For all 2603 frames of real data and 2000 frames of simulated:
--------------------------------------------------------------

1) FITTED SPOTS ONLY; AVERAGE OF VALID OFFSETS SUBTRACTED 

In each frame, the average offset in x and y of all fitted spots on a frame are 
subtracted from each spot offset (no filtering).  Only the fitted spots are
used to compute the mean ellipticity 

DATA (dashed_line)
-----------------------
e=0.225021

SIMULATIONS (solid_line)
-----------------------
e=0.19844	low=96 m	high=17cm
E_Histogram_No_Text_5-All_Nights-2005__fitted_only_avg_sub_low_96.71_high_0.0174.png


2) FITTED SPOTS ONLY; KX=0,KY=0 REMOVED

The FFT of the frame is taken, the kx=0, ky=0 bin is set to zero 
(no filtering besides this), and the inverse is computed to get the offsets
back.  Only the offsets of the  spots that were nonzero before the filtering 
occured are used in each frame to caclulate the ellipticity.

SIMULATIONS (solid_line)
-----------------------
e=0.226877      low=96 m         high=17cm
E_Histogram_No_Text_5-All_Nights-2005__fitted_only_kx0_ky0_sub_low_96.71_high_0.0174.png

DATA (dashed_line)
-----------------------
e=0.241583