o  Incident photon energy was 4.5GeV for every event.

o  The nuclear recoil is included.

o  The cross-section tables were generated using code left over
   from the 2000 eta primakoff studies. This used the Browman
   code (not Faldt).

o  Events were only generated, not tracked.

o  Target length was 10cm in beam direction (no extension into
   X/Y plane). It was centered at 750.0 cm upstream of HYCAL.

o  Detected photons were smeared using Gaussian functions. The
   sigmas were determined by where the photon was projected to
   hit HYCAL:
      PbWO4:    Energy: 3%*sqrt(E)  position: 2.5mm/sqrt(E)
      PbGlass:  Energy: 6%*sqrt(E)  position:   5mm/sqrt(E)

o  About 2.7M events were generated.

o  At places where a fiducial cut was applied, the fiducial
   cut was defined to be at one PbWO4 block in from the beam
   hole in both X and Y and one Pb-glass block away from the
   outside edge of HYCAL.

o  The pi0 angular distribution came from sampling Primakoff,
   nuclear coherent, and interference pi0 production spectra.
   I'll have to dig around to find relative phase shift used
   since I don't recall how I left it.

PS	This is a plot of the missing mass given the incident photon 4-vector and the reconstructed pi0 4-vector. The pi0 photons were smeared in position and energy and the incident photon energy was smeared by 1%. This also includes the vertex smearing for a 10cm target.
PS	This plot is the ratio of the reconstructed to thrown pi0 theta distributions. The reconstructed values were smeared by detector resolutions and vertex position and then a fiducial cut was applied. The cut-off at about 3.5 degrees is due to the fiducial cut. The ugliness at low angles is due to a small binning effect in the thrown histogram(denominator) which is smoothed out by the smearing of the reconstructed pi0(numerator).
PS	This is the angular distribution of all the thrown, final state photons.
PS	Here is the energy distribution of all the detected final state photons. A fiducial cut is applied to each photon included. The energy is smeared.
PS	This is the pi0 mass reconstructed from the photon four vectors after smearing both energy and position. The top plot shows the mass dependance on the vertex position. The bottom plot is the projection of the top one onto the Y-axis.
PS	Here is the reconstructed pi0 angular distribution on a log scale. The blue spectrum comes from the total pi0 production cross-section including Primakoff, nuclear coherent, and interference components. The red spectrum is only Primakoff.
PS	This shows the reconstructed angular distribution of the pi0s. Energy, position, and vertex smearing are all included.
PS	Here is the angular distribution of the thrown pi0s.
< PS	Here are the reconstructed positions relative to the thrown positions at the front face of HYCAL. This shows the overall effect of the applied smearing functions.
< PS	This plot shows the difference between the thrown and reconstructed pi0 angles vs. vertex position in Z. The dots are a profile histogram filled with the same data as the 2-diminsional scatter plot. The profile histo was fit with a line which you may be able to see if you squint really, really hard.
< PS	This plot is the projection of the one above onto the y-axis. The red is from values resulting from a kinematic fit to the Energy only. The blue is due to a MINUIT fit to both energy and photon position at HYCAL (I don't know why it's actually worse). The black is with no kinematic smearing.
< PS	This shows the results of a quick Monte Carlo calculation of the angular resolution of the HYCAL for a 15cm proton target. The angular resolution was taken from the sigma of a plot of thrown minus reconstructed pi0 theta. The reconstructed value came from a kinematic fit to the parametrically smeared data. The kinematic fit varied the energy of one of the decay photons, the positions of both photons on hycal, and the Z coordinate of the vertex. The pi0 mass was required strictly, but the proton mass was required loosely through an additional term in the chi-squared. The weight of the proton mass term (in MeV) is what is plotted on the x-axis. It should be noted that without the proton term, the missing mass resolution is about 40-50MeV. This is very close to the minimum of the plot (i.e, best angular resolution). The data flattens out to the right, as it should, due to the condition that the proton term becomes so small, that it does not significantly contribute to the chi-squared. The important thing here is that even though we can improve the angular resolution slightly by imposing the proton mass in the fit, the improvement is small.
< PS	Here are two plots of mandelstam "t" from the simulation. The top plot shows t calculated from thrown values for the incident photon and resulting pi0 in red. The blue shows t calculated from the fully smeared, kinematically fit photons (combined to get the reconstructed pi0 as described above). The blue curve also uses a smeared initial photon energy (1% systematic + 0.1% statistical). The curve on the bottom is the ratio of the upper two and represents the fiducial acceptance as a function of -t. The spike near t=0 likely due to the fact that we're taking a ratio of a smeared to an unsmeared spectrum.