The wavefronts at each position of the detector have been calculated and compared to those of PIPE2D-489, since (1) wavefront differences likely drive the FRD residuals, and (2) discrepancies between these wavefronts and those in PIPE2D-489, made from the same data set, would indicate an issue with wavefront incorporation in my FRD code. Wavefront_test2.png demonstrates the wavefronts I extracted, which qualitatively mirror that from PIPE2D-489. The FRD residual jump in 2DResidualsTotal.png between the third and fourth rows is also correlates with a large wavefront change between these rows. No clear reason driving the 5th and 6th column jump is apparent, however.

PSFs with and without the masking due to the camera, spiders, and slit were compared; these likely do not drive the large residual jump.

No_Recentering.png displays the residuals in 2D across the detector when the recentering algorithm is not implemented in the code. The most immediately noticeable trait is the continuity of the residual magnitudes across the detector across columns. This raises the strong possibility that the PSF alignment introduces the large discontinuity seen in No_Recentering.png. While recentering attempts in the oversampled regions were attempted, they failed due to poor downsampling implementation. Thus resulting in recentering in the downsampled region. Of course, recentering in the downsampled region can lead to artifacts in the image that may manifest with enough of a change in the initial centroid (which correlates with detector position). Currently I am rewriting my recentering code for the oversampled data after advice from Neven; the hope is that incorporating this will resolve this issue.

It should be noted that a similarly continuous 2D residual distribution across the detector was found when setting the z4, z11, and z22 components of the wavefront to 0; however, these circularly symmetric Zernikes were not expected to be the cause of the strong asymmetry of residual magnitudes across the 5th and 6th columns.

2D_ZeroCenter.png shows the 2D residuals when the center of the illumination and the fiber are set to have zero offset. Notably, the discontinuity in the 2D residuals is not present in this figure, indicating that the fiber-illumination offset was driving the discontinuity.