for the upper limits.
Brennan: not sure what to quote
jonah: ok. let's see what/if any number has been quoted historically.
Siong: my next comment is separating results and conclusions. its not clear from the results section that the thresholds are set with the box closed. that should be made more clear.
Siong: something is fishy with figure 12 - no events at 6.2 w/ gamma
Jonah: yeah, good point.
Jonah: Explain to me why data quality is different from vetoes.
Brennan: I think DQ is treated differently from vetoes
Brennan: With DQ cuts, we bascially had the DQ cuts on a per-instrument basis. So, we didn't use any time lag whatsoever. We looked at the effect on a single instrument. It is a different method for selecting them than the vetoes. You can't have the same standard for what makes a DQ and Veto becuase we are looking at single IFO compared to combined, and the rates are just different. I just tried to set some reasonable standard where I could say this data quality is doing something. 1.7 is where this data quality is doing something - looking at the table, you can see the DQ cuts are doing their job. the auxillary channels is not something i had a lot to do with - we just did the same as low-freq.
Jonah: ok. but i see in the paper a note in the DQ section about "coincident" triggers. it looks like the DQ flags WERE done with 5 second lags
Brennan: oh yeah. i hadn't thought about that at the time. there's no reason i couldn't have done it that way.
Jonah: Has there been a review/documentation of the vetoe selection software.
Brennan: I think the main reference is the S5 low freq paper
Jonah: we should confirm that has been reviewed.
- Looked at "high" statistics of low-gamma foreground - stats. at ~2 sigma level, and is consistent with an incidental occurance
- Brennan should make many of the changes suggested by the reviewers.
- Reveiwers should check with Jolien that veto selection method will be/has been reviewed
- Is 20 ms a reasonable choice of coincidence window? How is Gamma threshold affected compred with 10 ms?
- Investigate missing background events in figure 12
- Worry about biasing issues due to choosing DQ flags based on "real" background set
- Ensure small block time (16 s) is OK in Q-pipeline