Tracing real rays
Revision 2020 entry by Nupo.
In this video a "lasered" version of "GNSS-stylus" (search the web for it
if you want to know more) is used for the first time to scan (or "raytrace")
a 3D-object. The idea to scan a toilet seat originated from a legendary
Amiga demo called Hardwired (released 1991) by Crionics & Silents. There
was a decommissioned toilet seat lying in the shed so why not? It actually
has interesting and challenging shapes. :) Brush and paper reel visible in
the end of the video were also "raytraced", but there was so much hassle with
them that I decided to leave the "tracing" of them out of the video.
Collecting the data and generating point clouds and stuff were done by my own
(QT & open source) program: https://github.com/GNSS-Stylist/GNSS-Stylus.
I will probably also release the Processing sketches, objects and data in some
form later (once I clean them up a little).
Other programs used:
- QT creator for creating the SW used to communicate with the devices and for
logging/manipulating the data.
- Processing for creating CGI-parts
- MeshLab for creating 3d-objects (out of the point clouds) you can see on the
CGI-parts.
- OpenShot for video editing
- Audacity for audio editing
- WinUAE to grab a video of the original demo
It was quite a rush to put this together. Started working on this wednesday
and deadline for the compo is sunday 0:00. It's now about 12:30 saturday and
will definitely try to not touch anything any more :). Most of the things
needed were of course already ready, but SW needed some fixes and all the
generating/editing/Processing sketching and stuff took it's time. Partly due
to this and my lack of "stylistic" skills the production values in this video
are not up to par with general demoscene-level. I tried to avoid making the
video overly long and boring by leaving out most of the yada yada. But now
it feels even a bit rushed(?) But well, it is what it is now, no time to fix
it any more.