Become a better telescope mirror maker
Well sure, apprentice with an expert, study the literature paying
particular attention to the writings of the first opticians who faced
the initial barrage of obstacles to overcome, and check-in with fellow
There is more than that though. Here’s my story or how I learned to be super cautious with optics making.
I'm the type to read all the books, study all the literature,
especially going back in time to the early practitioners who faced the
initial blast of problems.
My first mirrors, well, to be blunt, were pretty hit and miss. Some
were awesome and some were terrible. How could that be? I dutifully
followed the instructions in the telescope making books. Was it going
to be necessary to star test each mirror, then redo it if it turned out
About that time I took physics lab in college. It wasn't like in high
school where we could 'fudge' the experiment with a wink and a nod.
What I learned above all else is that experiments are hard for very
real and for cognitive reasons.
I learned to over compensate for my over-confidence by surrounding
myself with every conceivable test. I used Foucault and Caustic and
Poor-Man's Caustic and Ronchi and Star Test; much later I experimented
with Hartman. But you know what I learned from all that is that it is
stupid to blindly over-test. It lowers the odds of a bad mirror
escaping for sure, but I didn't really learn more than testing is hard.
What I really needed was an understanding of what I wanted to
accomplish with the mirror and how to test for those seminal
requirements. So I studied various ways that the masters broke down
mirror making. I picked an optimal path of fewer steps with sharpened
goals: generating the curve by rough grinding, fine grinding to repair
the damage caused by rough grinding, polishing to repair the damage
caused by fine grinding. And most of all, I studied the various
approaches to parabolizing such as small and large tools, deepening the
center versus deepening the edge. I selected tests to verify each
stage. For instance, the polishing stage doesn’t need an accurate wave
rating because the mirror is nowhere near parabolized, but the
polishing stage does need a test for large and small scale smoothness
in order to answer the question, “Am I done with polishing? Can I
commence a parabolizing run?”
And it turned out that that wasn't enough.
What I discovered over the years was that I needed a self-adapting
process that I could learn from, that would alter and improve my
technique for every unique mirror situation.
So that's why I am so eager to grab people by the shoulders and get
them to see that it's not just about the stages and the tests, but how
they are used and for what purpose and at what stage in the mirror
That's why I do not think it very enlightened to use a single test all
the way through. But that's the common approach, and so people end up
arguing test 'x' is better than test 'y', and nonsense like, 'I'll show
you a bad mirror made using test y, therefore test y is bad'.
Keeping in mind that there are 20 ways to make a mirror for every 10
opticians then surely we can do better. With the OODA loop (Observe,
Orient, Decide, Act) I have been able to push hard on the boundaries of
what is commonly accepted as the limits to amateur mirror making.
One of the good uses of OODA is the idea of validated learning (I hate
the concept of 'failing quick' by the way, it's perverse). So I don't
as much make mirrors as attempt to learn from them. For example when I
made my 13.2 inch [34cm] f3.0 meniscus constant thickness mirror, I
systematically attempted to parabolize the mirror a dozen different
ways. From that emerged a standard approach to smoothly parabolizing a
very fast mirror using of all things a slightly oversized pitch lap.
It’s not sufficient to learn a test well and to stick to the same
approach for every mirror. In my mind these are but ingredients to be
used in making a tasty meal. Shall I say, “Food for thought?”