For all previous FPD analysis, fiducial volume cut of 0.5cell width has been used.

For the preliminary Eta analysis, however, we have removed this cut in order to enhance the Eta acceptace especially at mid energy.

Since cell by cell gain calibration is done based on the pi0 photon pairs that pass the fiducial volume cut, it is unclear how much we can trust the energy scale for photons hitting outside the traditional fiducial volume. Furthermore, the reconstructed coordinates of the photons may have larger errors at the edge.

Here we look at some of these effects to evaluate the minimum "safe" distance from the edge using existing flat-distribution GSTAR results.

Single Photon Reconstructed/Generated Energy vs. Reconstructed X coordinate, with Y coordinate in the central 4X4.

Single Photon Reconstructed/Generated Energy vs. Reconstructed Y coordinate, with X coordinate in the central 4X4.

X-coordinate seems to be affected by the incident angle with the FPD at the middle position. (Xoffset~30cm) But for both coordinates, the underestimation of energy is very mild up to 0.35cell width from the edge.

Single Photon Reconstructed vs Generated X-coordinate, with Y coordinate in the central 4X4.

Single Photon Reconstructed vs Generated Y-coordinate, with X coordinate in the central 4X4.

Here again, X and Y coordinates behave somewhat differently, but the common story is that reconstructed position has effective minimum distance from the edge at ~0.25cell width. Above 0.35cell width, reconstructed position tracks the generated value reasonably well.

So we try 0.35cell width from the edge, as opposed to 0.5cell width, as the alternative fiducial cut and look at the pi0 mass distribution.

Here blue indicates events that pass the 0.35 fiducial cut, whereas black indicates those events that have at least one photon outside of the fiducial volume.