Through The Eyes Of A Child | September 2, 1997 |
Base image as 2 CD covers
tteoac0.mpg (3.5MB 33secs 997frames)
tteoac1.mpg (5.5MB 58secs 1750frames)
tteoac_src.zip(33KB)
EMAIL: dick@buckosoft.com NAME: Dick Balaska TOPIC: Magic COPYRIGHT: 1997 BuckoSoft, Inc. TITLE: Through The Eyes Of A Child COUNTRY: USA (Connecticut) WEBPAGE: http://www.buckosoft.com/ RENDERER USED: Povwin 3.0[12] TOOLS USED: RENDER TIME: 36 hours ENCODING TIME: 1 hour 20 minutes HARDWARE USED: Dual PPro/200, 128MB Ram, 4.3GB Ultra Fat Hairy SCSI Imagine 128 @ 1600x1200x65536, Princeton Ultra 20 P166, 48MB Win95 P133, 32MB Win95 IMAGE DESCRIPTION:This was going to be my "Magic" entry into the IRTC. The Magic of Christmas morning, i thought. About the time that i had the idea, the IRTC started seriously discussing an animation competition and began evaluating the tools available to determine the feasability of such a competition and to lay down a basic set of rules.
I tried to finish the scene for the competition. I was going to render one frame and submit it. But, the mechanics of the motion routines got the better of me and i invested my time in that.
DESCRIPTION OF HOW THIS IMAGE WAS CREATED:I have been wanting to raytrace the train from my company's logo for some time... This seemed like a great excuse to.
I used four CPUs to render the 990 frames in the animation. NT Explorer started to be inadequate to manage the rendered frames
after a bit so i wrote a perl script to manage the frames and give me some reports. Ain't HTML wonderful. :)
Anyway, Here's a sample of what the output looks like. You can get the script too.
The black entries are frames that have been traced, The red ones need to be traced, and the white ones are in progress.
I started with this script on Linux, but the SMB link to my NT box was strangely slow to read the directory.
I got perl running on my NT box under IS; it's very fast but its kinda like kissing your sister. (It _is_ kissing, but *yech*)
I also put together a couple of other pages to help me with my pov work.
In general, i noticed the following speeds from each CPU:
2/P200 (128MB shared) 8 minutes per frame rendered. 30 seconds to parse a frame
P166 (48MB) 12 minutes per frame. 40 seconds to parse a frame.
P133 (32MB) 20 minutes per frame. 300 seconds (5 minutes) to parse a frame. Grind grind grind.
The 990 frames are all rendered as PPM files, because that is a native format for Berkeley's mpeg_encode, which is the encoder i used.
Ray |
Art |
Home |
Dick |
News |
SiteMap |
Prefs |
Pinball |
XPilot |