To: atm{at}shore.net
From: mdholm{at}telerama.com
Reply-To: mdholm{at}telerama.com


Once you get down to about 1/250 wave cell deformation, there is no simple
reason to push for less.  Why?

1. It is a 99.999% certainty that something else in your system will
contribute more error than this, negating any theoretical improvement.

2. You couldn't see the difference if you tried, even if you tried really
hard.  The Strehl ratio for a PERFECT mirror on a cell giving 1/227 wave
RMS is 0.9968 with no secondary mirror or spider diffraction.  Adding a
secondary and spider will bring the Strehl significantly lower than this
and making the cell more perfect will not compensate for the effects of the
diagonal and spider.

3. Plop, although a very useful program and a big advance for ATM's, is
still an approximation.  At this level it is likely that the
"errors" resulting from Plop's approximations are as great as the
residual cell deformations.

There are more complicated reasons why a cell giving lower theoretical
deformation might be a good idea, but it is not possible to examine these
with a simple application of Plop.  Also, it is not possible to evaluate
these with a simple examination of Plop's results.  One has to get rather
sophisticated about it.  Some work in this line has been done by Jeff
Anderson Lee and David Lewis.  Even ignoring these higher order
considerations, it is still quite likely that a cell Plop says produces
only 1/250 wave or so RMS deformation will give a very good result.  It
might not be exactly 1/250 wave, but it will be good enough that you will
have a very hard time telling the difference.

Mark Holm
mdholm{at}telerama.com

--- BBBS/NT v4.01 Flag-4