From:	FOCUS::BURROWS      "Chris"  7-JUN-1990 09:32:50.96
To:	@wfpc

CC:	BURROWS
Subj:	 WFPC Vignetting?

One possible explanation for the apparent change in the WFPC images across the 
field of the cameras is unexpected internal vignetting in the camera. Richard 
tells me that some apparent vignetting was observed in flatfields on the ground, 
but was always attributed to misalignment of the stimulus. I am not saying that 
it is present, simply raising the possibility.

 If there is any additional internal vignetting in the WFPC, 
the best way to detect it and characterize it would be to introduce a large
SM despace by RT commands. The resulting image would directly measure the pupil 
function with all its warts. The vignetting introduced by the change 
in beam shape away from focal planes (ie the detector and pyramid) should be 
negligible, while the image would be easily analysable if a 4-8 arcsec image 
were made (corresponding to 440-880 actuator steps, or 10-20 waves peak to peak 
focus error).  
This image would give us a direct measure of any vignetting, and also the 
curvature of the wavefront is directly proportional to the image brightness
(after flat fielding etc) at the corresponding image position. (and so would 
help with OTA focus activities)
I haven't computed S/N for this despace, but I imagine that the 12th magnitude 
star in 180 sec would give good results. (as presently proposed for the upcoming 
SI activities and Bootstrap phase B part 4)

I have changed the sign of the WFPC obscuration motion in my software 
(corresponding to moving the star to the symmetrically opposite part of the 
chip.) The effect on my best fit image is small (but not negligible), and 
apparently in the same sense as the observed field dependence, but definitely 
not as large. Thus known sources of vignetting do not make such large effects 
with the sorts of aberrations that we have been considering to date. Sandra 
Faber hypothesizes that the wavefront is particularly bad near the secondary 
mirror obscuration. She is checking the energetics, but it seems unlikely to me 
that the pupil function changes enough as the secondary obscuration moves to 
explain all of the observed variations, although that is not the case for the 
present fits. 

If there is additional vignetting, all bets are off on the present wavefront 
estimates. I'm passing on the above speculations in the hope of some good ideas 
as feedback ( FOC images?, Are there field dependent changes in the instrumental
magnitudes compared to ground truth?, Could the 'fingers' be caused by scatter 
off such a vignetting surface?, are there any known ghost images that would 
indicate vignetting in a deep overexposure (on the ground data - or should we 
take one now?)) Clearly, it is very desirable to eliminate this possibility.

Since writing the above I've talked with Rick Lyon (Hughes). He has 
independently concluded that there may be internal vignetting - and also has a 
possible explanation (the OTA misalignments could cause the f/24 beam to be 
off centered.) If so, after upcoming SM adjustment (tilt), the effect should 
disappear - and no long term problem would exist. (not only that but the fingers 
should go away as well - if caused by scatter off an  internal baffle)
Chris
