User-Agent: Mozilla Thunderbird 1.0 (Windows/20041206)
On Phone: A. Bernstein, D. Dale.
Present at the Meeting: M. Ito, R. Demirchyan, D. Lawrence, D. McNulty,
Y. Prok, A. Teymurazyan, M. Wood
Secretary: M. Payen
Please note: Last week, we projected the screen in the room on the
internet. However, one must have an account
to access this feature. What is seen here at JLAB is exactly what you'll
see on your end as well, and therefore,
eliminating the need to navigate your way through a meeting. Obviously,
this would make the meeting much more
efficient and easier to follow. More this on that later
*_Topics of Discussion_*
*__*
*__*1. *Update of Compton Analysis*
Yelena showed some interesting plots from compton runs that were
compared to Geant simulations.
These plots are located at
//~yrok/primex/compton/car//on//b/ctogeant.html/. Generally, the data
and Geant
showed good agreement. However, Dan noticed in the /first plot/
/xynoveto /that there's a faint image at the
right of beam hole. Yelena hadn't noticed it before. This perhaps needs
further investigation.
In addition, Aron was curious of the peculiar shape of her
plots, and she stated that the shape was
attributed to the fiducial cuts being used. On /plot thnoveto, /there's
a noticeably dip between 0.7-0.8 degrees.
This is also present interestingly enough in Geant. The dip is in the
order of a half degree.Yelena stated that it's
closest to the minimum opening angle, but she wasn't sure if this
correlates. It appears before any cuts were
placed. Ilyia believes it may be because of the trigger set missing.
Also, the /plot/ /thveto /shows a low energy
tail that is not present in Geant.
Mike believes the distance between Veto and HYCAL that hasn't
been incorporated into the analysis/vetobank,
may explain what we're seeing in Yelena's plots. Dan questioned the top
four plots, but in particular /xynoveto. /The
plots appear to be a bit asymmetric. The blog on the left is bigger than
the blog on the right. Dan suggested it could be
the effect of a residual field, because the magnet wasn't degaussed when
turned off.
A procedure used to estimate compton pairs in comptons runs is
located at //~yrok/primex/compton/car//on//b/counting.html./
Raphael strongly believes that using the delta phi constraint will help
minimize the background and provide more accurate
estimates.
In //~yrok/primex/compton/car//on//b/findz.html, /the distance
from target to HYCAL agrees with Dustin's analysis a
month ago, but still doesn't agree well with the survey. It's believed
to be off by ~7cm. A summary of compton runs with their respective Geant
plots can be found at //~yrok/primex/compton/car//on//b/sumtable.html./
2. *Update on Veto Reconstruction*
* *Mike will address this topic next week.
3. *Order in calling of tagger/hycal reconstruction*
* *In the PrimEx analysis code, Dustin would prefer that the tagger
code be reconstructed first, and he wishes to incorporate some new
definitions and clusters to be linked. However, after a long discussion
and a few informed opinions by Dave, Mark and Dustin will have to give
this matter some more thought on how to proceed.
4. *Anomalies in clock scalers and other normalization issues II: Dave
and Aram*
* *Aram believes he has a function that may correct the problem with
live1 & live2. Dave noticed that problem is localized at bit 16 & bit 24
for live1 and live2. However, Dave needs to confirm there's a problem
with bit 8. He may have to look at more statistics to do so. Ed, the
designer of the Trigger Supervisor Module and part of the Data
Acquistion Group, has performed tests to reproduce this problem. At the
core of this problem is the implementation of four 8 bit scalers instead
of the preferred 32 bit scaler. It appears systematic enough to correct,
but unacceptable as per Dustin.
5. *See
/group/primex/html/weekly_meetings/slides_2005_04_29/**time_combine_05058.eps*
* *Dustin's first plot shows some irregularities with the time field
of the event ID bank. This needs to be examined further. The bottom plot
is simply the projection onto the y-axis.
6. *More reconstructed electrons in tag_m banks than in tag_t banks*
Aram reminded us that on numerous occasions he has noticed more
reconstructed electrons in tag_m banks than in tag_t banks. It was
suggested that Eric should have a look at the code and the calibration
constants to perhaps rectify this problem or at least understand what is
really happening there.
7. *Next week, Dustin will give update on pi0 analysis.*