Re: ATM Foucalt testing

[Donovan Buck <donovan@bravoz.com>]

At 11:08 PM +0100 12/22/99, Nils Olof Carlin wrote:
>Everett Cairns wrote to Donovan Buck:
> >...... If the source were EXTREMELY narrow then the the ray from 
>one zone would be
> >on the knife,while the next zone would be off. At least if 
>diffraction effects
> >were negligible. The classic bowl would be flat but 
>deep.Transition from light
> >to dark would be instantaneous.Perhaps we could say the image had high
> >acuteness.
>
>Often it is said that you don't have to consider diffraction effects 
>with Foucault testing - I feel it is an over-simplification. 
>Diffraction means the transition will not be very abrupt even with 
>infinitely narrow source.

Okay, I think I'm getting a clue. Slit width does not necessarily 
correspond to contrast between the lightest and darkest areas of the 
shadow pattern (as far as we're concerned those areas are either 
fully illuminated, or not at all illuminated), but it does affect how 
quickly the transition from light to dark occurs. Diffraction also 
affects this transition distance since it effectively increases the 
width of the source.

I've built a Foucalt tester which uses a Quickcam to image the shadow 
pattern, and I'm using a variation of the digital data reduction 
techniques described in the most recent ATM Journal to evaluate the 
images. Currently the transition area is fairly soft, and I would 
benefit from a more sharply defined transition. Besides the technique 
described in Tex. what suggestions would the list have for making a 
narrow slit? Also, how can I minimize the diffraction.

I have considered using an imagesetter to make a slit. I believe this 
is how most Ronchi screens are made, albeit with fewer "slits" per 
inch. I am concerned however that the film substrate would affect the 
quality of the test. Perhaps if the emulsion is facing away from the 
light source this will not be an issue.

Here's another question... assuming the substrate is not a problem, 
why not just use a Ronchi screen to define the slit on a Foucalt 
tester?

At 11:08 PM +0100 12/22/99, Nils Olof Carlin wrote:
>If the light from the "narrow slit" nearest to this edge can barely 
>pass the KE when returning, the inverting action of the mirror 
>ensures that the light from the next "narrow slit" outside it is 
>fully cut out by the KE

This would be true of the Ronchi screen as well correct? It would 
make it a lot easier to position the tester since there would be more 
than one "slit" to choose from.

Also, because of the relatively narrow slit currently on my tester, I 
was easily able to see I had a TDE since I could not fully illuminate 
the left and right edges of my mirror. Would this benefit be lost 
with the slitless tester?

Thanks for your patient explanations,

Donovan Buck
Bravo Zulu! Interactive
www.bravoz.com