PDA

View Full Version : Character Animation Feature Requests



madrenderman
10-29-2003, 07:41 PM
first i must tell that you can do character animation with lw, but is not a simple task.

in lw there are a lot of interesting tools, many good ideas, but developed for half or less.
some example?

endomorph : good idea, better than other software, which must use a lot of obj copies, but.... try to use it on speaking char
you can't. morph mixer mix pose to pose in additive mode, this mean that you deform not linearly the point, but you add value to value modified, this mean is very difficult to transofrm a good phoneme to another, be cuase in 80 case on 100 you see the lips of character deformed before to have a correct final shape.

another trouble is the miss of no linear transformation, which create a lot of trouble during special task like complex cartoon eyes.

sock monkey : a bad compy of old puppet master (by fory oruwa, lw metanurbs creator, messiah developer, plugin developer, moprh gizmo developer).
why not work correctly?
it not is additive, this mean you cannot mix it with other deform plugin like morph mixer, if move some point group, it reset every kind of displacement before applied.

spline deform good tool but is not simple to control the mesh by it, you cannot create a spline to deform a face (like messiah from that is copied), and every node cannot be controlled by motion plugin or expression

a japanese guys develope a more flexible, fast and powerful spline deform plugin, and it' sfree, guys of NT check and hire this guy!

bone and weightmap, subpatch nice tools but simpel and easly to dominate. you cannot paint weightmap on layout, and in modeler you can test effects of weight and bone only with vertex paint, but is not enought smart, and if you use subdivision, the different level of subdivision can change a lot the result of final deformation.

control of bone influence is complex, and rich, bit not smart, try to use maya bone setup and can ind a more fast and confortable weght paint system.

i hope that now with ortho tools will be more simple to work in layout, but a merging on one packages is the real solution

speed of playing there aren't a way to have a correctly speed of scene playing, this mean we can't understand correctly the speed of animation, to check the real final animation time.

deofrmation cage to optimize the deformation on obj, using of endomoprh linked to bone by expression, or plugin like join morph.



gimbal lock
without quaternion system of rotation, we must continue change from a coordinate system o another to avoid gimbal lock, or we need to add tons of rotazion bone. too primitive solution, please rewrite it.

optimize background image loading
with or without previw in background, the task is slow, optimize routines to load, compress and use sequence in memory when we must work with long background sequence.


editing of rest and basic motion without pain like now.


and this is only the start of problems...
anyone?

p.s. i animate with lw from 3.5 on amiga computer, but i see that often is not the best solution

hrgiger
10-29-2003, 10:31 PM
Character animation isn't easy in any package but I would agree that Lightwave could use improvements in the character animation department. Fortunately Lightwave 8 looks like there are many improvements.

Just a few of your statements I'd like to comment on;

Gimbal lock: As far as I'm aware, Lightwave is not the only program to suffer the problem of gimbal lock. You don't need to add tons of bones to avoid it, you need to add one to control your heading. Nesting your controls like this will avoid gimbal lock altogether.

Speed of playing: You can render a preview of your motions before you render your animation. So you can see what your final motions look like before you waste time rendering a bad animation. You can even change the fps on the preview as it's playing.

wacom
10-29-2003, 11:15 PM
I haven't experienced any gimble lock in my stuff since I went through this tutorial:

http://www.newtek.com/products/lightwave/contests/dec-02/winners/tutorials/layout/Chris_Olsen/index.html


Have you tried it?

EyesClosed
10-30-2003, 01:59 AM
LightWave lacks even the most basic animation features like a dope sheet, ghosting, connectable attributes, good constraints, etc.

There seems to be some progress with LightWave 8, but still looks like there's a far way to go.

Nemoid
10-30-2003, 03:58 AM
Good, Madrenderman you stressed in a clever way the huge probs of Lw about CA!!
every app have its weaknesses, but since we are in 2003, we want to see many of these probs solved.

the fact is that Nt didn't implement in a clever way things like sock monkey and other CA stuff in the past.

I really hope this thing is starting to change with Lw[8] even if I think this is only a(probably good) start and it will need some time to face and fix these kinda probs.

obviosusly introducing tools like ortho and other free plugs, even if very good isn't enough if they don't fix probs at the base.

lunarcamel
10-30-2003, 08:39 AM
I feel this pain ;(

I rarely animate in LW because I just don't think it's up to speed with some other packages. I doubt the 8 upgrade will wow me enough to change my mind either. NT prove me wrong ;)

Noclar7
10-30-2003, 09:22 AM
I'm pretty certain you will be wowed for character animation in 8. Most of the focus on enhancements for LW8 was on animation, So much in fact that some people were even complaining that there was too much effort put on this one area. I've been waiting for some good character tools as well, and I'm betting that when my electronic download is ready, it will deliver.

cagey5
10-30-2003, 12:08 PM
Speed of Playing there aren't a way to have a correctly speed of scene playing, this mean we can't understand correctly the speed of animation, to check the real final animation time.

What about clicking on 'play at exact rate' under general options or am I missing something?

Nemoid
10-30-2003, 12:23 PM
Originally posted by Noclar7
I'm pretty certain you will be wowed for character animation in 8. Most of the focus on enhancements for LW8 was on animation, So much in fact that some people were even complaining that there was too much effort put on this one area. I've been waiting for some good character tools as well, and I'm betting that when my electronic download is ready, it will deliver.

I hope so, but sometimes, when I'm very pessimistic., I think that Nt will put into [8] the greatest plugs, like Ortho tools, maybe Ikeda tools, the plug for undos in layout, and stuff like that. integrate them in some crap way, and label this like Lw[8].:confused:

actually, I hope they are working good as I'm thinking, especially now that they own all the code as they say, but its true that most of the probs Madrenderman stressed exist from a long long time.

most probs are at the base.

however, CA is the main area to work up, as well as rendering engine.

so go, Nt , make our jaws drop!! :)

hrgiger
10-30-2003, 09:50 PM
Originally posted by Nemoid
I hope so, but sometimes, when I'm very pessimistic., I think that Nt will put into [8] the greatest plugs, like Ortho tools, maybe Ikeda tools, the plug for undos in layout, and stuff like that. integrate them in some crap way, and label this like Lw[8].:confused:



It goes beyond plugs. Perhaps you're forgetting animatable points, improved softbodies, rigidbodies, bone dynamics, faster fk and ik, instant ik, etc....

Nemoid
10-31-2003, 03:33 AM
I'm not forgetting all this stuff. that's why most of times i feel positive about Lw[8]
only, I'd like to see tools implemented in a clever and complete way. i'm not blind.
since in the past in some case things were not implemented in a deeper way
e.g. sock monkey, the Viper itself. (its rendering is not so faithfull with the final render, as it happens in Maya) and other tools, that's why sometimes i am worried.

another example could be releases like 7.5 b and partially c wich demonstrate not so much cleverness, since they have many bugs.

so, in one hand I am confident that Nt is working hard for[8] on the other hand sometimes I am worried.

Dodgy
10-31-2003, 04:01 AM
I think that although a lot of the improvements for 8 are bought in plugins, this is merely a stop gap measure to give the new team some time to reorganize and start looking at a lot of the bigger concerns of Lightwave. We can see they are already putting in some new stuff, which is welcome, and I can see them looking to improve the system as it stands.

hrgiger
10-31-2003, 04:20 AM
What could be more clever then right-clicking to instantly setup an ik chain?
I wasn't suggesting you were blind Nemoid, just pointing out that I think LW8 is a leap forward in CA features. There is still much room to grow though....
I myself only have two bones to pick(pun definately intended) as far as things I'd like to see fixed in Lightwave as far as Character animation goes. These are the only two things that I think absolutely need addressed, the rest is all candy to me...

1) Interaction speed: I'd like animating in layout to come close to real time. Many of the rigs I've used in Layout can slow to a crawl depending on the setup, not to mention using more then one or two in a scene. Let's see some real time animation like in Motionbuilder. Even using a low bounding box threshold does not always speed up bone rotations and is quite hard to work with sometimes especially when trying to accurately have two characters interact physically.

2) Precise Deformations: I really wish there was a way to correct deformations directly in Layout. I'm hoping for good things with having animatable points in Layout (a long over due feature there...) however, from everything I've heard about it, I don't think it's a total solution quite yet.

Non-linear endomorphs would be nice and I think would be easy to implement. I think all Newtek would have to do is to place a 0-100 slider in modeler that is represenetive of the slider you have in Layout so that when you create an endomorph in modeler, you base is at 0, your extreme is at 100. Then you can put your points at anywhere inbetween those numbers to get a non-linear deform.

madrenderman
10-31-2003, 08:12 AM
like i tell before, i use lw from 3.5 on amiga, and i have a deeep knowledge of lw and its engine(i love it from many years), but this not mean i0m blind on its lack.

1) speed of playing is important to avoid a continue preview in ram, and if an old plugin like first messiah which come from lw sdk, can be more fast than actual lw, i think lw must be otpimized.

2) opengl not are enought modern, many lack and limits to use a modern videocard, i have a 2000$ video card and lw are not enought fast like other 3d apps.

naturally when i tell of CA optimizing i mean in correct setup of scene, with complex character hided, disable, and working on visible proxi obj at very low resolution, under 300 poly.
but if i start to have many character, lw be slow (on 3dboxx dual xeon 2,9 ghz, where maya and messiah work well.

many useful plugin for CA not work in realtime (updated only if there is time event, not a parameter event), but you must do Autokeyframe active, or (very bad) they are updated only when you scrub in time line, this mean most of settings must be done by expression (more efficient solution).

play at exact rate is a solution only for a too fast scene, but not for too slow scene where frame are skipped to play at exact rate, and you cannot understand correctly the time of action.

gimbal lock is a pain on complex character, i know two or three tips to avoid it, but are annoying work add, which many other software avoid.

naturally i try read and learn everything i see on net, or on many other media, but i olso work with maya, messiah, and many of this problem are lost....

i hope this simple list is useful to help newtek to do better.

pauland
10-31-2003, 09:50 AM
Considering who your host is, I think you might have chosen a better thread title, such as 'LW and animation'.

Whatever the merits of what you have to say, the thread title is rather impolite.

Paul

UnCommonGrafx
10-31-2003, 10:33 AM
When I first saw this thread I thought there might be something of interest here. My, how I was wrong.

If you have MotionBuilder, why complain? Use your tools and make money.

All of what I've read here has been mentioned ad nauseum over the past few years and NT set about addressing such concerns. LW8 isn't even available yet and the complaining about it seems louder than ever!

If you have tools for the job, use them. If not, create them within the environment you can.

Tools is tools. An artist complaining about the tools is a sad state of affairs, to say the least. "Damned this brush; it just doesn't make the strokes like I want!"

Why don't we all just wait until 8 comes out to make more informed opinion than the mad supposition we have here.

Nemoid
11-01-2003, 08:56 AM
I can't agree completely with you.Uncommongraphics.

first of all, its obvious that one uses the tools at his disposal, but this doen't justify huge lacks of an app.
there are persons wich can't afford nothing than one software and choosed Lw for some good reason, but the lacks, present since years don't allow him/her to work efficiently as in other apps, especially for animation.

ok,[8] isn't here yet, but I think there's room for some request and highlighting of problems, even because the new staff can have many ideas to work on. if not, this entire forum for feature request have no reason to exist.

can seem boring repeating things over and over, but since the probs stressed from Madrenderman and many others are real and IMO important I think it can be good.

MB is a great app but I'd like more to work entirely in Lw for example . learning to use one app is better than learning several apps indeed, and the workflow increases its speed.

there are many cases of mixed pipelines to get the job done, but IMO a very good app allows you to work very efficiently.

when i say this, i'm thinking especially to a solo artist, wich can't afford no other than Lw, but wants to make good animations in a decent time. since Lw has many strenghts I think facing these probs would increase its power more and more.

So, I think that animation issues are the ones in wich Nt MUST focus in a deep way as well as UI workflow and enhancements, especially in editors ( look to MB and you can understand what i mean : flexibility, spedness, power.).

I really hope Nt is working into this direction.

pauland
11-01-2003, 09:09 AM
I'm sure that Newtek will be waiting on the words of a Sage to guide them, particularly with less than 60 days to go before release. How they must wish these posts had been made before Siggraph.

Nemoid
11-01-2003, 11:54 AM
irony! :)

these probs are around since years,and were never fixed. reminding them its not bad IMO... they are also valid for the next future. :)

I think I and other people simply want Lw to become better and better, being also highly competitive with other apps.
to say it better, we want that on the base of Lw strenghts (and there are a lot) and good philosophy,weaknesses are fixed with the animator in mind, and taking some ispiration from other apps , Like MB Maya, XSI too.
without loose Lw own identity,of course.


personally, I am also happy that I and other people can express opinions in this forums. I think Nt will keep everyone's opinion and ideas into account.

lunarcamel
11-01-2003, 08:25 PM
In my opinion Layout is a half *** solution for animation. When it comes down to it, Layout needs to catch up to Maya's toolset and workflow for me to consider using it seriously. If I had all the time in the world sure I could make it work but in a studio envirionment it just doesn't cut it. Now if Layout was as powerful as Modeler then it would be a whole other story ;)

What I've found too is that in a work situation the companies I have worked for have always put toolset/workflow ahead of cost per license - because in the end its no good if it takes you twice as long to do something that takes half as long in another package. I hope to see some serious workflow improvements in this update.

Ok Im done complaining - Im gonna play subspace and brain drain :)

animotion
11-02-2003, 09:59 AM
It's good to complain if you have valid reasons to do so, belive me these are really good ones.
Its only too late when your six feet under.
Its never too late to keep NT on their toes.

Keep hope alive :D

AngelDream
11-06-2003, 08:50 AM
I'm just a newbie on Lightwave, but I'm skilled in "software engineering" and I think I can say a word or two about this. First of all, I agree about that tools are tools, and you must have your way using them. However, though NewTek sould provide NewTools for this, it's not an easy task. Just thinking about a civilian engineer being asked to add two more bathrooms in each floor of a building and three more floors to the top, does it sounds reasonable?? The whole building would just fall apart!! Now, you must be thinking: "it's just some stupid f** computer code, just add some more code lines and it's done". The fact is that NewTek, as far as I can see, does a tremendous effort stuying kinematics, dynamics, fluid dyinamics and particle motion, not to mention other esoteric stuff like sub-surface scatter. All these things cannot be get together in the limited time we ask for all of them, it's not just a code rewriting, it's a whole new application what we're talking 'bout. Fortunately, Lightwave has some great future changes prediction principles and it has loaded many things into plug-ins, if that wasn't the case, forget about improvements!!!
Nothing can be done in the 60 days remaining, hey, I'm developing an information system and every time the client adds something, many things must be checked (for those who knows, not too much since we're using all information hiding, abstraction and so), imagine it with a system like our lovely Lightwave. Just give NT some time, answers will come. Keep Lightwaving!!!

Elio.

PS:if you note some strange writing, I'm from Argentina, so english is my 2nd language
PS1: there's not such thing like software engineering, is like saying a man is engineer, and is building ships and the next day he introduces himself in a plane engineering plant, jaja.
PS2: hope ur not bored by now, sory!!!:D

madrenderman
11-06-2003, 07:33 PM
my big frustration is to see a lot of good ideas in lw, but half developed, and never completed.

i'don't know why many powerful Ca tools of lightwave are not completed, but is a bad thing.

using many application mean many software knowledge, means good import export way, means a lot of work added to standard work.

if i can, i avoid to do ones work with different 3d application, be cause often means a lot of wasted time in exporting, synch and correction between different application.

lw has a very good base, but incomplete.
if every user tell its problems (real problems) to newtek, probably we can have a better software, also for ca.

i use a lot other software, and no one software are perfect, but if we (the users) tell the problem, bug, or suggest, software can grow, and do better.

Nemoid
11-07-2003, 06:23 AM
Originally posted by AngelDream
I'm just a newbie on Lightwave, but I'm skilled in "software engineering" and I think I can say a word or two about this. First of all, I agree about that tools are tools, and you must have your way using them. However, though NewTek sould provide NewTools for this, it's not an easy task...etc.

Talking only for me, I never, never asked for a quick 60 days implementation of incredible tools and astonishing bug fixes.
most of Lw probs are at its base, from a programming POV,(like one undo in Layout wich means that you can add some undo capability here and there, but not for all functions of the app)
and are there since years.

this means a not clever developing during these past years. in fact. Layout remained weak in time, while it was a part to develop more.

e.g. at a certain point, Nt introduced skelegons, wich are a way to setup bones in modeler.
but editing tools in Layout for bones were poor and remained like this till now.

if they though at that timet : "hey, the animator could setup its char in Layout instead, and so , test bones deformation and behaviour with the mesh immediatly. correct errors, etc.Simple! powerful!

Since current tools for bones are not so user friendly, lets project tools to aid him to make this in an easy way directly there!

Now what about weight maps ? uhm..a char must have weights, but the poor animator have always to go in modeler to setup them. got to find a way to create them too in Layout..."

ok. this is a way of thinking. with animator's needs in your head.

wich is the simplier and powerfull way to do the job?

this is the main question a dev team have to always try to reply.

ok. this isn't exactly happened during the past, and because of this behaviour, now we have tools half implemented.

now, in these days we have a new team, wich will work for developing Lw 8.x releases.
the only thing I ask is : develop tools with the animator in your mind. test them hardly. check what's good, mantain it, and develop it to the maximum amount of power.

that's all.

wacom
11-07-2003, 09:25 AM
I think that the complaints should continue! It might also be good to compliment them on what they are trying to do. Though I can see being bitter about upgrading from 6 to 7.5 in the past...

I was worried about many of the new features in LW8 being just plugins 'till I found out that the people who made them are now part of the LW development team. Take the Ortho Pack for instance. People keep saying that Newtek just took the plugin off the shelf and slapped it in as they have done with other plugins. But they bought out the people and had them rework their code to be more intergrated AND more advanced than it was. They've put in a whole new IK system to go with it. Doesn't this show that NewTek is ready to start reworking LW from the ground up in key areas?
Plus look who the new poster child is for NewTek- Proton the CA freak. I'm not saying that LW 8 is going to be the second comming, but my god people can't you say it's going to be a BIG setp in the right direction for CA. They know they have no choice but make major changes- they know what is out there.

I'm hopeful that the new IK will be MUCH faster (I don't know about OpenGL though which has me worried).
Many of the objects demonstrated at Sigg had an insane amount of bones in them yet they were responding well. Granted Proton was using a crazy BoXX system but still...

I guess well see when it comes out though...

The nit-pick system or the reward system...