This is the first blog about the differences seen in the data between run9 pp500 SL09g and SL10j productions. There are going to be 8 plots, 4 from each library (sorry for no on-plot labels) and the 4 plots will be new DBV timestamp, new DBV timestamp+oldSC&GL, old DBV timestamp, old DBV timestamp+newSC&GL. All the plots below are made from the .event.root files looking at the highest rank vertex. 

From the plots below, I timidly draw the following conclusion, that the source of the loss of tracks at high pT is predominantly from something coupled with the new DBV timestamp and the SL10j library. I get this from looking at the following. If I look at the SL09g to SL10j comparison for when the new DBV timestamp was used (top 4 plots) I see a loss of tracks at high pTs from the left to the right and I don't see the same loss of tracks between SL09g and SL10j using the old DBV timestamp (bottom 4 plots). The new SC&GL may be causing a small portion of the loss, but it doesn't look the be the dominant feature. Also it looks like the just the new library alone is causing some of the loss as well as can be evidenced in the comparison between SL09g and SL10j using the old DBV timestamp. I ran other tests and there looked to be changes that affected the spectra at SL10f, SL10g and SL10h (I ran 200 events with the old DBV on every library between SL09g and SL10j and looked for changes).

SL09g (left) and SL10j (right) using the SL10j production DB timestamp

SL09g (left) and SL10j (right) using the SL10j production DB timestamp, but the old SC&GL values

SL09g (left) and SL10j (right) using the SL09g production timestamp

SL09g (left) and SL10j (right) using the SL09g production timestamp and the new SC&GL values