I would focus on the first part of this rather than the acting bit at the end...it's a bit too much of a big take right now that happens really quickly without a lot of necessary settle into it.
I like the idea of the happy march in...be careful to really push the breakdown frames so he's not hitting the poses in such a linear rate. Make sure the back foot is planted enough to get a real push from it.
Saturday, February 27, 2010
Posted by jeff at 11:30 PM
Wednesday, February 24, 2010
I was putting together a lecture on runs and took a minute to do some really rough studies to try to understand them a bit further. This is mostly me noodling around trying to understand the mechanical relationships beyond just doing one and having it look good. I try to do these as much as I can/need when trying to solve a motion problem. Pose and motion relate completely in animation, but I find in doing cycles, the motion itself is a kind of character, or, rather can be a really strong indicator of what's going on in the body. Runs in a way are almost an animation of a strong pose, and the character and energy applied can really be found in the little things...how much hang time in the kickoff. What are the arms doing? Are they pumping and driving the motion like a sprinter? Or going along for the ride as an extension of the torso rotations in the case of a jog.
I did a couple of studies...one as an 18f run without a huge stride. It's more of a jog and designed to get a general sense of cadence and rhythm. No rotations yet, just a general idea and an attempt to get the y of the pelvis relating to the movement of the feet...always the first thing I try to nail down.
Next, the 18f run with torso and pelvic twist. This turned out to be the most important thing to get. It was easy to overanimate everything, but the twisting of the torso really made it feel more like a run. I started out with a lot of compression after the feet contacted the floor, but it was looking really rubbery and, well, wrong.
After this came the 18f run with compression. Whereas in a walk there's a nice, rhytmic, sinuous motion to the spine, in a run, it seems to be more of a reaction to the impact of the body from the pumping of the legs. This of course, becomes inviolate when doing a full on sprint, Where the body is held much more rigid. I'm mixed on the compression idea...I think if it's done right, it can add a lot of weight and character to an animation that by it's nature can be somewhat mechanical (ameliorated by offsetting things..having hip and shoulder rotation happen early and late relative to the body, etc). The thing I'm mixed on is that it can look kinda floppy if overused (as my example is a bit) What I was happy with was the slight offset of the y motion in the pelvis to torso...the offset compression made it feel like it was a connected body rather than a series of parts. It was still looking static so I added a minute bit of forward back in the z, which made another rhythm that seemed to help make it less static.
Next, I started taking a stab at the 14f run...more of a sprint, though the feet are completely wrong...there should be a higher arc in the front and a quick descent, but again, these are studies, and that's the next step in refinement. I changed the stride length, positioned the legs further back, and angled the pelvis forward. I should probably explore a less slouched position for the upper body. Working through them, I ended up with one with a lot quicker torso twist, added hands to show the pumping action, revised the arcs of the feet.
These are super rough studies, so don't judge! They're a great refresher for quickly testing yourself on what you actually know about motion and what's actually going on in a cycle in a way that gets the information out as fast as possible.
I hope this is useful, and feel free to add/dispute any of the above...
Posted by jeff at 1:04 AM
Saturday, February 20, 2010
Looks pretty good - the one thing that's most noticable is the weight transfer onto the stiff leg. The stiff leg is too far back in general and probably takes too big of an arc from front to back. The sense of drag is good but a little uneven..something endemic to in place cycles. In doing cycles where one leg is stiff, you generall push onto it, rising a bit to get the momentum to fall onto the other foot, which would fall hard. You've animated the dominant foot in a little more of a cartoony manner with extra overlap on the way down, but that works for this model/style of animation.
The arms could use some refinement in the spacing, but in the degree of drag in the poses and in the amount of noise in the settles.
I didn't touch upon the spine, but it does need a bit more compression (which would be stiff)
Hopefully this screencapture will makes sense..
Posted by jeff at 8:58 PM
This looks pretty good - it's very readable right from the start as being a kind of cross between a regular walk with a bit of a runway walk thrown in. I put up a quick sketchover and will follow up with a more detailed crit.
Right off the bat, the hip forward/back rotation is the opposite of what it should be.
Timing of the feet is good, but spacing could use some decelleration after the kickoff...right now it's happening right before which would cause a stutter if put in motion.
the hand gesture is fun, but a jitters a bit - i'd keep it more fluid and get more accelleration in the spacing of the hand back to front a bit to get some snappy motion to it...too much noise after it falls forwards...tone back some of the breakdowns. Lead with the elbow a bit more in that gesture to get more naturalism/fluidity to the motion.
Add more frames of the right hand matching the body it's resting on so it won't look off - even a few frames right off the bat will get it pretty close.
Posted by jeff at 4:13 PM