Omar does it a bit differently. It has something to do with the hit time being computed by the amplitude-weighted average of t0 times (I think for the 2 sensors in each layer?). We stopped doing it this way.
However, even that is probably not a direct comparison and what is here is still noticeably worse than my collaboration meeting talk. On this page, it's filtered by "clean tracks" (a basic selection of FEEs) and only single strip hits are used and the fitted distributions is just the raw hit time. I used hit time residuals (hit time - track time) in my talk and I don't remember if I cleaned these up.
I have asked Mike to re-bin these also, since bins are probably too big, so that may change a little bit.
Plots are updated. Resolution still looks considerably worse, though a direct comparison may not be appropriate Problems I need to fix include: -I'm not sure why the hit time plots are cut off at +/-12ns despite looking good in the root file -the error bars on SN and cluster amplitude are set to 0
/EPP
3 Comments
Tim Nelson
So... why is time resolution so much worse than Omar showed in Paris for early analysis of 2015 data?
(slide 14 here) Omar @Paris
Matthew Solt
Omar does it a bit differently. It has something to do with the hit time being computed by the amplitude-weighted average of t0 times (I think for the 2 sensors in each layer?). We stopped doing it this way.
A better comparison is to my talk here slide 15: https://www.jlab.org/conferences/hps2016nov/talks/wed/MattS_SVT_CollaborationMeeting_111616.pdf
However, even that is probably not a direct comparison and what is here is still noticeably worse than my collaboration meeting talk. On this page, it's filtered by "clean tracks" (a basic selection of FEEs) and only single strip hits are used and the fitted distributions is just the raw hit time. I used hit time residuals (hit time - track time) in my talk and I don't remember if I cleaned these up.
I have asked Mike to re-bin these also, since bins are probably too big, so that may change a little bit.
Unknown User (testa)
Plots are updated. Resolution still looks considerably worse, though a direct comparison may not be appropriate
Problems I need to fix include:
-I'm not sure why the hit time plots are cut off at +/-12ns despite looking good in the root file
-the error bars on SN and cluster amplitude are set to 0