The discussion so far is about matching focal lengths (image size). In
addition to focal length, we have to worry about aperture matching.

I don't know if Bill has any evidence to the contrary, but my tests of
35mm lenses show that the apertures are matched well. In the last round
of testing with Contax/Yashica 35mm lenses, I could not detect a
difference in exposure between different lenses. I did my testing in 3
different apertures: f2.8, f8 and f22. Again, I only used my eyes and
judgment, no measurements.

I am aware of a pair of Russian Zenitar16mm fisheye lenses (which I own)
that are used in a Pentax-mount RBT camera where one lens needed a full
1/2-stop aperture adjustment to match the other lens. It is a good thing
that the lenses have 1/2 stop increments. Once used properly, the
exposures are matched. But this is Zenitar and an extreme FL. I think
most mainstream lenses from major manufacturers are well matched, within
the eye's ability to detect the changes (what difference in f-stop can the
eye detect?)

George Themelis

Hi Doc

> The discussion so far is about matching focal lengths (image size). In
> addition to focal length, we have to worry about aperture matching.
>

Of all the variables in stereo photography, I would
suggest apt. matching is rarely ever a problem.... if it was a problem,
it would surface in 2d photography as poor exposures - this is one of
the few similarities we share with 2d photography.

> I am aware of a pair of Russian Zenitar16mm fisheye lenses (which I own)
> that are used in a Pentax-mount RBT camera where one lens needed a full
> 1/2-stop aperture adjustment to match the other lens. It is a good thing
> that the lenses have 1/2 stop increments. Once used properly, the
> exposures are matched. But this is Zenitar and an extreme FL. I think
> most mainstream lenses from major manufacturers are well matched, within
> the eye's ability to detect the changes

Agreed......

> (what difference in f-stop can the eye detect?)

The eye is very poor at detecting slight
variances in f stops. The eyes ability to allow for such a huge dynamic
range (exposure latitude), has one drawback.... we are very poor at
detecting variances in light levels. I have used a light meter for many
many years....and I still marvel how I can't determine the proper f stop
setting of certain scenes, which i see over and over.... in many cases,
I am off by a stop, sometimes as much as 2. But more importantly, in
stereo viewing, if one image is brighter than the other image, either by
variance of exposure in the camera, or variance in the back lighting
between the two sides..... the brain will not detect this difference,
UNLESS it is extreme (I would guess a min. of 2/3rds of a stop or
more)....instead, the image we perceive is an average of the two
different light levels.

> But more importantly, in stereo viewing, if one image is
> brighter than the other image, either by variance of
> exposure in the camera, or variance in the back lighting
> between the two sides..... the brain will not detect
> this difference, UNLESS it is extreme (I would guess a
> min. of 2/3rds of a stop or more)....instead, the image
> we perceive is an average of the two different light
> levels.

I heartily disagree in the case of exposure because, given
slide film's rather narrow exposure latitude, small
exposure differences can lead to retinal rivalry,
especially in highlight detail. I can't quantify the
tolerance, but have noticed slight exposure mismatches
where one chip has certain highlight information that is
lacking, or appears different, in the other chip. So even
if an exposure mismatch of half a stop is practically
unnoticeable across most of the brightness range, depending
on the scene, there's a good chance that a 'perfect' stereo
effect will be spoiled.

John G.

>
> I heartily disagree in the case of exposure because, given
> slide film's rather narrow exposure latitude, small
> exposure differences can lead to retinal rivalry,
> especially in highlight detail.

this is possible, I agree....

> I can't quantify the
> tolerance, but have noticed slight exposure mismatches
> where one chip has certain highlight information that is
> lacking, or appears different, in the other chip. So even
> if an exposure mismatch of half a stop is practically
> unnoticeable across most of the brightness range, depending
> on the scene, there's a good chance that a 'perfect' stereo
> effect will be spoiled.

the more degredation of a stereo image,
the less enjoyable the effect....and I do agree this exposure issue
"can" be a problem under certain conditions... my only point was, any
differences in apt.'s are very marginal with camera lenses, as they all
comply with the same f stop system of exposure.