[hackerspaces] More Digital Film Technology Re: first time seeing tron

Stewart Dickson mathartspd at gmail.com
Thu Sep 6 05:13:26 CEST 2012


One last note about AT&T pre-divestiture (pre-1984) -- The primary 
governing principle of the
Regulated Monopoly was "Universal Service" -- Sounds like the opposite 
of the "Digital Divide" to me.
(Maybe even Net Neutrality?  Assuming all connections are two-way -- You 
can produce as you consume.)

And on to "Lawnmower Man".

1987 was THE FIRST GREAT CGI CRASH -- The Big Three 3D computer 
animation houses all went
bankrupt simultaneously -- Digital Productions, Cranston-Csuri 
Productions and Omnibus.  Then
came the Writer's Guild strike.   I arrived at The Post Group in 
Hollywood in January, 1988.  The Writer's
Strike was already well underway.

Pacific Title & Art Studio teamed up with The Post Group to develop a 
CCIR 601 4:2:2 digital video to
35mm theatrical film transfer process with digital image processing for 
improved picture quality and
a one-light printing process which produced electronic effects on camera 
original background material
which could be freely inter-cut with the camera negative.
"Video Signal Processing System". United States Patent 5,191,416; Issued 
March 2, 1993.
In *most* types of scenes you couldn't see the video scan lines, though 
1024 lines is stretching it for
35mm film.

Bill Villarreal wrote the Cray drivers for the Laser Scanner at Robert 
Able & Associates.  Stephen R. ("Bruno") George was Optical Supervisor 
on "Ghostbusters".   Bill specified the CRT to use for the film printer,
mounted the CRT on the optical lathe bed and developed the camera 
control hardware and software driver.
The camera controller counted vertical scan intervals in the video 
output (1024x1024) and blanked
the video after a fixed number of scans were recorded.   This kept 
exposure constant and eliminated
the shutter-bar effect.   Bruno did the film end of the optical 
process.   I did the programming.

The system was called The Gemini Process and it filled a niche at a time 
when it was still prohibitively
expensive in terms of disk storage capacity and network bandwidth to do 
electronic effects on
full-resolution theatrical film.   The Post Group, on the other hand had 
CCIR 601 (Sony D-1) Rank
Cintel flying-spot video digitizer (TeleCine) and the Abekas A-10 4:2:2 
video switcher.  You could do
zeroth-generation, lossless video editing.

There was also a thing called the Abekas A-60 digital disk recorder.  It 
was kind of nifty.  It was a
hardware RAID system which could record and play back a minute of CCIR 
601 4:2:2 uncompressed digital
video.  At 2/3 MByte per frame, that's 1-1/4 GBytes at 20 MBytes per 
second in and out.  It also had a SCSI
interface so that you could record video into it from Sony D-1 tape and 
then pull the digital image frames
into a general-purpose computer for processing.

In The Gemini Process, the A-60's SCSI port was on the high-speed 
interface of a Pixar Image Computer
(PIC).  The same kind that was used in Disney's CAPS system.   The PIC 
did 4-way-parallel 12-bit integer
pixel arithmetic and had 10-bit Digital-to-Analog Converters (DACs) in 
its video output.   Plus the
Catmull-Rom cubic spline image resizing algorithm was extremely 
high-quality and fast.  I wrote (with help
from Loren Carpenter at Pixar) a 4-way parallel converter from 
16-bit-per-pixel YUV D-1 video to RGB in
Pixar Channel-Processor (CHaP) Assembly Language (ChAs).  I could pull a 
frame of video off of the
Abekas A-60 and process it in about the same amount of time it took to 
the camera to record the frame to
film.   The PIC had dual frame buffers, so you could be doing image 
processing in the inactive frame buffer
while the active frame buffer was displaying a finished frame to the camera.

All of the electronic effects for "Lawnmower Man" from Angel Studios in 
San Diego and Xaos Effects in
San Francisco went through The Gemini Process on their way to get cut 
into the film.

Meanwhile, I had gotten my digital sculptures on the cover of Science 
News on August 3, 1991.  Brian
Vandellyn Park was in the Marketing department of one of the Rapid 
Prototyping vendors (DTM Corp.) to
whom I was sending exotic STL files for them to build to exhibit at 
their trade shows.   Some times they
would send me copies of what they built.  Brian was the inventor of the 
Flogiston Chair, which is what Jobe is
sitting in when he is learning at an accelerated rate in Virtual Reality 
in "The Lawnmower Man".   Brian
put me in touch with Jacqui Masson, the Set Dresser for "Lawnmower Man", 
who rented four of my
mathematical stereolithographs to put on the shelves in Sebastian 
Timms's office.
http://stewartdickson.cgsociety.org/gallery/876907

So, that's how I wound up in multiple roles on "Lawnmower Man".   I got 
to go to the Cast & Crew screening
at Columbia/Tri-Star Studio and to the Premier in Westwood.

On 9/5/12 2:18 AM, Frantisek Apfelbeck wrote:
> Thanks Stewan for the info and links I will do some more studying!
>
> And by the way as a child from behind the "Iron curtain" The Lawnmower 
> Man in which you have been involved was I think first "computer heavy" 
> graphic movie which I have seen, I guess around 1994-95.
>
> Sincerely,
> Frantisek Algoldor Apfelbeck
>
> biotechnologist&kvasir and hacker
>
>
> http://www.frantisekapfelbeck.org
>
>
> "There is no way to peace, peace is the way." Ghandi
>
...
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.hackerspaces.org/pipermail/discuss/attachments/20120905/97d7b221/attachment-0001.html>


More information about the Discuss mailing list