Mailinglist Archives:
Infrared
Panorama
Photo-3D
Tech-3D
Sell-3D
MF3D

Notice
This mailinglist archive is frozen since May 2001, i.e. it will stay online but will not be updated.
<-- Date Index --> <-- Thread Index --> [Author Index]

[MF3D.FORUM:527] Re: Exact fl match for paired lenses.


  • From: Paul Talbot <ptww@xxxxxxxxx>
  • Subject: [MF3D.FORUM:527] Re: Exact fl match for paired lenses.
  • Date: Tue, 23 May 2000 22:52:26 -0500

Bill Glickman wrote:

> I tested my paired 80mm Mamiya lenses, and from the outcome, it seems to
> have been an expensive mistake.  I promised I would report in my results...
> 
> The first problem I have is the lenses centers are 5" apart when both
> cameras are mounted next to each other on the slide bar.  This was my first
> experience shooting at this base seperation.  The results were poor.  When I
> mounted the chromes I learned why this practice is risky... you can not
> prevent double vision on the side edges.

I assume here you are talking about the doubling of the window
frames (inability to fuse the stereo window) caused by the stereo
window not being set properly.  True?  Or perhaps there is too
much non-stereo information at the edges?

> Although I must admit the center looks awesome.

That's a good start!  :-)

> But the whole effect is destroyed by the double vision on
> the edges.

Imagine...just one week ago this same guy would never have
even noticed a window problem like this!  That's incredibly
fast progress!  :-)

> I assume this is simply the consequence of having a base wider
> than it should be..

Maybe.  But aren't you shooting 6x7?  There should be plenty of
room to set the stereo window in the RMM 50x50 mounts with 6x7
chromes.

> It is possible to get the base in a bit tighter by
> mounting the cameras vertical.... it would be cumbersome and only help a
> bit... it will not accomplish the 65mm base required.

65mm is not the only useful stereo base...

> Next I took some shots with subjects further in the distance.... I used John
> B's formula...and I still had the same problem... probably because there
> still was some near subjects in the scene, like grass.

One of the variables in the formula is distance to the nearest
object.  When they say "nearest" they mean nearest.  You have
to either crop out near objects you don't want in the scene,
or use them as the point for measuring the near distance when
calculating the lens separation.

> It increases the Wow
> effect on the far subjects but somehow everything else does not look right?
> Has anyone else experienced this?

You are probably experiencing the miniturization or Lilliputism
effect we discussed.  That and some problems fusing the whole
scene, perhaps.

> My conclusion is this.... mounting two cameras side by side on a stereo bar
> is a very difficult thing to accomplish.

It has its challenges, but it's certainly doable.

> Not many cameras fit together like
> Greg Erkers do.  So this prevents you from shooting any subjects which are
> near.

Yes, compact body width is an important criteria in selecting
cameras for twin camera stereo.

> Moving subjects therefore are limited to greater distances only, due
> to the wider stereo base.

True.

> Timing the shutters is very hard to accomplish
> with two cameras, many variables exist in this... however, with a bit of
> fine tuning, this is poosible to get close.... if the subject is not moving
> too fast, it should be OK.

Agree.  Your situation is complicated by having one mechanical
and one electronic release camera body.  (Anyone have any tips
on dealing with that?)

> The wide stereo base does not seem to produce
> images as pleasing as the normal stereo base....

That depends on the scene and the viewer's sensitivity to
hyperstereo distortions.

> Focussing the
> cameras perfectly in sync is also difficult and time consuming.

Were you trying to focus on a particular object in the scene?
Focus in stereo is usually done by hyperfocal settings, due
to needing sharpness throughout the z-axis.  If your lenses
have good focus synch, you should be able to set them to the
same distance reading and fire away.

> I also experimented with using a shorter fl lens, 43mm using the 65mm stereo
> base.  This once again did not create the realism the 80mm lens did?  I
> guess our brain is wondering why the scene is so damn wide???    our eyes do
> not see like that!

Anytime you vary the stereobase from 65mm and use non-matched
shooting and viewing FLs, you will be impinging on the realism
of the reconstructed scene.

> after this experiment, it surprises me that the Gilde MF stereo camera is
> the only new camera on the market today... but at $12k for the camera and
> one pair of stereo lenses, its just a bit pricey for my taste.

Well, with only perhaps a few hundred people in the world doing
MF stereo, there's not a lot of enticement to camera makers to
bring something to market.  :-(

> Unless someone offers me some great advise,

Try a Rolleidoscop.  :-)

> my MF stereo dreams are drifting
> away.... It seems the best I can do is shoot with one MF camera with one fl
> lens, and simply move the camera between shots on the stereo bar.  This
> eliminates all motion shots.... even trees blowing in the wind just a bit,
> right?

Pretty much.  But I think we can get your twin rig results
working well enough...for appropriate scenes, anyway.  Don't
let go of those dreams!

> Maybe I am making some mistakes or overlooking something major?  I am worn
> out and have exhausted my 3d knowledge....  Any input would be helpful, at
> this point, I'll even accept a pat on the back.... and a "Nice try" remark
> :-)

How about some encouragement from the cheerleaders?  Let's all
chant now...
  Go, Bill, Go!  Go, Bill, Go!

;-)

Hang in there, Bill.  We'll figure something out!

Paul Talbot