(present: IG, TI, NK, AA, ML, MF, RK)

Status Reports
Tayfun showed some plots of the energy-weighted average position of the PS and the 3 EMEC layers. He also plotted beam chamber position vs the average EMEC position (one EMEC layer at a time). He finds a correlation with a peculiar background pattern, which needs to be investigated. Naoko says that the beam chamber information is based on a `cluster' of wires. In order to see what is behind this background pattern one might have to look at the data in athena. Richard suggested to look at pion and muon data to see whether the same effect is present there.
As to the event with zero signal he has been seeing: Naoko informed us that those are events, where the global cubic time has not been reconstructed.

Margret is writing a simple cell-finding routine to practice working with TBRootAna. She is starting from the beam chamber (x,y) position, looks for the channel with the highest energy close by and collects up cells within a radius of that channel.

Naoko reported that a bug, which had been found by Michel and Tamara has been fixed for the calibration files. She also started working on simulating data here at UVic, and pointed out that it takes 40 hours for 1000 events of 119 GeV e data!
Then Naoko showed very impressive slides of her analysis of EMEC data: She studies the pedestal noise as a function of channel number (and layer), fits each channel with a Gaussian and then selects channels with E>2sigma. Of all the channels with E>4sigma, she takes the one with the maximum signal as a seed cell and then successively collects adjacent cells all around with E>3sigma. If there are no more adjacent cells left, the cluster-building is complete. Using this method, she obtains an energy resolution of sigma/E = 1.9%.

Michel reported that he is debugging the calibration package and making it work. He is also looking at characterization of noise. His aim is to produce a cell number and its noise as an output.

TBRootAna
Ian mentioned that he has fixed a number of bugs and is now working on making the code portable. It now runs on lxplus and several other platforms. Once its portability has been confirmed the aim is to make the code public.