PDA

View Full Version : What grown-ups do,...


scotttygett
03-16-2003, 07:21 AM
I attended a recent Siggraph event where there were hundreds of animation newbies. As I listened to the producers and recruiters and HR people describe their ideal match, I couldn't help wondering what a "job shadow" of the typical animator would include? At Friedman, I learned that once a model was close, it was sent to the texturing department, and they would bring it back if it had serious UV issues. So, instead of grieving over texture adjustment, the model would be adjusted to get it more workable?

I have a bunch of questions along these lines. A few MAY have different answers for non-LWers, and I'd like to hear those answers too, if possible:


Are ALL objects UV-mapped these days? It sounds like this is what's going on.

Motion capture: do animators use gloves/wacom tablets? Do they use puppets with Kaydara?

Do animators pantomime on a webcam?

Does anybody morph heads any more? How do you keep the morphing realtime, if you do?

Is object capture/photogrammetry/polhemus used for props? For character sculptures? I've seen how a CAD object can be used to base a more animatable mesh on. Is it common?

Does anybody use collision detection methods for dump stuff like keeping a hand from going through a desk, or is that just a newbie problem?

When you get a really crowded busy scene, is there a different pallet you use for the objects to keep the different positions clear? I heard it was common to use a bright lighting set-up.

Thanks...

twidup
03-16-2003, 08:52 AM
I will answer part of the mocap one.

at my previous game jobs, here is how it goes with mocap. One of the animators, usually the lead, is on set for the mocap shoot, directing the actors.

Animators rarely will use the gloves to control anything. Usually, the mocap data is delivered to the animation team already on teh skeleton that they provide.

As for capturing props, the only systems I have seen for that are optical systems, such as Giant, Vicon, and Mac. They are also solved and delivered to teh team by the mocap team. animators typically dont touch the mocap data at all, unless they are an animator on the mocap team there to touch it up.

-Todd

leigh
03-16-2003, 02:17 PM
I will answer the mapping question.

No. Not all things are UV mapped - technically, UV mapping is often actually less precise than conventional mapping techniques! That, added to the fact that it is extremely time consuming, makes it less than ideal to use for absolutely everything.
Very basic objects such as buildings, etc that have lots of straight planes and whatnot are easier and quicker to texture using planar projections.

UV mapping is for characters, and other things that would be tricky to map it conventionally.

DaveW
03-16-2003, 06:56 PM
Are ALL objects UV-mapped these days? It sounds like this is what's going on.


Nope, but there are some apps that only do UV mapping, and in games everything is UV mapped. A lot of places use NURBS for props or landscapes, and NURBS have built-in UV's so it would make sense to take advantage of that.


Motion capture: do animators use gloves/wacom tablets? Do they use puppets with Kaydara?


I have heard of some people using gloves or puppets, but it doesn't seem all that common. Never heard of anyone doing mocap with a tablet, but I know modelers who use tablets.


Do animators pantomime on a webcam?

Some do, others use DV cameras, some don't bother at all. I personally find it very helpful and you can get a good webcam pretty cheap these days. I use a DV camera at work and have a webcam at home. DV is a lot more expensive (for both the camera itself and the tapes) but it's much higher quality.


Does anybody morph heads any more? How do you keep the morphing realtime, if you do?

Yeah people morph heads all the time. I don't understand what you mean about keeping it realtime though. Assuming your computer is fast enough there shouldn't be any problems morphing. If things are going slow you can always split characters up so the only part deforming is the head.


Is object capture/photogrammetry/polhemus used for props? For character sculptures? I've seen how a CAD object can be used to base a more animatable mesh on. Is it common?

Depends on the studio and the project. It seems the larger studios always make maquettes of the characters for the director to play around with, and once s/he decides on one, it is scanned in and used as a template for the modeler to build an optimized mesh. I don't do much prop modeling but it makes sense to use every trick in the book to speed things up and keep costs down. Sometimes you can just use very simple block objects and project and image onto it to get the effect you need.


Does anybody use collision detection methods for dump stuff like keeping a hand from going through a desk, or is that just a newbie problem?

Newbie problem :) If the camera doesn't see it, then it didn't happen. There were scenes in Toy Story where Woody had his feet a inch or two deep into the floor so that he wouldn't appear too tall next to Buzz. Since the shots were from the waist up or shoulders up, it didn't matter. Live action shoots do similar things, in some of the James Bond movies the actresses would stand on crates (or they dig ditches for him to stand in) because Sean Connery is pretty tall. Other times the male actor will be standing on a crate because he's too short.


When you get a really crowded busy scene, is there a different pallet you use for the objects to keep the different positions clear? I heard it was common to use a bright lighting set-up.


I'm not exactly sure what you mean here, but I try to keep my scenes pretty empty and animate objects in their own scene, then load from scene (merge scene in other apps) and combine all the elements together. I also use bright, even lighting from all angles when I'm animating so that I can see everything, then when the animation is done I turn off the bright lights and turn on the proper scene lights for rendering.

scotttygett
03-17-2003, 09:44 PM
The other day I accidentally stumbled on a worthwhile bones-rigging technique for INSPIRE3D/LW5.6 and I about had a cow.

Thank you everybody for pitching in to say how it's done.



(PS: By the way, the technique is to lay down undrawn joint bones and their children only at joint positions (then record/rest them) and probably make them inactive but visible, then draw only child bones from each joint (reposition and rerest those).)

CGTalk Moderation
01-14-2006, 03:00 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.