At 3:59 AM +0000 1/16/03, olympus-digest wrote:
>Date: Wed, 15 Jan 2003 15:45:47 -0600
>From: clintonr@xxxxxxxxxxxxxxxx
>Subject: Re: [OM] Testing for light leaks
>
>A significant number of the original Stylus Zooms (35-70mm, black body)
>develop light leaks around the back cover. The solution is to adhere a
>strip of self-adhesive felt along the top edge of the back frame (above
>where the back door closes) and a strip of foam along the bottom edge of the
>back door outside the rubber seal. Corrects the problem 99.9+0f the time.
>
>Actually, from my experience, it's alot easier to wait for a sunny day to
>check for light leaks. I've fired a flash 100+ times from every angle at a
>camera with film in it to check for a light leak only to have it come back
>after only a few minutes in the sun. Things can happen as you twirl the
>camera in sunlight with light constantly strking it that don't happen when a
>flash fires 1/10k second....
I would guess that the problem is that the light from a flash is very
directional, and one may have difficulty hitting upon the exact right angle.
One way to solve such a problem is to put the camera inside an "integrating
sphere" and fire a studio flash into the sphere. The effect of the sphere is
to surround the camera in a pool of diffuse light coming from all directions.
What's an "integrating sphere"? It's a spherical cavity painted flat white on
the inside, several times larger than whatever is being measured. The defining
property of an integrating sphere is that the illumination within is perfectly
uniform regardless of the illumination pattern of the source, so integrating
spheres are widely used in the measurement of the total optical output of
sources. A major manufacturer of integrating spheres for scientific use is
LabSphere <www.labsphere.com>. Their website has lots of application notes and
tutorials.
For crude uses like testing for light leaks, the "sphere" need not be even
close to spherical, and the painted inside of an ordinary cardboard packing box
will do nicely. I use a standard "china barrel" box (2x2x4 feet) spraypainted
with white enamel, but any indoor latex housepaint will do. The best latex
paint would be "pastel base" white, the stuff to which colors are added to make
all those oddly named designer paints, applied with brush or roller. Put the
box open end up, hang the camera down in the box a foot from the floor, and
fire a few full-power flashes down into the box a few times. A painted
cardboard cover with a hole just large enough for the flash to poke through can
be put on top of the box.
One can also use the barrel to measure the total optical output of a flash.
For this, add a white-painted diagonal baffle one foot deep that divides the
opening into two triangles, and fire the flash into one triangle and receive
light to a flashmeter from the other, arranging things so there is no direct
lightpath from flash to flashmeter. Total optical output is what matters when
one is using an umbrella or softbox or the like.
Use of a cube versus a sphere causes some non-uniformity (especially in the
corners) and thus some error, but it's good enough for most photo tests.
Better is a cube with corners filled, yielding a "truncated cube" polyhedron.
The unit of total optical output in the metric system is the lumen-second, this
being one lumen for one second.
Joe Gwinn
>Chuck Norcutt wrote:
>
> > I have an OM-PC and a Stylus 35-70 zoom that have light leak problems.
> > I think the PC needs new foam despite the fact that it looks OK and the
> > same for the Stylus, I think it needs a new rubber door seal... also
> > despite the fact that it looks OK.
> >
> > The question is, short of ruining a whole roll of film, is there some
> > means to test whether the leak has been repaired? I don't have any
> > developing tanks anymore but the only thing I've been able to think of
> > is to insert a short strip of high speed B&W into the film gate and then
> > let the camera soak in the sun for many hours. Is there a better way?
> >
> > Chuck Norcutt
> > Woburn, Massachusetts, USA
< This message was delivered via the Olympus Mailing List >
< For questions, mailto:owner-olympus@xxxxxxxxxxxxxxx >
< Web Page: http://Zuiko.sls.bc.ca/swright/olympuslist.html >
|