How to do basic 3D animation?
interpolator
Nothing too hard. I'm just in the mood to animate something 3D and make it look like Looney Tunes. Sort of like the Crash Bandicoot games on the PS1.
Software I have: 3DS Max & Unreal Engine 4.
I wouldn't even start with a character. I'd animate a teapot.
Software I have: 3DS Max & Unreal Engine 4.
I wouldn't even start with a character. I'd animate a teapot.
Replies
If you're already familiar with it start to do some studies and see what's the difference in animation on Avatar versus Road Runner. They both use the same principles but in different degrees for different results. There is no recipe for success only iteration and keeping it simple to begin with.
But what are you actually asking? How to do the animation or how to create a game like Crash Bandicoot with your tools at hand?
I know there are 2 different forms of 3D animation, so I was trying to go for the one that allows for the most freedom. The Crash Bandicoot example is based on vertex animation which is what I want, but I couldn't find many tutorials for it.
Try 'Sprite animation tutorial'
- http://design.tutsplus.com/tutorials/how-to-create-an-animated-pixel-art-sprite-in-adobe-photoshop--cms-20428
- http://www.manningkrull.com/pixel-art/walking.php
*http://gas13.ru/v3/tutorials/sywtbapa_breathing_life_into_sprites.php
edit: Nevermind, just saw that you wanted to do it in 3D.
Animating a full character using vertex animation/blend shapes is incredibly time-consuming, so you'd be better off building a rig for it
That squashy stretchy, noodle arm style, is probably some of the hardest stuff to rig up and animate in 3D. Also I'm pretty sure Crash Banicoot had a skeletal deformation system of some kind. It was the dark ages but it wasn't that dark...
I also think the payoffs are really worth it. Breathing life into each vertex has a certain charm I find very desirable. Just like how old 2D cartoons look so good because it's all hand manipulated.
I tried my first animation. I made the snout move and expand. So much fun!
[ame="
The developers said they did it all through vertex. Everybody else did skeleton animation (but not naughty Dog)
http://all-things-andy-gavin.com/2011/02/04/making-crash-bandicoot-part-3/
I also wasn't trying to imply cartoons were easy. I was referring to learning 3D animation from the basics as opposed to the very professional level.
http://wiki.polycount.com/wiki/Category:Animation
It's a lot of fun! But it's also really hard to get it right, as others have said. Thankfully people have shared their knowledge.
Think about that for a second, manipulate every vertex in a scene to produce an animation. That would be like 2D animators using microscopes to place each granule of graphite.
Take a scene like this for example:
[ame="
I use vertex animation, every day. As part of the rigs I create but it's not the only thing I use, just part. When used correctly with other things, it really does give the characters some extra kick.
I look at animating one of those people in the background behind the steam shovel, only with vertex animation and I say screw it, there will be no joy, it will take 3-4-5x longer and I'd probably rather be unclogging toilets with my bare hands. In the time it would take me to animate one of those guys, with only vertex anim, I could do the entire scene. It's not because vertex animation is something new and the learning curve is steep, its that its so shallow that it can't pull off what I need it to do.
Are you really ready to manage and manipulate EVERY vert in a scene, or would you rather build groups of verts to interact with? That's all bones allow you to do, build groups and blend between them. Using traditional, well established, decades old, rigging techniques will help you build systems to drive the bones in predictable and helpful ways. Those systems handle the micro management so you can concentrate and focus on the good stuff. The stuff that actually breaths life into the animation.
For some reason you think vertex animation is pretty awesome? You seem pretty set on doing it. But ask yourself is that just to get out of learning something new and more complex? You seem to want to stick to vertex animation because you think you sort of understand it, but you don't.
You think a teapot is a simple object? Why? Because it's easy to make? Click drag done? But if you stop and think about it, it's not an easy object to animate with vertex animation.
A ball? Maybe, it's one contiguous piece of welded geometry. But the multi-part teapot primitive has parts that just clip into each other, they aren't contiguous. That will get annoying trying to keep the spout, handle and lid in the right spot.
I suggest you fix that before it turns your project into a micro-managing nightmare. What if you did a bunch of vertex animation and then found that problem and wanted to weld it all together? What do you do? Your options are A) Redo the animation, or
If you have the lid fly off, how do you get it back into it's original position? You better hope you have some keys that get it into the right spot and that everything else is in the right spot so it all aligns because you don't have any systems that can help you go back to that pose.
If you're dead set on doing it, you should give it a try. I think you will find out quickly why people don't do it. Personally I think you should listen to the advice that "the stove is hot, don't touch it" but if that's the only way you'll learn, the stove is in the kitchen...
Just know it will more than likely end in failure and frustration. Hopefully then you will be ready to learn the actual systems that have taken years, decades to invent and perfect? Because even if you're project doesn't make you hate animation, you'll still have learned there are limits to vertex animation and they suck and that you have to learn the actual way to handle those things.
You will end up coming to the same conclusion as everyone else. But if you want to take the long weird and winding path that's your bazaar choice to make. Who knows maybe you'll be that 1 in a million person who invents something amazing and teaches the world a new way to think. Personally I don't think that is likely to happen but who knows...
For a first animation... that's... ok-ish. No, wait I can't soft pedal it. Sorry, but it's horrible, you can't call that animation. If that's your quality bar then great, you met it, congratulations. But honestly I think you should learn the systems that would enable you to something close to what you're saying you want to do?
[ame="
If you're looking for something to practice vertex animation with, pick a ball not a teapot. If you master that, add a tail to the ball and then after week or two if you haven't jabbed or thumbs into your eyes, you might have what it takes to add legs to the ball, THEN you'll wake up in the ER blind as a bat.
Now, I'm not saying you can't pull off Mrs Potts with vertex animation, it its technically possible and her motion isn't that complex, but there are easier and faster ways to do it. Also keep in mind that because the rest of the world uses the tools and systems that have been created and perfected over the last few decades and you plan to ignore all that, you will probably have to write some tools yourself in order to get even the simplest project finished.
So yea vertex animation has a place and is very valuable but it's not the star of the show, it never will be and trying to get it to act at that level will only end up in misery.
No they didn't, read that again. They used a joint based system that allowed them to blend the vertex weighting between joints and then dumped the vert data into the game because the hardware only supported simple vertex weighting for bones.
Why did Naught Dog do it?
Was it because vertex animation was some superior form of artistic expression? Or was it that consoles launch with shitty out of date hardware, than then sits around for a decade. Meanwhile the industry blazes forward with new techniques (like weighting verts between 2,3,4 bones).
That's why they did that, the hardware sucked. Not because vertex animation allowed them do free the characters. The industry does goofy things (like ND did) to use current techniques on ancient out of date hardware, all the time. For every character you see in an Uncharted game there are 2-3-4 skeletons that never make it into the actual engine. They all allow them to manipulate the characters in ways that are intuitive and easier than using a single skeleton. If you only look at the final result, you'll never really understand how they made it happen, you will only draw the wrong conclusions and go down paths that have dead ends. Don't be that guy, don't waste your time like that.
Lets back up for a second, lets say you successfully animate a walk cycle for a character using only vertex animation. Now lets say you have another character that is similar but not identical and you don't want to redo that hard work, how do you transfer that walk cycle? Answer: You don't. Unless they are identical in every vert right down to the way the verts are numbered you can't share the walk cycle. But with bones you can, the bones don't even need to be in the same positions or the same lengths or have the same number of bones or similar names. You can put a midgets walk cycle on a giant and vice versa. You can edit that cycle to give a slightly different personality all the time building off of work you already did, that you don't have to redo.
So two teapots walking down the street takes you 2X as long to animate with vertex animation. Unless you rob them of their character and make them identical. Hooray for shitty limitations that cause a lot of busy work and keep you from the fun stuff.
I wasn't trying to offend. I think it should have been obvious that I've never animated in 3D before and thus it's no surprise if my comments sound really ignorant to you. Which is understandable from your perspective. But I hope you understand it from my perspective as well.
That said, I still appreciate the time put into your post and I will take up the wealth of information there.