[MF3D.FORUM:924] Re: Black mount obsession?

[Tom Hubin <thubin@xxxxxxxxx>]

Hello Richard, Bob, et al

I am an optical engineer with an interest in stereo projection. I have
designed, in my own mind at least, a better 35mm slide projector.
However, I never seem to have time to build it. Anyway, I have given
some thought to many of the illumination and projection issues for 35mm.

So I have a few questions about MF projection. I like answers with
numbers where possible.
 
> Bob Aldridge writes:
> > I'm trying to put together a projection system for these mounts, and in
> > their current, white form, they are not opaque. Using black - or blackened -
> > board would help immensely in this respect...

Acually, the absorption or optical density qualities of the material are
what are important. The outside color controls reflected light. So you
need to look at transmission per mm of thickness to determine how much
light makes it through the mount. I'm guessing that 1% transmission
through the total thickness might be little enough. On the other hand, a
really dark slide might be in the 10% or less range itself. So maybe you
want less than 0.1% transmission through the mount.

Metal mounts would stop all of the light. I expect you are avoiding this
for economical reasons. Comments?

To test a mount I would fill the window with metal then project it and
measure the light outside the window. That would allow you to quantify
the transmission of the mount. You could probably use a good photo light
meter. You would need to specify the magnification or the distance to
the meter and the projection lens used so others could reproduce the
results.
  
> I'm working on MF projection too and came to the conclusion that the
> chips are too close together in the Spicer mounts for the optics of MF
> projection.

What is the chip size (width and height and diagonal) and the center to
center spacing?

> Edmund recommends condensers 1 1/2 times the diagonal of
> the chip.

I might challenge this. Where does Edmund's say this?

The condenser system chosen depends on the light source characteristics,
the slide dimensions, and the diameter of the projection lens. The
condenser is to direct as much light as possible from the source
(filament or emitter) through the slide and into the projection lens.
Usually the near field of the source (the filament or LED emitter
itself) is imaged onto the entrance pupil of the projection lens while
the far field of the source is imaged onto the slide. The far field is
imaged onto the slide because the far field is usually evenly, or at
least smoothly, distributed. 

To be practical, the design is often slightly different than I have
described. But these are the theoretical goals of many condenser
designs.

As to the f# of the condenser, it is chosen to be as small as possible
to collect as much light as possible. However, any light that is outside
the slide or outside the projection lens entrance pupil is also lost.

> All the suitable lenses I've tested are too large also to
> center on the chips...The big Rolleis are way to big, and the Buhl
> series 3 216mm f3.4 lenses, which are made for 35mm/superslides but have
> just enough coverage for MF, are a bit too big, but too big
> nevertheless.

Lenses do not have to be round. You can chop off the overlapping parts.
But be aware that condeser lenses are often tempered to withstand great
heat. It may be difficult to remove the excess glass from a tempered
lens. You may need to use an untempered lens or find one that is square
or rectangular. You could also anneal a tempered one then remove glass
then retemper if desired.

The condenser systems could be staggered slightly so that the lenses do
not overlap. Performance for the two channels might differ. Then again,
the difference might be so slight as to be unnoticeable. Keep in mind
that two lamps that age differently already result in differneces that
most folks ignore. Kind of a wild notion but I thought I would throw
that in just to get you to think "outside the box" a bit.

You can also use mirrors to redirect the light. The two condenser
systems could be at right angles to the slides. Above, below, to the
side, or just folded for compactness. Then rectangular mirrors used to
turn the light at right angles (or some other angle) into the slides at
notmal incidence.

As for the heat, if you use a very good IR absorbing glass plate before
the condenser lens then there won't be as much heat on the condenser
lens or slide. Then untempered glass is practical.

You may be trying to do too much with one lens per condenser system. You
may be better off with a 2 or 3 lens system. If the slides do not
overlap then there is no reason why the lenses must overlap. 

Give me the dimensions and seperations of the slides, the dimensions and
wattage of the filament, and the focal length and f# of the projection
lens.  

> I should think black mounts would be a liability for projection if they
> are entirely black due to heat absorbtion.

You can eliminate the heat with good IR absorbing glass somewhere
between the source and the slide. Schott KG1 is good stuff. 

You start with lamps that are maybe 90% IR and 10% visible. That is
optimistic. 95% IR and 5% visible is not uncommon. Anyway, that means a
100 watt lamp is 90 watts IR and 10 watts visible. 9 times as much IR as
visible. Both are sources of heat when absorbed by the slide or mount.
But the visible has a useful function in projection while the IR is
purely destructive. So I would prefer numbers like 10 watts visible and
1 watt IR. 

With that in mind I want to pass less than 1% IR and more than 90%
visible. Look at the transmission characteristics of the glass and the
thickness and do the math. I think I came up with 3mm of KG1 as adequate
once. If you want thicker glass to endure more heat then I would choose
a less absorbent glass so you don't give up too much visible light. 

IR absorbing glass is often called heat absorbing glass. That's too bad.
Because it absorbs IR and radiates heat. I think it got the name because
the slide no longer gets hot so the glass seems to absorb the heat of
the slide.  

> Ideally they should be black
> on the lens side and white on the lamp side, like RBTs.  The white
> reflects light back into the source, and the black reduces internal
> reflection in the projection optics.

Absolutely true. They should also be opaque so that no light transmits
through them.

> With the dimensional obstacles I abandoned single-mount projection and
> opted for a 2 (or 4 for dissolve) machine setup.  At least that way
> there is some standardization, with gepe mounts.  (Another issue there -
> I've tried glassed and unglassed, and the focus uniformity is so much
> better with glass that I've abandoned the economy and ease of mounting
> (2 surfaces to clean instead of 6) of unglassed.

I would guess this is because the slide is flatter. The need for
flatness would depend on the depth of field of the projection lens. A
low f# would have a short depth of field. A large f# would have a
greater depth of field but might collect less light if the condenser
system was designed for a large aperture projection lens.
 
If the condenser was designed for a small aperture projection lens then
replacing the projection lens with a larger aperture will not get you
any more light. If all of the filament image is within the projection
lens entrance pupil then the projection lens is already getting all the
light that is available to it.

This is very different than cameras where the source of light is much
larger than the lens aperture. There the light collected is
proportionate to the area of the aperture.

Tom Hubin
thubin@xxxxxxxxx
AO Systems Design