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]

P3D Hyperstereo, lens spacing, and ortho



Hi again,

I was just reading Mark's post and wanted to make sure people knew that
the focal length of the lens does not have enything to do with whether
the image is hyper or hypo stereo.  The hyper and hypo are only related
to the stereo base, whether the separation between the camera lenses or
amount the camera is moved between each exposure is greater or less than
the average distance between people's eyes.  Hyper and hypo effect the
apparent size of the subject of the stereograph.  The Hyperstereo of the
3Discover images (having the cameras 10 feet apart) make the cities or
landscapes look like little scale models.  The Hypostereo of Andrea's
tabletop make the Mantis look huge.  For me, with close to 60mm distance
between my beady eyes, the Realist is noticeably hyperstereo, making
people look nearly 15% smaller than they are.  I envy those who have 70mm
eye spacing.  Maybe I ought to get the TDC with the smaller distance
between lenses, or the RBT S1 with a 59mm base.

It is true that when shorter focal length lenses are used on your camera,
the stereo base can be increased and still produce a viewable stereo
pair, but it will be hyperstereo.

Ortho stereo does require that they be niether hyper nor hypo, but
"orthoness" (mimicking what you would see if you were there in person)
also requires that the focal length of camera be the same as the focal
length of the lenses of the viewer you use to see them.  When the focal
length of the viewer is too much greater than that of the camera it
causes the depth of the image to appear greater than it really is.  This
is called stretch or the "ET" effect.  That is why portraits taken with a
Realist or Kodak or other 35mm camera makes it look like the person's
chin is sticking out a foot from their neck.  The opposite, squash, is
caused by a camera viewer focal length mismatch in the other direction.

When the image is expanded, in the case of making a print from a 35mm
negative,  you multiply the focal length of the camera times the
magnification used to get your print, e.g. to get a 3 inch (76 mm) tall
print from a negative that is 24mm tall, you have to magnify it about 4
times before cropping.  This comes out to 35mm (in the case of a Realist
of Kodak) x 4 equals 140mm.  This would be equivalent to taking a 3 inch
tall slide with a 140mm lensed camera.  This works out to be close to
ortho for a holmes Viewer which has a focal length of about 150mm.

While I'm at it, cardboarding is caused by the image or viewer being out
of focus a little.  There are enough details in the image to tell that
Bob is in front of the wall, but not enough detail to tell that Bob's
head and belly are round.  

I hope this helps someone:)

Grant Campos


------------------------------