Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!elroy.jpl.nasa.gov!swrinde!zaphod.mps.ohio-state.edu!rpi!uwm.edu!bionet!hayes.ims.alaska.edu!raven.alaska.edu!milton!hlab
From: kilian@poplar.cray.com (Alan Kilian)
Newsgroups: sci.virtual-worlds
Subject: Re: Polarized light 3D systems: info wanted
Message-ID: <1991May31.040533.27531@milton.u.washington.edu>
Date: 31 May 91 02:01:09 GMT
Sender: hlab@milton.u.washington.edu (Human Int. Technology Lab)
Organization: University of Washington
Lines: 116
Approved: cyberoid@milton.u.washington.edu




  Well I guess I'll take a shot at explaining the LCD polarizing filter 3D
shutter system. First we need to plot what happens when in the CRT/LCD system.
If you are doing 120 frames per second, one frame is 8.3ms long (120^-1)
O.K now all that time will not be used for displaying the image. The CRT takes
some time to get the electron beam back to the top of the screen at the end
of a frame. So we'll say it takes about 1ms. It really spends .5ms above the
top of the frame and .5ms below the bottom of the frame but you don't care
about that. 

  A few other numbers from Thad Starners post of May 27:

  1.5ms Maximum time for phosphors to decay from 100% brightness to 0%
  .35ms Maximum time for switching from left eye to right eye.
  3.2ms Maximum time for switching from right eye to left eye.

So here is a time line. Each row is .5ms with things that don't line up
listed with their "real" times.

Time in ms   Video action                          LCD action
----------   ------------                          ----------
 0.0         Gun at top of CRT
 0.5         First raster line being drawn
 1.0         This is the right eye image
 1.5         
 2.0         
 2.5         
 3.0         
 3.5         
 4.0         
 4.5         
 5.0         
 5.5         
 6.0         
 6.5         
 7.0         
 7.5         Last raster line being drawn (7.8ms)
 8.0         Vertical sync (8.3ms)                 Start to switch to left eye
 8.5         First raster line being drawn (8.8ms)
 9.0         This is the left eye image
 9.5         
10.0         
10.5         
11.0         
11.5         top 40% of screen drawn               Left eye fully transparent
12.0         
12.5         
13.0         
13.5         
14.0         
14.5         
15.0         
15.5         
16.0         Last raster line (16.1ms)
16.5         Vertical sync (16.6ms)                 Start to switch to right eye
17.0                                                Right eye fully transparent
17.5         First raster line (17.1ms)
18.0         This is the right eye image
18.5         
19.0         
19.5         
20.0         
20.5         
21.0         
21.5         
22.0         
22.5         
23.0         
23.5         
24.0         
24.5         Last raster line (24.4ms)
25.0         Vertical sync (24.9ms)                 Start to switch to left eye

O.K. so what does all of this mean? Two things to look at:

1) When the left eye becomes fully transparent we have already drawn 40% of the
   left  eye image. If the phosphors decay at the minimum rate (1.5ms) The
   upper 17% of the left eye image has already decayed to black. So the upper
   17% of the screen is not useful to the left eye.

2) When the right eye becomes transparent we are before the first raster line
   of the right eye, BUT the lower 9% of the CRT still contains the left eye 
   image so the lower 9% of the screen is not useful to the right eye.

So that leaves 74% of the screen useful (100% - 9% - 17%)

  If your phosphors take longer than 1.5ms to decay then less of the left eye
is useful and more of the right eye is useful.

  If your phosphors take less time to decay then more of the left eye is useful
and less of the right eye is useful.

  These problems do not have to do so much with the switching time of the LCD
but due to the difference in switching times left to right and right to left.

  I use a pair of Sega(tm) LCD glasses with my system. They have equal 
switching times and I find that I cannot use about the lower 10% of the CRT 
because ofthe phosphor decay time. The Sega(tm) LCD glasses switch very fast. 
About .5ms to go from dark to clear. The bad thing about these is that they 
don't go very dark so you have to use a filter over them to get good blocking 
and this makes the image dim. If you turn the CRT brightness up you start to 
see the other image through the dark lens.

So with the Sega(tm) glasses I use the top 90% of the CRT and all is fine.

I hope this help everyone evaluate LCD swiching for 3D display.

 -Alan Kilian kilian@cray.com                  612.683.5499
  Cray Research, Inc.           | If god had meant us to use the metric system
  655 F Lone Oak Drive          | he would have given us ten finger and ten
  Eagan  MN,     55121          | toes. Judith Stone _Lighter Elements_


[MODERATOR'S NOTE:  Thank you, Alan.  Very well done. -- Bob Jacobson]