#Particle FX
42 messages in this thread
To all who long for Special Effects in A.M.
Partical systems and most other similar effects, such as explode, rain,
disintegrate, etc. cannot be produced in a spline based program. They can only
be produced by a polygonal program. Even Wavefront, Alias and all the other
high end programs convert their spline and nurbs models to polygons before they
allow any partical effects.
Animation master can never do this unless they write a convert to
polygon module just for special effects. That would allow the writing of
external plug in programs. Spline patch data is not going to yield polygons for
special effects.
So…..the only solution is to have Hash Inc. add a polygonal
modeler to Animation Master….which pretty much defeats the purpose of
animating with splines. It will take a much bigger program to do this kind of
thing.
If someone can figure a way, it would be great. I don't think plugins
will be possible without doing this. Also the renderer will have to be able to
render polygonal files…which it won't do now.
Regards, Sanford
Well… would it be possible to modify the effects to work on the Spline patch
data??? May take a little extra work, but it must be possible…
Grady III
Grady,
As far as my studies have revealed, the spline patch is a mathematical
expression of a space. I don't know the real theory behind its construction. I
only know that in every software package where dynamic particle effects are
available, spline patches are not used. In all cased the objects are polygonal
surfaces, or are spline or nurbs surfaces that were converted to polygons.
Polygons can be detatched from each other and still retain their integrity as
objects. These objects can then be further modified and scaled to produce
various effects. A spline patch ceases to exist when you remove a spline. This
could affect many patches at once if it is a large object with a single spline
running all the way around it.
Maybe someone like Martin Hash has a better answer.
Sanford
Grady,
>>Polygons can be detatched from each other and still retain their integrity
as objects. <<
So why couldn't pathes be detached from each other? The edge that is common
would have to be duplicated, but it seems like that is what also has to happen
with polygons, a polygon must have an edge, so must a patch.
It would seem that the logical progression of coding would be to first add the
ability to have sculpture automatically subdivide patches, once that solution
is worked out, have it double up the splines on each edge of the patch, now you
have seperate objects, which could them be scaled/rotated/altered
algorythimically.
Dealing with polygons is just easier in once sense, because all of your edges
are straight, while patches allow curves, so more math is needed. I don't think
its impossible, and can't figure out why you say it is, UNLESS you know that
Martin Hash is the kind of guy who will get his fur up when someone says that
something is impossible, and stays up for days at a time, until he figures it
out! <G> If THAT is your intention, great, if not, why be so condemming on such
on a great product.
Where's John Dowdell when we need him???? 🙂
Jim Jonah // not trying to war, just defending a great product!//
Jim,
I love Animation Master. In fact I am now selling models made in AM to
guys who run Alias because AM is a better and much faster modeler. But,
Impossible is a big word. If we all worked at Hash, we mi might know more about
why they have not implemented any kind of particle systems. As another forum
cruiser said recently…I am switching to Lightwave…Well, I'm not switching,
I already have Lightwave….It is a polygon modler and has lots of great
effects. Martin, lets do lunch, OK? There some ideas we need to discuss.
Jim, I like very much the idea of adding a module to convert to
polygons after the animation is finished. That would also allow Phong
rendering. Then I could get some work done. As it is, I don't have time to wait
for raytracing.
Sigh…I wish we had phong, but without polygons..I don't know.
Sanford
I was talking about this to thread to Joe Krause, a friend of mine who has had
Lightwave for some time now. He says Lightwave didn't always have the neat
special effects like lense flares, etc. The stuff had to be developed first.
What
you have with Lightwave is the fact that it came bundled with the Toaster and
the Toaster has so many users that it made sense for companies to develop
programs with other features for those users. The more A:M users we get, the
more
sense it will make for developers to spend time and money on solving some of
these problems. I wouldn't say anything is impossible. I never thought it would
be this easy to run a computer program until Windows came along.
The guy who developed Sparks, a particle effects package for Lightwave, spent
months breaking into the various routines of Lightwave to get it to work.
Guys like Dale Myers are super interested in A:M because of the speed, ease and
low cost involved in developing characters with a spline based program. They
are tired of polygon based modelers time and other problems. They've made good
characters, such as his Robo Jr., which could fly as a Saturday morning series.
ReBoot has primed the pump, I think, for a surge in interest for programs that
can
easily do character animation.
Kevin,
Just curious since I have never used a polygon-based modeler for 3D
animation. When you say some of these guys are tired of the time it takes to
create characters in a polygon modeler, are you refering to specifically to
characters like human forms and animals, or does this include inanimate objects
like cars, buildings, and spaceships. I have found that AM's lack of CSG
functions (in some cases) turns a relatively simple 2-3 hour modeling job into
a 8-10 day job. I am a very detail-oriented modeler, an although I love AM, the
lack of CSG and some other modeling functions, added to the lack of F/X, I
forcing me to look at switching to Lightwave when the Windows version is
released.
Until that time, however, I am continuing to create detailed models in AM
and posting images here and in other forums, to let people see the detail that
can be achieved with AM in the hopes of attracting as many users as possible to
encourage 3rd party developers to consider creating F/X for AM.
Michael S. Flynn
The time it takes to make believable, moving characters with polygons is very
time
intensive from what they tell me…this includes human, animal, alien and
robotic
characters with personalities. Seems many "cheat" and use models you can buy
for objects like cars, furniture, etc. just to save time.
Also, I believe they've encountered movement problems in Lightwave, but
someone said they are getting IK. I'm not completely sure about that, though. I
didn't know that Lightwave didn't have IK.
Kevin
Kevin,
If this is true for the 3D applications in the $1000 or less range, maybe
it isn't worth it to switch from AM.
Michael S. Flynn
I wouldn't switch, but I wouldn't ignore what other programs have to offer. The
guys I've run into here that are doing this for a living and usually in some
video
production or industrial application, use many programs for different things.
No one
program gets an overall endorsement from them. If Lightwave hadn't been bundled
with all those Video Toasters, who could say what program we'd be talking about
at the Animation SIGs here in Detroit.
After all, these are Amiga owners and Playmation (I think it was called
Animation
Apprentice then) debuted on the Amiga…so did Imagine and I guess a few
others.
I bought a PC so that's where my interests are at this point. They are thinking
of
going the PC route if nobody of strength picks up the Amiga chipset.
If your money is tight, I'd wait and see what the new version of A:M has. Steve
has dropped hints and so has Martin in his "Minutes".
All I know is what some people have said after their jaws dropped while
watching
the Playmation video. They are amazed at the ease and the power of the program
and how much quicker Windows makes things with multi-tasking.
Kevin
P.S. I can't remember all the hints Steve dropped, but Martin has been working
on
an interface allowing access to all areas more easy. Let's hope they've been
working on some effects. I know Martin stated if there was a strong desire
among
us users for some plug-ins, we should call them and tell them which ones we
wanted. I have Kai Power Tools on order, if it works for me, I'll be calling!
Kevin,
My main requests are for CSG functions in the modeler and effects. So far,
it seems the new interface is getting the attention, which, I guess, is
necessary to attract new users and maybe get back some old ones. If you are
getting Kai's Power Tools, I would suggest getting something like Adobe
Premeire if you don't have it. KPT plugs work with Premeire and are great for
adding some effects while doing compositing and sound.
Michael S. Flynn
What do you need CSG for, that you cannot do with splines???
Grady III
Grady,
It's not so much that you can't do certain things with splines, it's that
CSG would make modeling much faster. I am currently working creating very
detailed models to send to LucasFilm and LucasArts and creating Star Wars
models is extremely time consuming.
Just as an example, take the Death Star. I haven't started working on it
yet, but for this example, it will work because it is very basic as far as
shape. The basic shape is a sphere. Outside of adding the surface detail, the
only really distinguishing feature is the scooped-out dish section of the
planet-destroying laser. You can't just place a circle segment for the rim
because the segment won't conform to the sphere's surface area. There is no
fast way of doing this. It would take hours to manually set each point, going
in and out of RENDER until you have a smooth surface. You would also have to
have a specially designed sphere for the Death Star so that splines connected
from the sphere to the rim all matched up. Once you got that done, you would
need to re-orient the pivot to the center of the rim so that you could Lathe
the dish. If there were CSG functions, this whole thing could be done in a
matter of minutes by creating a spehere for the Death Star and a smaller sphere
for the dish. Place the dish sphere so that it intersects the Death Star at the
desired point and choose the CSG function Difference. Viola! And one of the
other benefits of CSG is that in most cases you would need far fewer control
points in objects.
Take one of the new TIE Fighters for example. The hull is comprised of 12
segments. There is space between each segment. The Front, Back, Top and Bottom
all need sliced off at a certain point. With CSG, you create a sphere for the
hull. Next, create an arced segment for the space between each hull segment.
Place it in an intesecting position on the hull. Clone it and rotate it 30
degrees and repeat the function until you have 12 arced segemnts encircling the
hull. Group the arced segemnts into one object. Then use Difference and you
would have a hull with 12 segments with space between each segment. Next,
create a plane and use the Slice function to cut of the top, bottom, front and
back. If you look at the hatch of the new TIEs, it has slots in the top. Creat
your arc and Lathe it for the Hatch. Create rectangles and extrude them to the
desired width of the slots and then place them so they intersect the hatch,
group them into one object and then use Difference. For the wing struts and
wing frames, create each piece separately, and then use Union to combine them
into one object.
The only real way to see where I'm coming from is to try and create these
models yourself. If you or anyone else can come up with a faster or better way
to create these models at the high detail level, it would be a big help to me.
I am about finished with the TIE Interceptor – it should be done today.
After I test render it and am satisfied, I will be uploading the project files
for it and the TIE Fighter.
Micahel S. Flynn
Michael,
If I remember, Steve said booleans were coming!
Jim Jonah
Jim,
That will be very cool!
Michael S. Flynn
You Bet! Booleans are coming with version 3!
Steve, Hash Inc.
Kevin,
I am involved in promotion the use of A.M.'s modeler on the SGI
platform as well as for 3D Studio. I like most of the program and I agree that
it is great for character development and animation. But we both know it needs
more development before it gains wide acceptance for serious production. Having
"most" of the features needed doesn't cut it in the marketplace. I see the
current version appealing to companies that traditionally did cell animation.
The problem is that there are great 2 d programs like Cambridge, Linker, and
Softimage that do everything an animation house needs. Where is A.M.'s Post
Processor? Why can't photoshop read a A.M. targa file? This list of problems
goes on.
I am using A.M. because it is the best modeler….hands down. The rest
of the program need more work before I can make money with it.
Regards, Sanford
Sanford,
>>Having "most" of the features needed doesn't cut it in the marketplace.<<
Oh really??!! If that was the case, NO program would ever be sold! 🙂 Come on,
do painters only use one brush, mechanics only one wrench, page layout people
only one program? 🙂
A friend of mine has Alias PowerAnimator, guess what, he uses photoshop to
create textures, Premiere for editing the video (adding transitions), Cosa
After Effects for alot of compositing AND, believe it or not, some of the
special effects that you (and I <G>) wish were in A:M!
His, and my, strategy is to do as little as possible in the 3D program and use
blue screening and compositing as much as possible i.e. similar to cell
animation techiniques, usually it's faster!
Jim Jonah
Jim
<most of the features>
I have been in on many software purchase discussions here in Hollywood
and I must admit that many people are intrigued by A. Master, but are not
willing to try doing production with it at this time. I know of a number of
companies who have it on Pentiums. They have tried it and won't use it. I can't
explain it except to say that their clients are demanding Alias file
compatibility.
I just finished a Gopher in A.M. for a guy who imediately converted it
to DXF and sold it to his client as an Alias model. There just hasn't been
enough good work done or published to give A.M. any real momentum in the
professional marketplace.
The only place I have found it being used is in
schools.
Those were UCLA, and the classes taught by the Animator's
Union in North Hollywood. Do you know of a major company who is using
it for regular cartoon production as their main software?
Sanford
Sanford,
>>Do you know of a major company who is using it for regular cartoon
production as their main software?<<
How many major companies producing 3D cartoons are their? Most cartoons I see
are still 2D, hopefully REBOOT will change that! Considering that the only
company I know producing a regular 3D is the one producing REBOOT I'm not the
person to answer that.
All I am pointing out is that A:M is a great product (IMHO), that NO product I
use does everything I want it to, AND to not condemn a product because it
doesn't have every trick in the book. If I could afford Alias and was doing 3D
all day every day, Alias would be one of the many tools in my toolkit.
Sometimes I have to use/buy a product I don't like because the client demands
it (Autocad in my case for a project), but that doesn't stop me for rooting for
the Product I do like, hoping that by supporting the product that enough users
will buy the product, generating enough revenue to hire more programmers so
that we can all have the tricks and toys in the program that we want!! <G>
Jim JOnah
Jim,
Hopefully we can both support it. I'm becomming a dealer if I can. I am
also writing a how to book.
I just did a flic file in A.M….29 frames, 11 hours, 34 minutes, "LOW"
resolution, 320 X 200…40 minutes and 09 seconds to render frame #29. I will
now load the fil [Ce as a DXF in 3D studio, and render it with the same
specifications. I used shadows (photo quality) and anti-alias. I will upload
the results.
Sanford PS this is the CITY file without the
figures..just the buildings.
I can't say too much, but I can say that GRAFX will soon be another company
producing a Saturday morning cartoon all in 3D, but using 3D Studio with
Puppeteer. We are also working in 2D using 3D studion for the current cartoon,
Where in the World is Carmen San Diego?
Sanford,
No argument from me on what A:M needs. I like it specifically for the character
animation abilities. I just wish more people owned it so Hash could do more and
there would be more incentive for additional add-on programs like the font
program
I bought from Branded Research. Money talks…
Kevin
Kevin,
I am going out there to try and sell A.Master. I have faith in it. I
use the program daily, and am very happy with the sculpture module. Many will
buy it from me after they see the models you can make. I have great faith that
things will improve for all very soon.
Sanford
Kevin,
Did you see Reboot Saturday? It's great!
I did notice a couple of this about it though:
1. No shadows
2. No particles. Even though things blew up.
3. Things in the distance were just as clear as things up close. No "fogging"
or "hazing" of objects.
I'm sure the reason were for rending speed, which makes me smile, in a sick
way <G>, that even the guys with the big bucks using the SGIs (softimage) have
to cut corners. Makes me appreciate A:Ms speed on the PowerMacs even more!!
I still would love to see a tape of Robo Jr.
BTW, did Steve get some demo tapes to you to pass out at the meeting?
I'll see ya!!!
Jim Jonah
Yes, I saw ReBoot…it was a re-run of an earlier episode…they're probably
padding to make more time for a couple more episodes to be done. I had a
message from Chris Wellman at ReBoot Productions in alt.3d on the Internet a
few
weeks back and he said they were working concurrently on episodes 5 & 6.
They hadn't expected ABC to go with them as soon as they did and they took
late delivery on some equipment last spring.
They are trying to save time in rendering and we all know shadows take time. I
can't fault them that. They are doing a marvelous job with what they have and
are pioneers since they have the first regularly scheduled computer animated
show.
I haven't heard from Steve yet…perhaps they are getting ready for Comdex. He
has my addresses if he wants to send them. I'll be glad to pass them out.
Kevin
Sanford,
>>That would also allow Phong rendering. Then I could get some work done. As
it is, I don't have time to wait for raytracing.<<
The phong render I have seen doesn't have shadows, which can improve the
rendering time, so why not just turn on the anti-alias button, and turn off the
Photo button?
The longest I've had to wait for a frame was 5 minutes, some of my projects
have been less than a minuter per frame, what kind of times are you getting
(granted it really is dependant on the Machine, RAM, textures, Model
complexities, etc.)
Jim Jonah
Jim,
I have a 496 66 with 20 meg ram and two 500 meg hard drives. Windows
for Workgroups. I am rendering a flythrough of a a city in D1 Photo,
anit-aliased quality. I have twenty figures in one scene using figure
substitution. I tried rendering 30 frames, and after 18 hours, I shut the
machine off and went back to 3D studio. That program took 28 minutes per frame
at 756 X 486, but then I decided to dump the project for now. The 3D S file I
ended up with was 2.85 megs. The glass textures don't look good in 3D Studio,
but it is much faster than A.M.
I will prepare a bench test of the scene in both programs and upload
the results.
Sanford
re:" Where's John Dowdell when we need him???? :)"
Forum Drift! :>
John is one of those great company tech support guys (like Steve here) who's
_really_ into doing cool things with the product.
I'm working on a project that uses both companies' products, of course…
Gavin.
Sanford,
I am not a programmer, and I don't know if you are either, but it seems to
me that the only reason there aren't particle systems for spline models is
because no one has actually expended the effort to create one. I don't know
about the beginnings of computer animation, but I would assume that it was all
polygons in the beginning. And people who wrote effects for early animation
started by writing stuff for polygon models. The demand for the future will be
curve-based modeling, and someone or some group will find a way to do it.
One way that occurs to me to make a spline-based sphere explode would be
to write an algorithm that calculates the surface, uses some sort of fractal
mathematics to break the surface up into randomly shaped patches, and then use
something like skinning, only in a reverse fashion, to disconnect each patch to
become a separate particle to be animated.
Have you ever wondered what type of air travel we would have today if
there never evolved any animal, bird, or insect that had wings and could fly?
No one would have seen a bird and admired the way it could soar and glide and
wished or wondered that there was a way for man to do it also. But they would
have seen leaves blow in the wind, or kicked a dandelion and watched a summer
breeze carry each piece up, and away.
There is a reason why programming languages are called languages. Look at
the English language. Extremely versatile, you can use to make almost any
thought understood by almost anyone. It has certain limitations just like
programming languages do. I look at Turbo C++ as a rather young, growing,
language that has not been explored fully. But it will be. I'm pretty sure that
when someone was wishing for a curve-based modeler years ago, that the polygon
experts were the first to say impossible. Maybe it isn't a good idea to ask the
polygon effects creators to create spline-based effects. Their way of thinking
may be too closed to trying something that they already think is impossible.
I have a friend who works for Precision Cast Parts, and a while back he
was telling me about a situation where some people wanted to see if an
automobile could be designed that could meet certain specs such as 100 mpg, and
some other things I don't remember. The point is, is that naturally went
straight to the automobile designers who said "impossible". That was expected.
Then they went to a different group of engineers with the same challenge. No,
they didn't know a thing about automobile design, but they went for it. And
succeded. Who were these engineers? Believe it or not, they were hydraulics
experts. They even designed a set of brakes that stored inertia as kinetic
energy that was released after a stop starting the car moving at a reasonable
without using the gas pedal at all.
I'm no expert, but I believe that spline-based effects are certainly
possible. All it is going to take is an individual or group of dedicated
programmers to make it a reality.
Michael S. Flynn
Michael,
>>>it seems to me that the only reason there aren't particle systems for spline
models is because no one has actually expended the effort to create one.
I am sure you are right on this. The real question is: will there be somebody
who will do it? This is how I read Stanford's message — We can't really use
the polygon based methods and to reinvent one for A:M world is too much to ask
for…
The real problem we have (as A:M users) is that ALL WE HAVE is Hash, Inc. and
their hands are full. (Of course, we love them for what they have created for
us).
I am sure Stanford knows that anything can be done — It's just that the
economics that determine what _will_ be done…
Pawel
Pawel,
The question of wether or not somebody will create a particle system for
spline models is a serious one, at this point. It does not matter how great AM
is, if it can't do the things people see on TV and the movies, it will remain
in a niche that may be impossible to climb out of. Some of the things I need,
and probably others, aren't just particle effects. Lens and flare effects are
becoming extremly important. Collision detection should not be a real problem
since I think it works by knowing surface areas, both of which have to be
calculated for either polygons or splines.
Another thing that should be possible is using ACTION to morph entire
figures of the same class, including any movements, maps, etc., as long as the
figure you are morphing to has an equal or less number of control points. We
should also be able to morph the MAG, ALPHA, and GAMMA of control points.
It seems that we are in somewhat of a Catch-22 – we could have many more
features if there were a lot more users to create revenue, but you can't
attract users without the features.
Myself, I am going to wait and see what AM 3.0 is like. Then I will be
looking at switching to Real 3D for Windows – $1,300, or the upcoming Lightwave
3.5 for Windows – $950. With the price of upgrading to AM 3.0 being about the
same as going the competitive upgrade route to Lightwave, AM will not be the
only application losing users. Unfortunately, I am currently in the unemployed
mode. If I can get back to work before the end of the year, my tax refund
should allow me to purchase Lightwave, while keeping AM, but to be honest, I've
already made up my mind about 80% in favor of switching to Lightwave.
Michael S. Flynn
re:"The real problem we have (as A:M users) is that ALL WE HAVE is Hash, Inc.
and their hands are full. (Of course, we love them for what they have created
for us)."
Actually, Vance Gloster (vance@netcom.com) has created a 3rd-party product for
procedurally constructing complex characters such as buildings. You create one
character, and then can duplicate it an arbitrary number of times along various
paths. Very, very cool for non-organic stuff. Saves zillions of hours.
Gavin.
Gavin,
Do you have a phone number for Vance, is his product for sale? It sounds
cool!!
Jim Jonah
Vance Gloster: (818) 574-3067 or from compuserve mail: >INTERNET:
vance@netcom.com.
Gavin.
> Vance Gloster (vance@netcom.com) has created a 3rd-party product for
procedurally constructing complex characters such as buildings. <
Sounds cool. Who do i contact about it?
..mark
Michael,
I agree with your hopeful ideas. If you could re-process a spline patch
on the fly as the effect were applied, then you possibly still could use the
spline and morph animation tools to animate the model. After all, the
conversion to particles need only take place in the renderer. A good idea.
Sanford
Sanford,
Just curious, why do you say that the effects "cannot" be produced in a sline
based program, are you a programmer who has tried programming a spline based
system? You sound pretty sure of yourself on this, what's your basis for this
statement?
Jim Jonah
Jim,
No, I am not a programmer. I am an engineer and have been a computer
and circuit board designer. My conclusions are observations of every software
package that has emerged that did graphics since 1978, when I started trying to
draw pictures on my first Tandy. I have been dissappointed that I could not
find any effects done directly with spline patch models. The problem seems to
arrise from the inability to break apart the patches patches along a spline and
maintain their independent integrity. In all polygonal programs you can do this
easily. Have you ever seen a spline patch that could be turned into particles?
If so, what program?
Regards, Sanford
Stanford,
Are you sure that the Alias system converts to polygon first? If it does, then
why couldn't A:M be altered to do the same thing as well, if they run into a
wall trying to do it just with splines?
It is just a matter of priorities IMHO, enough people complained about then
interface, which really doesn't bother me that much, so time and money is being
spent on reworking the interface, as well as adding other features.
Personally, every program I have every learned has some quirks in its
interface, so I would rather have had all resourses applied to new features
instead! Oh well.
Jim Jonah
Jim,
I have been running Alias for some time (up to yesterday), and it waits
to convert the wireframe model to polygons until render time. At that time you
give it a "Save SDL File" command, it goes off somewhere and makes splines into
polygons, and then renders.
You can "quick render" the patch models all through the animation
process, but the final images are based on polygon versions. The Shader module
of Alias is part of their Render module. That is where all the effects are
applied. Thay can only be applied to the SDL file, which can be edited by
itself in a text editor. It is similar to a Renderman RIB file, which describes
how normals are oriented on poligonal surfaces, and how light hits them. If
Animation Master could write a RIB file, we could have the best of all worlds,
but a very complicated one as well. Renderman does allow many effects to be
applied, and can render any file converted to their format.
Well, I hope Martin is looking at this thread. We do need effects to
keep Animation Master afloat into the 2000's.
Sanford