te (Thilo Ernst) (05/17/91)
Hi Virtual Realists, Here's my version of a REALLY cheap stereo display device, which I invented some years ago visualizing and animating some wireframes on a hercules-equipped XT (top view): _______________________________________ | _______ 'table' plane | | / PC \ | | / monitor \ | | |_LI_____RI_| | | \ .| | | \ [ . | | | \[. | | | /[ | | | / [ | | | / | | | .___^___. | | /YOUR HEAD\ | | |(somewhat| | | \stylized)/ | | -------- | |______________________________________| Simply place a mirror ([) between you and your PC/workstation screen and make your software display the left and right stereo images LI and RI side by side, with one of them (LI) mirrored (by software, too). Surprise: LI's virtual image exactly matches RI, as shown above.Full 3D look-and-feel (almost) free of charge, although not really well suited for headmounting :-) A more convenient variant is to use special 'glasses' containing prisms or (small) mirrors to make the eyes look onto different images displayed side by side (no need for mirroring of one image). Works very fine, but yields headaches if wrong adjusted. I found this solution sufficient for experiments; might be useful for off line prototyping in a system like Pausch's 5$/day VR, too. COPYLEFTed, i.e. donated to the non-commercial, non-military public domain hereby, unless this is a (g)oldie (sorry then). Have a pleasant journey into 3space, Thilo Ernst, Berlin, Germany (te@gmdtub.uucp, te@prosun.first.gmd.dbp.de, ..!unido!gmdtub!te)
webber@csd.uwo.ca (Robert E. Webber) (05/19/91)
In article <1991May17.213646.27113@milton.u.washington.edu> writes:
.
.Here's my version of a REALLY cheap stereo display device, which
.I invented some years ago visualizing and animating some wireframes
.on a hercules-equipped XT (top view):
.
._______________________________________
.| _______ 'table' plane |
.| / PC \ |
.| / monitor \ |
.| |_LI_____RI_| |
.| \ .| |
.| \ [ . | |
.| \[. | |
.| /[ | |
.| / [ | |
.| / | |
.| .___^___. |
.| /YOUR HEAD\ |
.| |(somewhat| |
.| \stylized)/ |
.| -------- |
.|______________________________________|
.
.Simply place a mirror ([) between you and your PC/workstation
.screen and make your software display the left and right stereo
.images LI and RI side by side, with one of them (LI) mirrored (by
.software, too). ...
This doesn't seem to work for me. In order for each eye not to see
both sides of the mirror, I have to press the mirror to my nose and
forehead. Having gotten control over what each eye is seeing, the eye
that is seeing the mirror also sees the screen direct on its side and
the eye that is supposed to be looking at the screen direct also sees
the dull side of the mirror and the two eyes are not fooled into
thinking they see stereo-pairs. However, if you hold a bathroom shelf
mirror like:
S H E L F
/
/
/
^
nose
where the mirror angle is such that the left eye cannot see the back
of the mirror, then even though the images that each eye sees are
completely unrelated, the eyes fuses them into a ghost overlay.
However, for cheap stereo for wire frame images, I don't see how you can
beat an old 8 color 6502-based box circa 1980 with a pair of red and green
filtered 3d comic book glasses.
But, it doesn't appear that stereo is the key issue in virtual
reality. Apparently, the important aspect of reality is not that it
is stereo-quality 3-d, but rather that it surrounds one and that one
is, in some sense, in it in real-time. The minimal VR human seems to
be a movable eye with an arm (although at the high end, we find a more
fully articulated human; this minimal VR human appears to be what was
advocated by the $5-a-day (3 year-wait) scheme).
--- BOB `of the dented nose' (webber@csd.uwo.ca)