Blender.org

I like Harkymans proposals and
hopefully some of this stuff will make it into a release, but on the other hand I would be glad to wait a little longer for a complete new approach to animation in general.
Why?
I think that it all will be obsolete in a matter of a couple of years.
In 10 years time ppl are probably laughing about the good old days, when frame by frame key settings were necessary to create animation.

I had this thought when Ive seen a demo of an euro 49,- pc game I think (not sure)it was fifa soccer : http://www.easports.com/games/fifa2005/home.jsp
one can use a web cam to create the shape and the texture of a players head. Even morph targets are created by the use of 2d reference points.
This old link came strait into my mind:

http://www.blender.org/modules.php?op=m ... pic&t=3083

this would give an animator a complete new role to play. Exactly: to play and act! Because an animator would then actually act in front of a camera and see his/her action in realtime on the mesh.

Also character animation will become eventually a completely different ballgame:

One can use default settings like this
http://www.biomotionlab.ca/Demos/BMLwalker.html
and with a couple of clicks create a walkcycle, or with the development of mo-cap infra-red cameras, just skip and jump around in front of the computer and see it all in realtime.

This is where it will be heading! Im pretty sure of it.
Hardware for that, namely infrared cameras for mocap will eventually become the same price as web cams today and everyone will be able to afford them.

With the captured movements, one can build up a library of actions and then just use the NLA to edit the final animation, and/or fine tune it.
Some stuff, like this library could be integrated as is:
For example:
a walkcycle can be adjusted to the characteristics of the actor and confirmed. Further actions from the libary are then mixed to it in the NLA.

Sure it is a long way to go, but I think we should think more about the future and have these things in mind, when developing new animation functions.
interactions with collition bodies, as in the game engine become ipo-curves and actions (there is a python-script for it, but I doubt that it is developed further) so the actors can interact and create keyframes while doing so....well...just thinking loud here

but this could also mean, that blender would set a whole new trend in animation and out-perform other apps

as I said, this is just a thought I had the other day, but maybe, just maybe blender could be the first in actually give access to such a technologie

Yeah sure... mocap will become easier and more affordable, but you still need a good anim system to adjust the results and I think you're looking further into the future than you realise...

The already is an animation system that has no keyframing: video

But I always thought that (computer) animation was a tool to visualize things that are not out there in the world. Starting as basic as flying logo's to simple as explaining a factory process to as difficult as growing alien forrests.

True, hollywood (and games) has it's focus on fotorealistic syntetic actors, but as people will be able to grasp more visual ideas, hollywood will have to follow.

Then there is another little part that makes motion capture not always a helpfull tool; the people doing the motion need to be very good body actors will the result be more then boring. (ofcourse soccer players tend to be very able to play soccer players).

So please, let the tools begin !


ps. still looking for a woman to motion capture a character called 'elastigirl'.

People have been predicting End for Traditional Animation for years now. Mostly people who don't know much about it and are awed by technological tools like motion capture and 3d.

They say: "why to draw, when you can so easily make these photoreal teapots over chessboard?" And "animation will surely die, because now we can capture a twitch in the tiniest muscle, it's so real!"

Animation has never been much about _real_. And even if the result needs to be realistic, it is pretty hard for any actor to behave as a giant squid. Motion capture or not. (Unless the actor is a giant squid, in which case it can't act and shits on the set.)

edit: fixed a typo

heathens !

but seriously: That's like saying painting is dead because photgraphy exists. Animation can be realistic, but it almost always isn't- you could say it has "impressionist" and "expressionist" varieties.
mocap has it's place, but it is not animation, though you can have a mixture of both ( currently, you almost need to since some "cleanup" seems to be always required)
anyway, I'm just repeating what has been said. Silence now.

well, if one tool would care about the cheap way, is Blender

Actors are expensive, as been said, I mean for a hobbyst to pay it.

And as said, they have to be reallly to reproduce what you want.


Machines are really expensive. As emuch as a webcam...hehe, not yet.

I was near to buy a 3d scanner, back in '98...I thought modelling by software was coming to an end...no way...there's allways the need of computer touch, and allways anyway the alternative... films are done with clay modelled figures, often, but there's a load of work later on by computer. Mocap systems are really expensive for a lot of big game companies , let a lone a hobbyst or individual professional...

I can model in clay or similar, but I have had to learn digital modelling...

Guess what, if I go to a game or video company saying that i can paint and model in clay, and that's all I do...they'd laugh at me...

I think Blender is needing a lot some animation improve, in much more basic features, like(foot planting, limiting axes, and the ones from Harky's proposal..) ...well, like many features needed.

Well if you love final fantasy, mocap is for you. But some of us think realism is just boring, I don't think I could animate whitht walk cilçcles "presets", just read a book like animators survival kit, theres so much fredom on the poses