From: "James Lerch" 
To: "Jeff Anderson-Lee" ,
"ATM List"
        
Reply-To: "James Lerch" 


----- Original Message -----
From: "Jeff Anderson-Lee" 
>
> "Cary Chleborad" 
> > Marshall electronics is a good place to get stuff....
> >
> > www.mars-cam.com
>
> Well maybe, but Supercircuits seems to have a better selection of B/W
> cameras at somewhat lower prices.  You can even pick up a 6 to 60mm CS mount
> zoom lens for $160 or a 12.5 to 75mm zoom lens (1/2" C mount) for $280 or
> even an 8-56mm motorized zoom lens for $370 if you want to skip machining
> your own mount.
>
> http://www.supercircuits.com/store/home.asp

** WARNING **

Dave Bevel went the route of C/CS mount zoom lens (not sure which, C or CS,
and not certain which zoom range he selected) and was meet with unhappy
results (Vignetting)

I have come to the conclusion that there exists a relationship between the
F/Ratio of the camera optic and the F/Ratio of the mirror being tested.  I
don't
know what the relationship is, as I don't know how to calculate the F/Ratio
of my existing camera optics (I've taken apart several camera lens sets,
and there are just WAY to many lenses in there to make any sense of!)

I can offer a few data points on my cheap Sony camera (circa 1990's).

#1 F/4.5 mirrors are no problem
#2 F/4 mirrors require that the field lens is as close the KE as possible AND
properly oriented to avoid vignetting the edges of the image.
#3 F/3.8 was no Joy (could get 90% of the surface to illuminate, but not the
edges)
#4 F/2 was seriously no Joy, only getting a little more than half the mirror to
illuminate


I believe its possible to do without the lens, and just use the bare CCD if
we use a pinhole or point source light source.  However this approach has
its own unique set of problems:

#1 Point source light, or really small pinhole = Lots of diffraction effects
#2 "Zoom" will need to be controlled by distance between KE and CCD
#3 The plane of the CCD must be kept perpendicular to the KE
#4 You won't get an image until everything is properly aligned
#5 Dust and stray light on the CCD must be addressed

Just my $0.02 worth, and while I'm tossing pennies around, let me relate some of
my experiences with video capture hardware.

WinXP and an Internal capture card often equals Trouble.  Why, you might
ask, well M$ has two version of Video Capture technology, the old and the
new

#1 The old method is called Video for Windows.  VFW is supposedly supported on
WinXP, and actually does work rather well with USB external capture
devices. However, most newer internal capture cards don't support VFW, or
kind of implement it and it doesn't work right..

#2 The "NEW" method is called DirectShow video capture, which
USED to be called
ActiveMovie (all part of the M$ DirectX technologies).  On WinXP this is
the method supported by the internal cards I've played with (Pinnacle &
ATI).  This will probably be the method used with FireWire and Digital
Video capture systems.

Here's the bad news, All three of the currently available Robo-Foucault programs
(Dale Eason's RTAFTc, Martin Cibulski's AutoFouc, and My RTAFT) all use the
older Video for Windows capture technology.  If you have WinXP and an
internal capture card, be prepared for problems!

The good news is M$ has an SDK for DirectShow, the bad news is it almost
requires a career change to implement it!  I've tinkered with it, and
almost have a method to support it with my Robo code, but the UI becomes
rather ugly!

Just something to thing about! 

Take Care,
James Lerch

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