Re: It's The Optics Stupid!

[Tom Droege <tdroege2@earthlink.net>]

Michael and all,

This particular set of lenses has light shades that are about 16" long.  I 
suspect more a lousy set of internals for the lenses.  There are no baffles 
and they are anodized, not painted.  The anodizing might look pretty shiny 
in the near IR.  It is also possible that the lenses are not so good at the 
edges.  There are also surfaces that can be seen from the lenses that might 
reflect into them.  TOM2 and TOM3 are in a better position in this respect 
as they are at the high point of the house.

It is just one more thing to work on the lenses themselves.  Perhaps it is 
time to get out the black paint and walnut shells.

I should note that from Pojmanski's papers, he may have a similar 
problem.  Perhaps we just must live with this for wide angle work.  Note 
that not many people are doing this kind of work.  A half dozen?  So there 
is not the wealth of experience that is around for long focal length, 
narrow field work.  Still, I notice the same magic talismans are 
used.  From long experience with High Energy Physicists I know that folk 
lore tends to be transferred to where it does not apply.  Just watch the 
"magic" used when otherwise sensible physicists are trying to get rid of 
electronic noise. We are trying to get rid of noise too, but there is not 
much experience here, I think.

One advantage that we have over those using camera lenses is that our field 
is a lot flatter, and so the corrections we have to make are smaller.   The 
other advantage that we have is the we built our own lenses, so if they 
need to be taken apart and have walnut shells glued in them, then we can do it.

Tom Droege


At 05:57 PM 10/20/02 -0400, you wrote:

> > Yesterday I had Dan make aperture masks for the I lenses.
>      ..
> > In the past, the V camera has always had significantly less spread in
> > errors than the I camera.  See for example, Figure 3 and 4 from TN-88.
> >
> > From this short run, the I data has less spread than the V data, and is