Re: [ATM] Foucault/interferometry comparisons

[Michael Peck <mpeck1@ix.netcom.com>]

At 09:29 3/24/05, Koehler, Steve wrote:
>where Foucault and interferometry don't agree.  The heart of the report is 
>eight
>graphs where I overlay surface profiles derived from interferometry and 
>Foucault
>so they can be compared directly.  The report is here:
>

Steve:

In several of these cases it appears to me that the main failing of 
Foucault is underestimating the size of zonal defects. That's a fairly 
predictable consequence of using too few zones in the test. Another 
consequence of too few zones is that the precise locations of zonal peaks 
and valleys will tend to be off.

One of the things I was trying to do when I developed my "semi-automated" 
algorithm was to get better approximations to true surface error 
derivatives by using finer grained data. My motivation for that was that I 
had burned myself the first time I tried a fairly large and fast mirror by 
convincing myself it was much smoother than it actually was. It seems clear 
enough that Dale's Robo is succeeding at that as well, since in cases where 
there are significant zonal defects it appears to be doing about as well as 
interferometry at tracking them.

A couple more suggestions:

You really need a simple quantitative metric of how close two curves are to 
each other. I suggest the (area weighted) RMS difference. Just eyeballing 
plots seems inadequate to me. For one thing remember that the inner 40% of 
a circle comprises 16% of it's area, which is the same as the outer 8.35% 
annulus. In some of these cases there could be less (or more) going on than 
meets the eye.

As long as you're cataloging cases why not collect all known examples where 
both interferometry and Foucault have been performed and at least some raw 
data made available? Just in the past several months we've seen data for 3 
mirrors tested by a couple dozen individuals, a dozen or so mirrors of 
varying sizes from J. Lerch, a few from Scott Mulligan, several from 
anonymous professionals, one from J. Mulherin, and even one from me. With 
enough data you might eventually be able to discern some real trend. Right 
now I'm having a hard time seeing one.

Mike Peck


_________________

Michael Peck
email mpeck1@ix.netcom.com
Wildlife photography page http://home.netcom.com/~mpeck1/index.html
Amateur telescope making http://home.netcom.com/~mpeck1/astro/astro.html 

_______________________________________________
ATM mailing list http://www.atmlist.net/