From: Frank Masci <fmasci@ipac.caltech.edu>
Subject: Re: No WISE WG meeting tomorrow, May 7 - v7.5 go, no-go poll
Date: May 6, 2014 at 4:47:23 PM PDT
To: Roc Cutri <roc@ipac.caltech.edu>, Tim Conrow <tim@ipac.caltech.edu>


Ok, I’ve thrashed this out (I think). Attached are residual delta-flats from stacking pre-normalized L1b frames in scan 48542a, for W1 (left) and W2 (right). Other scans show similar variations.

In general, the pixel-to-pixel gain-residuals are <~ 1.5% for W1, and <~ 0.5% for W2.

I’d say we’re all clear for ical v7.5, *however*, I’ve identified a possible lien which we may consider as a pre-conditioning step when generating coadds in future. Further below are delta-offset residual maps for W1, W2. There (still) appear to be a systematic residual patterns (striations) of about < +/- 1 DN which are not removed by the sky-offsets. These patterns are constant within calibration windows and warrant some sort of delta-offset correction prior to any additional gain-correction we plan to perform. +/- 1 DN may not seem like much, but it will affect faint sources, especially since frame overlaps on the sky are not perfectly random and there’ll always be a residual we can’t beat down.

Apart from creating delta-correction maps per calibration window and then applying these to the processed frames, I don’t have a simple algorithmic fix for ical. And no doubt, we’re out of time.

Regards, Frank

delta-flats:


delta-bias maps: