Design 950 Winter 2003 with Brian Stone
A list of things I worked on this quarter:
- Worked on my 3-D canine skeleton model until the beginning
of February. Compare the state it was in at the end
of Autumn 2004 (.jpg) with the current
state (.mov). The head was not modeled in detail like the other bones.
I decided to not model it any further so I could move on to other things.
- Created 2 animatics for my survey. One is an animatic of sequential
drawings of a dog trotting (.avi
or .mov). The
other is an animatic of sequential photographs of a dog galloping from Edward
Muybridge's Animals in Motion (.avi
or .mov). Both
were created by selecting all the images separately. Then, I copied all images
for a sequence into one Photoshop file in different layers. I lowered the
opacity of the upper layers in order to align the body image between frames
properly. I saved each layer out as its own file numbered sequentially. I
imported the sequences into Adobe Premiere and then exported them out as .mov
and .avi movies.
- I did rigging tests to see if I could get some geometry bound
to a kinematic skeleton that follows my dog mocap markers. See
details of the initial rigging tests done on a trotting dog. Next, I worked
on binding simple cylinder geometry to a dog's leg who is climbing a stair
case. I reached succes on my 12th attempt. View
my rigging notes. View
a short movie of the bound leg in Maya.
I then managed to export this to a VRML file and figure out what to change
within the VRML file to get it to work. View
the VRML file of the dog leg climbing and descending a short stair case.
(Get
Cortona VRML viewer for Internet Explorer from Parallelgraphics.com.)
- I tried to rig my 3-D skeleton to the animation skeleton
without success. I then decided to create "stick figure" geometry
of cylinders and cubes for the entire dog. These, I successfully bound to
the animation skeleton. View
rigging notes. It took a lot of attempts to get it to export to a VRML
2 file, and then it took lots of hacking at the code to fix the VRML export
file. View my notes on the VRML
creation process.
Maya kept locking up or ceasing to respond. I tried
selecting part of the stick figure to export to VRML, thinking that it was
the amount of data the computer had to process that stopped it from creating
the file. I also tried selecting just the dog's geometry. I was able to export
the geometry but it wasn't animated in VRML at first. I managed to fix the
problem though.
View the final successful
animated stick figure VRML file. It is a normal trotting gait.
- I rendered two perspectives from the successfully bound
stick figure 3-D model using Maya and Adobe Premiere. View the side-view movie,
full-speed,
half-speed,
quarter-speed.
View the overhead-view movie, full-speed,
half-speed,
quarter-speed.
- I did a muscle rigging tutorial from Maya Feature Creature
Creations by Todd Palamar. It didn't work. His rig broke when I moved the
ik handle around. The muscle went through the 3-D bone and the hand didn't
follow the animation skeleton. See
jpg.
- I read about survey design in "An Introduction to Survey
Research, Polling, and Data Analysis" by Herbert F. Weisberg, Jon A.
Krosnick, and Bruce D. Bowen, took notes, and thought of questions to add
to my survey. See PDF of possible
survey questions.
- Thought about the information architecture for revising the
learning module. Did some sketching of layouts. Noticed that I'll need to
add the links "walking," "trotting," "canter,"
or "stair climbing" before they choose a view: "lateral,"
"overhead," "frontal," or "posterior." Then
they will choose to view as "motion capture" or "video."
I also think it might be worthwhile to add another speed. So they could view
the current movie as "full speed," "half-speed," or "quarter-speed."
I'm not sure whether to use pop-up menus since these are frowned upon due
to lack of accessability for screen readers and for impaired motor control.
Maybe tab navigation would work better.