The questions:

1. In embedding data, what makers are necessary to properly include embedded EMC hits into the StEmcCollection for analysis?
2. How do the embedded EMC hits look like : spatial and energy distribution wise?
3. How well do embedded tracks match with the EMC points?

Tackling the questions one by one, let's first find out what makers are necessary in order to include the embedded EMC hits in the event to be analyzed.

The following are the relevant makers:

• StEmcADCtoEMaker - takes raw EMC hits in ADC and applies pedestal, gain, and status from the database to return EMC hits with energy values.
• StEmcPointMaker - takes EMC hits in BTOW, BSMDE, and BSMDP (output of StEmcADCtoEMaker) and clusters them per detector and combines clusters (up to three from the tree different sub-detectors) into EMC points..
• StEmcSimulatorMaker - takes simulated/GEANT EMC hits and converts them to ADC
• StEmcPreMixerMaker and StEmcMixerMaker - merge collection of simulated EMC hits with the collection of EMC hits from underlying real event.

So looking at the above, we should be able to reasonably assume that I need all of those makers to analyze embedding data. However, how then do I separate EMC hits from the event and from embedding in order to study the properties of those embedded hits? So I decided to try to run the analysis with and without StEmcSimulatorMaker, StEmcPreMixerMaker and StEmcMixerMaker. The idea is that the EMC data analyzed without those makers should be from the underlying event, and the EMC data analyzed with those makers will include the embedded EMC hits. Thus, the difference should roughly be the embedded hits.

So that was done. There shouldn't really be any difference in the event quantities, except those pertaining to the EMC hits in the event. And I found that the following quantities are unchanged (as expected):

1. Primary vertex coordinates
2. Number of global tracks
3. Number of global tracks satisfying quality cuts
4. Number of MC tracks
5. Number of MC tracks satisfying quality cuts
6. Number of primary tracks

What did change were the following (plots don't seem to be working at the moment):

1. Number of EMC points (plot)
2. Number of EMC points available for track matching (i.e. have 2 SMD planes) (plot)
3. Number of primary tracks satisfying quality cuts and matched to an EMC point (plot)
   

Everything seems reasonable so far. With the EMC Mixer makers, we get more EMC points per event and also slightly more (on average) primary tracks per event being matched to EMC points.

Next, I look at the EMC hits in detail.