I got a bit carried away doing an rig test and this was the result
Adventurer rig available here: 3dFiggins.com/Store
Adventurer animation
dorina
#2
God it looks mad, the animation is breathtaking, wouldn’t be able to do such thing myself. But as to the camera movevement- It would feel more natural if there was some delay when the camera follows the action ( character moves first, camera follows) You have to think of a camera as if it was handled by a human being. Easy way to get rid of this cheap CG-like look.
KielFiggins
#3
Very good point Dorina, I should dirty up the camera a bit more. Always an issue I have when I do the camera on my own animation, I get in the mindset of ‘if I animated it, I want the viewer to see it’, which leads to a little too perfect framing. Good call.
Roumenov
#4
I know this is way late. But if you ever think about revisiting, I’d suggest to delay him turning his shoulder for the roll. Right now it sort of reads like he was thinking about rolling there the whole time. He starts turning and easing his shoulder toward the roll pretty early. It makes it feel a little less frantic to me.
To the same vein of emphasizing the tone of frantic I’d say to maybe let his body squash a bit in the roll. Making it a little less deliberate and clean, so it’s like he’s falling as much as he’s rolling.
Also having his left hand be more open when he comes up, more severe drag. I just think the hand’s pose looks sort of casual. Maybe I’m overthinking it 
It’s really solid work, regardless.
Questions;
Does the rig have any form of dynamics, or is all the gear manually keyed? I’m thinking of grabbing it to try out, but I don’t remember that detail being clear on the page.
I really like that one article you wrote about blocking, btw. Thought that was a really interesting approach and I sort of did a similar thing before(but had less idea what I was doing, haha). Did you do that for this one, or just go straight ahead with the rig?
Roumenov
#5
Oh, I know what it is. His left arm comes around in the windmill action as he drops the stick and stumbles. He also drops his head super early, which removes any opportunity to see his internal, “oh shit!” thinking. But then it sort of eases and reverses its direction, rather than continuing the previous staggering that caused him to drop the stick.
Here’s a shitty snipping tool doodle with my mouse;
That’s what I meant.
Another question/comment. I really like the mech scorpion that’s in your reel. I remember seeing that trailer and particularly noting the awesomely successful sense of weight the scorpion has. What do you think is important for that? Is it the long follow through on weight shifts? I’ve previously struggled with that, so I’d be really curious for your input. Also, what was the name of that game?
KielFiggins
#6
Thanks for the feedback Roumenov, all valid points. I might open up this file again to flush it out a bit more. Thanks for the paint over as well.
The rig doesn’t have dynamic secondary, since those can be tricky for non technical animators to figure out the settings. And building a setup that has both and a blendable way was a bit to complex for the time I wanted to invest. So all the accessories are hand keyed.
The animation on this guy was primarly mocap with edits for camera, eye line and environment. It was more about showcasing the rig and telling a fun moment than animation polish. I have to do enough of that at work : ) Glad you liked the article I wrote about blocking, I always wonder if people read that or gain anything from it. Good to know.
The mech scorpion you’re thinking of is from the short Azureus Rising, you can read more about it here http://www.3dfiggins.com/Personal/AZ/ In the description is a link to the official site and short.
As for weight on something that large, I find two things to be useful.
-
Heavy objects have smoother curves. So his body for example, carrying all that weight would make the motion much more fluid and rocking instead of sharp twists or change in directions.
-
Delay overlap through zones. If the body adjusted one frame, I’d try to delay the hips a few frames later, then the base of the tail even more, and the tip of the tail even more so. This way the energy generated from the body takes a while to dissipate out. Think of it like a dinosaur, who could stomp the ground, but it would take a while before that ripple effected the tail.
Hope this helps.
Roumenov
#7
Ahhh, I hadn’t considered the usability issue. Hadn’t specifically thought about delaying overlap, though it seems obvious now that you mention it and I can see it in my head.,
Do you have a way to shoot the motion yourself? Or do you just happen to have a bunch of data on hand?
I’m real impressed by all these side projects you keep up on outside of work. =D
KielFiggins
#8
I have a bank of mocap cobbled together from various remote projects I’ve been a part of. Crowd stuff and soldiers mainly. I’ve looked into a Moven Mocap setup for personal use, but it’s not quiet cheap enough or refined enough. But I imagine in a couple years it will be.
I like doing these side projects, since it provides a nice sense of completion where as the larger projects I worked on professionally take years to produce and release.
