#Siggraph 94 report.
40 messages in this thread
I just got back from Orlando after two days of being blown away by all the new
toys that will keep me broke for another year. <BG> Here is a quick summary of
what really excited and surprised me. (no special order)
The biggest shock of all was the display of 3DSr4. My first exposure to this
was when I walked up to the Autodesk booth and saw Gary Yost proudly
demonstrating some of the newest features in r4. Well, so much for Autodesk's
policy about keeping this stuff a big secret until its released!. I asked Gary
why and he explained that with all the competition 3DS must move along at a
faster pace. My candid opinion of this is two fold. First I'm very excited at
having some of the new features demonstrated that Gary has fully listed in his
announcement here on the forum and will welcome these additions. They will
make keyframing a breeze for both machinery and character animation. From my
brief exposure to r4 I would say that not much change to the basic look and
feel of the shell is in r4; it really has to do with an expansion of the IPAS
interface, now to be called "plug-ins" On this basis I would form my second
opinion of r4 that r4 is really not a radical release change that r2 was over
r1 and r3 was over r2 but more like a collection of very advanced "plug-ins" to
r3. So, the naming of this version as a release 4 seems to be more of a
marketing hook than a major shift in the look and feel of the 3DS shell. I
spoke with several forum friends at the show and many had agreed this was the
case. For me, I don't care what Autodesk wants to call it, even if they wanted
to call it release 5.0 to be like the competition similar to what MS did, I
just welcome the modularity of the 3DS package. The open architecture design
of 3DS is maturing very nicely, with some very high end 3rd party effects.
Third party development is where Yost can expand his product while keeping the
complexity of the software manageable with his small development group. Wish
list items not realized in this r4 version most likely have been reserved for
plug-in development. IMHO, r4 is the beginning of the demise of major 18 month
waits for new features. Now, the plugin concept will allow us to custom build
our animation workstation around our needs.
DPS-PAR The newest software release was demonstrated to be available in a
couple of weeks. I was very excited that everyone of my wish list items was
put into this release. There will be wild card support for file selection,
non-destructive join, GPI moved to the front panel with a variable trigger
rate. More features in the scripting of join and a nice time lapse video
feature fully implemented. There is also support for two IDE drives now and
they don't have to be identical. You can even use a slower drive for files not
requiring motion. Second there will be a sound/video sync software which will
allow the animation to be linked with a wav file in windows. Looks like I'll
have to install windows and a sound card on my PAR machine now!. The wave file
is addressed from your regular drive so that it can be played back without
interfereing with the performance of the video file. I asked about the quality
improvement for the near future and I was told that this still depends on the
development of faster IDE's. As they are released we can all expect to just
increase the BLF from the current limits with the Mic 2200 series. The PAR
hardware is currently capable of the lower compression ratios of the MAX and
targa 2000 but we PAR owners are currently at the mercy of the drive
limitation.
Speaking of the MAX: Matrox finally has it working and it really looked good.
If fact, I thought it looked as good as the PAR on animation. They didn't
demonstrate real life live video capture but instead were showing an animation
video captured and played back. As those of us who own the PAR have observed,
the grabbing of real live video generates more artifacts in textures such as
flesh tones and real life images than seems to show up in animation whether or
not it is computer live of video tape grabbed. I was disappointed that they
didn't show this in their demo. I didn't question whether the MAX is released
(shipping) yet nor whether the priceing has changed. At any rate, The MAX did
look quite good and I'm sure Matrox will sell a few to those that feel they
need to spend more to do compressed DDR recording. I spoke with Martin Foster
and Alan Iglesias and they felt that the MAX looked just a tad better than the
PAR. While it may have to them, I require a much more radical difference in
quality to justify the greater complexity of the MAX system. One thing was
certain, neither the MAX, nor the PAR, came close to the quality I saw on the
Abekas DDR.
Note: This is a demo wish list to DPS and Matrox: In your demos, please
refrain from using any animation renderings that are not field rendered. These
animations such as those from the toaster appear jerky and almost strobe like
which people will judge to be a quality of the hardware you're trying to sell
when it is indeed a function of the rendering process.
A new contender in the compressed DDR race is FAST's new Digital Player
Recorder. While this system is really intended for NLE editing it will also do
what the DDR's, including the Abekas, will do. It will expand the Fast Video
Machine to NLE editing with a user definable compression from NO COMPRESSION to
as high as 100:1 using the 4:2:2 digital vidio processing on two video
channels. The cost of the system (DPR) will run around $13,000 with 9G. of
storage, fully loaded for broadcast betacam SP quality. The DPR concept uses a
self contained SCSI port and splits the fields of the video frame between two
HD's to achieve the high through put for zero compression DDR. The product is
expected to ship in Europe in Sept. The demo was quite impressive. It would
seem to me that if one is considering an Abekas or Accom technology and also
needs to do conventional video editing the FAST system is a good bet. It is
the only editing system that can mix seamlessly conventional machine control,
time line linear, EDL, off line, On line and soon NLE and DDR technology in one
package. There are other devices in the works to further expand the FAST into
the high end broadcast arena but are receiving low priority right now due to an
all out push to get the DPR shipping by Sept.
I additionaly enjoyed my tour of all the high tech stuff and was entertained by
the VR demos and 3D televisions and a holographic work table using a VR glove
and set of glasses (no, not a helmet). Using a tool like a 3D wacom tablet one
could draw fully real time rendered objects in 3D and move around the object's
holographic image suspended above the work table.
Well, there was much more but the above is what excited me the most and I'm
sure you will get many more reports what's happening in the SIGGRAPH world.
The Adesk party was great because I got the opportunity to meet many of my
forum friends and was only disappointing in that some of you couldn't make it
or that I just missed you last night. We all missed you and try harder next
year to BE THERE!
I'm sorry I missed you at the Autodesk booth. They really went all out with
their set up. I was really dissappointed that I got caught up here in
Gainesville and wasn't able to make the party last night. I was able to meet
some ASOFTer's at the booth but I wish I had more faces to put to names.
Looking forward to LA in '95,
Andrew
Thanks for the report Don. Besides the most discussed R4 plug-ins (IK & Patch
Modeler), the other plug-ins as well as the core code changes seem to add a
tremendous amount to this release – although the "look" may remain the same.
(I'm strictly going by the press release & Gary's upload). Throwing out those
two modules for a moment, the addition of a powerful BASIC-like SCRIPTING
LANGUAGE with canned routines, function libraries, etc., FAST PREVIEW features,
POSTSCRIPT OUTPUT, PERSPECTIVE MATCHING, major SPEED OPTIMIZATIONS in the
Editor, more WINDOWS SUPPORT, FIELD VISUALIZATION in the Keyframer, and major
IPAS INTERFACE enhancements – Gee, that's a bunch of additional power *now*,
and the IPAS enhancements alone should produce some amazing power enhancements
in the future. Then there's the IK & Patch Modeling. Sounds like a strong
upgrade that's not so obvious on the surface.
Did you see any of this stuff in action, and if so – can you comment ?
Thanks,
BILL
<<Sounds like a strong upgrade that's not so obvious on the surface.>>
Please PMJI, but it seems to me that there has been one gross oversight in the
R4 package: 32bit operation to take advantage of the Pentium machines! They
are either planning this and haven't yet announced it, or they blew it!
Tell me its not true! Someone please get me a paper bag; I think I'm
hyperventilating.
Tyler,
My local vendore says that R3 was optimized for 486's, and the new version
will lbe P5 optimized. I certainly hope so!
-Ralph-
Tyler,
>> My local vendore says that R3 was optimized for 486's, and the new version
will lbe P5 optimized. <<
I don't know who your "local vendor" is .. but I havn't heard anything like
that.
Ang
>> My local vendore says that R3 was optimized for 486's, and the new version
will be P5 optimized. <<
"optimized" can be a pretty nebulous word… sort of like "intermediate". Your
vendor needs to qualify what he/she means by optimized. -dave
Dave,
>> Your vendor needs to qualify what he/she means by optimized.<<
I am a vendor (3DS dealer) <g>. That was what I was saying to whomever I
responded to. At this point, r4 is just as optimized for the Pentium as r3
was.
Ang
How's the club doing since AEC? -dave
3DS is a 32-bit application. Compiling for Pentium optimization would appear
*_after_* R4 is actually de-bugged & released, don't you think ? …and more
data input on the potential hardware problems. 'Course I'm not saying that
there ever will be optimized Pentium code, or in fact whether it is now – I
don't know.
BILL
Gary and Jack demoed the IK over and over because, I guess that was the most
inpressive. They demoed a human mesh figure walking up the stairs which would
be a fairly difficult thing to keyframe by hand. It seemed as easy as the demo
I got on the wavefront. He also demoed the IK with a robot arm to show how the
IK could be applied to machinery animation. Since I do this type of stuff
often I'm looking forward to it.
Even with all the enhancements, 3DSr4 is still lacking some features I'd hoped
for in the new version that definitly were on the wish list. Maybe it will be
an option plug-in. One of these was a refractive raytracer rendering option.
Thanks a lot Don. Do you happen to know:
1) With the IK system, could linkages be established which, when pulling on the
end of the chain (hand) it would not only pull/rotate the arm components, but
would bend the torso heirarchy at the waist as well (if pulling down) – in
other words, linked heirarchies ?
2) In your example, was the subject's foot constrained by the step object, or
CD employed ? Or did you still have to line-up the bottom of the foot with the
top of the step ?
3) Are skeletonal stand-ins used when manipulating the heirarchy, or can you
actually work with the wireframe (even RT shaded), as the press release implies
to me ?
4) Did you see any patch construction demo using the Patch Modeler ? (from the
PR text, it seems that you have to start out with a planar or cylindrical patch
matrix, rather than designing freeform with control curves & construction
functions – I hope not)
<< Even with all the enhancements, 3DSr4 is still lacking some features I'd
hoped for in the new version that definitly were on the wish list. >>
Yeah, I know what you mean. Well, I really *don't* know 'cause I wasn't there,
and I'm not sure if those who were there are aware of the functionality of all
the new stuff. But from what's at least *_implied_* from the PR, it seems that
all of the dead-serious spline modeling/animating/motion graphing tools I'd
hoped for are not present. Maybe the next release will incluse them, and be in
about 6 months this time ! <g>
Thanks,
BILL
Collision Detection is not incorporated with R4 but at the booth they showed
how the addition of scripting allowed for you to use BASIC-like commands to
write in the collision detection.
Andrew
I'm really curious about the announced Keyframe Scripting Language KXP Plug-In.
Do you happen to know if this might allow individuals (like me), with no 'C'
programming skills to use it to develop custom animated functions ? In other
words, is the language more plain English (with some 3D/Animation functions
knowledge), or do you still pretty much have to be an accomplished 'C'
programmer & mathematician to use it ? This could be very powerful if it's easy
enough to use / pick-up on.
Thanks,
BILL
They went over scripting for collision detection, and it was about four pages
of script and it didn't look to confusing. I think I am safe to assume that
you will be able to develop custom animation functions, granting that you know
enough about programming to write them.
The language is supposedly "BASIC-like" in structure. I am not familiar with
'C' either but when they scrolled through the script, it looked pretty straight
forward.
This is an excerpt from the preliminary technical specifications:
"- Compose command scripts using a BASIC-like
interpreter to move, rotate, scale, hide, and morph
objects. For example, define a new pivot point for an
object and rotate it with floating-point precision.
– Execute a script in the text editor and immediately
affect the contents of the keyframer.
– Employ numeric and string variables, math functions
array and structure commands for creating sophisticated
data driven animations.
– Use 3D Studio dialog boxes to prompt for user imput.
– Reuse scripts, share scripts and embed scripts into .3DS
files for easy transfer to other 3D Studio users."
This was taken straight formm the Feature and Spec sheet for R4 they gave out
at SigGraph. Hope this helped.
Andrew
Andrew,
What CIS software do you use to upload a response? (not create a new message)
I'm sure you didn't type that response online right? Thanks for the report on
the scripting I am also very interested in it. (:-)
Mark Yankee
I'm using wincim, and it didn't take me long off line to type up the info.
Glad I could help.
Andrew
Thanks Andrew. I really don't know much about programming *specifically*,
although I do understand the overall logic & had to do it ages ago in
Engineering School. Hope it uses Animation-specific syntax.
BILL
I was wondering if the scripting is of any use at all to anyone who
knows absolutely NOTHING about programing (like me)? I'm still not sure exactly
what "scripting" really means. Does it let you just write up in text what 3ds
is supposed to do, as opposed to just doing it manually in 3ds itself? what is
the big advantage of doing it this way (aside from having a set of commands or
motions that you can just assign to an object)?
???
David
I believe that the Basic-like commands should make it quite possible to do
quite a bit of keyframe scripting even by non-programmers like us, David.
Check Gary's list of features in his initial message, and you'll probably get a
good idea of the functionality of the scripting feature.
-Alan
>>Basic-like commands…
I don't even know what Basic IS. Is it a programing language like C?
(Is that the right term to use… programing language? And thats what "C" is,
right?)
I did reread the feature list again, but most of the stuff mentioned in the
scripting heading kinda went over my head.
… I'm hopeless…. I know…
David
First: Congrats on the whale animation. As a diver, having experienced close
encounters, I can say your animation was very realistic.
As far as the scripting language is concerned, I agree. My first experience
with programming was with Dartmouth Basic on a TTY and punch tape. I did OK in
it but never considered myself programmer material. Some of the code I've
since written has been looked at by my supervisors as great but by true
programmers with a snicker/giggle. What I would like to see is a programming
language that understands human illogical thought. The problem with most
programmers is that they require the code to be in some logical (mathematical
order) Most creative people don't think in any kind of logical order. The
best example of this is an ex-boss of mine after unpacking his first computer,
plugged it in and then stared at it for a moment. After at least 30 seconds of
silence he raised both hands as if conducting an orchestra and said, "COMPUTE!"
Well everybody laughed and now that I think about it that is the kind of
programming scripting language I would like to see. Just the thought of FOR…
NEXT loop statements and Boolean algebra in 3DS gives me a headache. But on
the other hand, to be fair, we first need the power tools capability and later
we make the power tools easier to use so that users like me can crank out cheap
animations for the cheap, low budget clients I deal with.
Yeah, David, Basic is a pretty – er – *basic* programming language!<g> I'm no
expert, but it seems to be characterized by commands that are relatively
english-like and easy to use. Operative word there is relatively. Using the
scripting language, you should be able to create, save, recall, sell, trade,
etc. keyframe data in a much more robust way that ever before. For instance,
you could define a pivot point in an object and, using the scripting language,
rotate it with floating-point accuracy. Complete control of move, rotate,
scale, hide, and morph keys will be included in the feature set. You'll even be
able to create collision-detection situations along with other data-driven
animations, which should be a great thing for forensic animators.
I'm sure that when we see the examples that will ship with the release, many
more applications will become obvious to all of us. Either BobB or JoeF called
this feature the "sleeper" of the release, because of it's potential power.
Awesome whale animation, by the way, David. I hope to see what else you're
cooking up!
-Alan
David,
Lightwave has a scripting language built-in and it is quite powerful. Gary has
already expressed the fact that he doesn't like non friendly scripting
features. Basic is a programming language and is fairly user friendly. What I
envision is something along the lines of Macros which are great for repetitive
tasks. If you've worked with Ami Pro maybe you are familiar with the concept.
There are lots of possibilities in this area. But I expect they will be an
added feature not a replacement for the approach you currently employ.
You should be able to save scripts. For instance, you could type in some text
in the shaper, call up the script and it would apply beveling, lofting and
mapping coordinates to your pre-defined specifications. So that once you ran
the script, you'd go into the the builder and there the letters would be all
ready made.
One of the best scripts I've seen on the Amiga is one a friend of mine created
for doing a boolean operation on a block of granite, where with a random
function it broke up the object into random pieces while still maintain the
objects original shape, then applying another script you can make the object
fly apart relative to the objects position. Rather than having to keyframe each
object.
It will really depend on how far they take the scripting as to what you will be
able to do but I think once its implemented, you'll wonder how you got along
without it.
-JE
Scripting is the same thing as programming except for the execution. The term
"script" generally denotes an _interpreted_ set of commands, whereas "program"
denotes a _compiled_ set of commands.
The commands are usually oriented in a list-like fashion with some looping
controls to repeat commands or branch to commands. "Interpreted" just means
the commands are read as-typed (usually textual format), interpreted into an
action by a shell program, and then the action is carried out by the shell.
"Compiled" means that the list of commands is translated into machine-level
code that stands on its own (or almost on its own) so that it executes very
quickly.
Scripts generally run slower than programs, but from what I saw, the scripting
demoed at Siggraph executed sufficiently fast.
To answer your question though, I think you will find scripting very handy for
some simple things that might take a lot of manual labor otherwise. You will
learn the scripting quickly if you start out by trying to do simple things, and
then building on what you know.
— James — MAP —
James, John, Don, Alan, and all who jumped in on this one….
Thanks. I will have to wait until I get R4 to grasp completely the
broader scope of scripting, and to see if it is something I would find useful,
or easy enough to use. Will you be able to control steps in the other modules,
such as the 3d editor, to deal with repetitive modelling tasks? Or is this just
a Keyframer/animation thing for objects?
David
PS. and thanks for the good words on my whale animation! I can't wait to get
my hands on some of the exciting things coming up. They should make things like
that MUCH easier/quicker to do. (There were 26 whale morph targets in that 50
second shot… close to a million verts! Now THAT was a pain…)
SCRIPT is a Keyframe scripting language, allowing you to do just about
anything a KXP can do. This was the most requested (and most needed) area for
a scripting ability.
– Phil
p.s. You're whale sequence was well worth the work – very impressive!
The scripting language is a "BASIC"-like interpreter. I'm a total
non-programming bozo, but will definitely take a crack at this. Last time I
learned any basic was aeons ago on a Radio Shack 64k TRS-80 <g>.
John Tissavary (LUNA cie)
John:
Another former TRS-80 guy ! My first animation work was done
on the wonderful "Micro Movie" package. How far we've come !
Great meeting you at the party. I may be back in the SF area
soon, perhaps a beer or two ?? Talk to you soon,
|-\\/
Most definitely let me know when you'll be in SF. I'd LOVE to get together and
have a beer and a chat. Siggraph gets a little overwhelming for me sometimes,
and I don't get to really hang with everybody like I want to.
cheers,
John Tissavary (LUNA cie)
/* The scripting language is a "BASIC"-like interpreter. I'm a total
non-programming bozo, but will definitely take a crack at this. Last time I
learned any basic was aeons ago on a Radio Shack 64k TRS-80 <g>. */
Wwwwwwwwwwwoooooooooooooooooooooooooooooooooohhhhhhhhhhhhhhhhhhhhhhhhhhhh! The
year Star Wars came out!!!!! 😀 -dave
May the Swartz be with you! <G>
John Tissavary (LUNA cie)
I've never had any programing exposure at all, nor have I really had
the interest. I have to admit that sitting there and entering numbers into a
script file would probably bore the pants off me (I assume most scripts would
be somewhat lengthy – more than one page is "legthy" to me!), so I'll just have
to wait and see if I would want to mess with it. I downloaded the POV raytracer
a couple of years ago, but when I found out that it was entirely text based and
you had to enter all the data manually, I said "forget this!" and wiped it off
my drive. I have to be able to see what I'm doing or I get bored and wander off
to something else.
David
I've never really had the patience to learn programming, either, but I would
like to invest some time in this script thingy. It would be nice to do effects
and other stuff that's not available – bring custom stuff to the repetoire.
John Tissavary (LUNA cie)
Just think, John: You can finally get some return on that (TRS-80) investment!
😉
— James — MAP —
Yeah, I feel great about that <deadpan g>
John Tissavary (LUNA cie)
1) YES
2) YES
3) Don't remember but I do recall seeing the skeleton, I think one could pick
it up at any joint to begin the "follow through"
4) There was a patch demo but I don't remember many of the details to give you
a tutorial at this time <BG>
I saw hundreds of demos in the two days and only wished to report the hilights
of my findings. If you need more details I guess you will have to get them
from Yost. The first two questions I could answer because these points were
emphasized in the demos.
Sorry for the barrage. Thanks for what you have been able to share. I just now
talked to a friend that just got back from SIGGRAPH, and he told me that
Lightwave was announced & shipping for the PC (& SGI & Raptor). I'm thinking
that this might have been instrumental in causing R4 to be shown at this time.
Anyway, thanks again.
BILL
Don,
As usual, very insightful observations. Thanks for sharing them.
-JE
Don,
The only thing missing from that report was the "wop-wop-wop" of the helicopter
blades in the background <G>. Thanks for the report.
Ang