3d body adventure
18 messages in this thread
In the cd-rom 3d body adventure there is an area where you can interactively
rotate rendered viewpoint models in almost real-time by using your mouse. In
the directory of the program there is a file called wf.exe that uses the ascii
information from the 3ds files and is converted into a data type by a program
call asc2dat. I was wondering if you knew what the asc2dat file is, is it part
of being able to rotate the model, and if so how does one go about acquiring
it. Lastly do you know of how they actually accomplished the interactive
rotation? Any info would be appreciated. Thanks Sam Coor
In general, the way that Knowledge Adventure pulled so-called "real time"
rotation off is by a very elegant cheat. They rendered flics for each
allowable rotation "arc" and then put a UI in front of their own flic engine
that allows folks to effectively scroll back a forth within the flic in real
time. It works great. Beyond that, you'd have to contact the folks at KA for
details.
– G
Gary, Hi again. I have for some time been interested in human
character animation, and I'd like to read a little description by
someone working in that field about what's it really like: what's
feasible, what's not feasible, what the hardware and software
requirements are, how long it takes, what the skill set is. I am
particularly interested in realistic modeling of human expressions by
morphing facial musculature. (In fact, I've even toyed with doing a
book on the subject after I get finished with my book on multimedia
authoring in VB. Is there anyone you might invite to discuss this
topic here on CIS?) Thanks. Mitch
The folks at Sony Pictures Imageworks are doing the most interesting
human animation I've seen (Tim Miller, particularly). I'm not sure
if they're available on CIS, however.
The folks at Sony Pictures Imageworks are doing the most interesting
human animation I've seen (Tim Miller, particularly). Gary, That's
cause they've got all the money….
Actually, that division (the PC division, as compared to the SGI
division)_ does NOT operate with the high budgets that the
workstation guys have. Their operating philosophy is
bottom-line-budget, and that's why they're getting so much work now.
Tim's been doing great human work since before he was working there.
Money has nothing to do with talent.
Ah ha ! I was wondering where Tim ended up after Action closed ! Good
catch for Sony .
Really. (He's been there almost a year, now.)
In addition to being a beyond-mutant animator, he's a really nice guy!
You can check out two of his skin models right here in the forum. ALIEN.ZIP and
SOLDIE.ZIP. They have underlying skeletons that you animate in the KF while
the cross-sections are hidden, and then at render-time you hide the skeleton
and unhide the cross-sections. (This is basically the same technique that
Martin Foster describes in the new 3DS Special FX book.)
– G
Thanx,
See, these are the kind of cool uploads I'm missing now that we no longer have
those wonderful "What's New" postings in the forum messages. Looks likek I've
got some catalog DL'ing to do.
Bruce
>> Is there anyone you might invite to discuss this topic here on CIS?
<<
I'd love to discuss this (although not invited). I've dabbled in this
area for some time, but I'm clearly not in the league of people like
Tim Miller 🙁 However, you might find him hard to communicate with
since I don't remember him participating in the asoft forum. I met
him once and felt an inexplicable urge to polish his shoes, however I
didn't have the guts to ask if he'd mind if I did. I just skulked in
the shadows and basked in his aura ….
Anyway, that leaves me <g>. I've built some characters with skin
(check out the Siggraph '94 tape or heroes.jpg in the libraries
here). The characters don't do a lot of animation but I'm hoping to
liven them up a bit when I get IK in r4 or Puppeteer from Schreiber.
I personally feel I'd need IK, at least, if not optical motion
capture to make extensive human animations. I attended a Siggraph
class on motion capture and it was most enlightening. I feel that's
the way to go for human motion and speech. I'm sure some really nice
stuff has been done the with traditional keyframing but usually "less
is more" and the animators really need to concentrate hard on fine
tuning just a few motions to make them convincing.
If there's anything in particular you'd like to discuss, ask away!
Better that than me waffling <G>.
Hi Martin, Thanks for the reply.
I'm not familar with Tim Miller's work, mentioned by Gary and by you….
There is a brilliant book, sort of like an atlas, called Facial
Expressions by Gary Fagin (Watson-Guptill, ISBN 0-8230-1628-5). It
shows how the facial musculature is engaged to create six basic
expressions of emotion and scores of subtle variations. I guess one
thing I was wondering is how difficult it would be to model those
features and develop methods to make them stretch. The culmination of
this fantasy would be to have them behave under some kind of
higher-level control like a menu or a script. Same thing applies to
the expressions of speech.
(Part of the reason I'm fuzzy bout this is I haven't done much 3D
modeling. Version 1 of 3DS, which I received while doing my Animator
book, was so cranky and buggy on my system that I was never able to
justify pursuing it very far, and 3D animators sprang up like
mushrooms here in Atlanta, all grubbing for the available work. And
while Gary talks about talent, he means the entertainment industry.
In the world of commercial multimedia for corporate clients, talent
is far less important than lowest-bidder…)
Mitch
Mitch,
>> I'm not familar with Tim Miller's work, mentioned by Gary and by you…. <<
me neither, but Gary speaks highly of him. There are some examples of his work
in the forum libraries and also on the Siggraph '94 (and maybe previous years –
not sure). It seems like he's good at keyframing "action" sequences.
>> Facial Expressions by Gary Fagin …. shows how the facial musculature is
engaged to create six basic expressions of emotion and scores of subtle
variations….. how difficult ….to model those features and develop methods
to make them stretch. <<
Well, it's been done so it must be feasible. Another 3ds character animation
fiend, Steve Sherer, created a whole series of heads with facial expressions
and "phonems" (sp?). There's about 16 heads which you morph between. Modeling
typically involves creating an approximation of the head, then doing
appropriate booleans, then lots of vertex tweaking.
At Siggraph '94 I saw a bunch of different methods for lip-synching. Most
interesting. If I can find my notes (I took lots of 'em) I'll go into detail.
later
Martin
check out saturday mouring at 8.00 channel 7 i think, its called reboot. all 3d
human like models cartoon. actualy its the first fully 3d animated series i
have seen on a weekly bases. very cool stuff
sherwood animations
Sherwood:
Last year I caught an episode or two of 'Crash Test Dummies', an all Wavefront
3D animated series on Saturday mornings. Don't know if its still on the air
though.
Bob
Sherwood,
I too like reboot, we have it on ABC at 9:30 Saturaday (EST). It helped
me explain to my wife why I _needed_ to get 3DS, she does "child development"
stuff <g>. I do medical/toxicology stuff.
One of the things I noted was the lack of shadows, I guess to save
rendering/production time.
It would be interesting if a thread was started allowing comments on
methods and suggestions on a series such as this that appears weekly.
i definatly agree, do you know what platform and software reboot uses?
sherwood
I read on Internet comp.graphics.animation that Reboot is done on
Softimage by Limelight Productions of Vancouver.
thats what my quess was. thanks for the info.
sherwood