williamgaylord

Nice work, William. Guess I'll have to get back to work on my trees. In my case the goal is to make it grow from a sprout to a full grown tree. Quite a different challenge. I will be interested in how you do the foliage and how well it renders and how computationally intensive it might be. BTW, how do you set up a thumbnail picture with a link to the full size (as a curtousy to those with narrowband links or simultaneous rendering going on)? Tree growth animation thread... Bill Gaylord

Or maybe a Marshmallow munching on a bag of little humans? Bill

Haven't decided yet. Probably one of the simpler sight gag ones. (The list so far is on the first post of this topic.) Actually the first Marshmallow film will be a spoof of something Justin Barrett did. Can't say more--it would spoil the surprise. Bill Gaylord

Here is a closeup movie of the shoulder in action: *Warning: Imbedded Movie--will automatically play in a window.* Closeup of Shoulder Bill Gaylord

I'm just experimenting to learn more about various options. A single fan bone would likely work with this shoulder, with a bit of Smartskin to clean it up. The array of bones around the ring that pivots is SonofPat's idea--very similar to Victor Navone's mouth rig in some respects. They are basically fan bones, but one for each control point (children of the shoulder bone with orient like constraints to the upper arm bone). You adjust the enforcement for each one. You can find the explanation on this technique on the ARM site...under the "constraints" tutorial section. Works great for hip joints, too. The "translate to" constraints on the "target" bone keep the "axel" bone's "aim roll at" target to one side. If it were to cross to the other side, the roll handle would suddenly pop to the other side. That makes for a very badly kinked shoulder. It also limits how much the shoulder rotates to follow the roll of the upper arm. The "axel bone" idea is was one way to get the shoulder to rotate to follow the twisting of the upper arm bone to help reduce kinking, even while the joint is at an angle. A "roll like" constraint might be simpler. BTW: What's got into Major Innovation? His little dance in your icon looks like he's telling a funny story to some kindergarten kids. Looks very natural by the way--quite impressively so--but just a bit goofy. Bill Gaylord

Yes. Another annoying habit of mine is to directly imbed relatively large pictures, instead of a small replica with a link to the larger one.... Guess I should provide clearly labeled options. Thanks for the suggestion. Bill Gaylord

Here is a little movie of how the shoulder rig works. It's a bit small to see the shoulder but you can still get an idea of how well it performs. *Warning: Imbedded Movie--will automatically play in a window.*Shoulder Rig Test Movie Bill Gaylord

Haven't had much spare time in the past two weeks, but have squeezed in time to experiment with various ways to rig the Marshmallow's arms. Here is what I did for the shoulder. The idea is to make the main shoulder ring pivot like it is fixed to an axel. The next ring is the actual pivot point and acts mostly as a hinge joint. The result is similar to a universal joint that connects two rotating shafts at an angle. This is a simplified shoulder design for simple characters where the shoulder does not really need to be joined to the body. The following picture shows how it works: The "axel" bone is a child of the shoulder bone. It has an "aim roll at" constraint to a target that is in turn controlled by the angle and roll of the upper arm bone. The target bone has two "translate to" constraints--one to a bone that is a child of the upper arm bone and one that is a child of the shoulder bone. This combination allows the arm to rotate forward beyond 90 degrees without the target bone passing to the opposite side of the axel bone. If it did, the roll of the axel bone would suddenly pop to the other side causing a gruesome kink in the shoulder. I use Sonofpat's idea of using several bones with "orient like" constraints around the shoulder ring associated with the pivot point. That does most of the work of smoothing the shoulder mesh. I then give it some final tweaks using a Smartskin on the upper arm bone, setting just four extremes--forward, backward, up, and down. Works great with the TSM2 arm rig! I'll do the same with the hip joints as the Marshmallow has no hips, strictly speaking. ("He has no butt!!!") I may post an animation of the arm rig to show off it's performance. Bill Gaylord

The happy expression on his face reminds me of Louie Armstrong. All he needs is a trumpet! He's looking great! Bill Gaylord

I'm not sure how to avoid the crashing, but I find that some times adding or changing something in one of the poses makes the crash go away. What I can help you with is how to properly delete poses. The confusing thing is that you have to delete a pose in two places to completely remove it. First delete the pose listed in the Project Work Space under Relationships and then click on the model and delete the same pose under User Properties in the Properties window. Bill Gaylord

I think I finally have all the fine mouth controls sorted out. Here are sample pictures: Now he can even whistle!

Most of the facial rigging is "bone free"...mostly muscle mode or simple translations controlled by percentage poses, except for the eye rig, so not much would be visible as in a normal skeletal rig. I will share as much as I can, though. Bill Gaylord

Actually this was in an action window, so lighting is not ideal. This was just a test of how well the "puckering" of the lips works. Normally they won't "pooch" out like that. I am very pleased with the lack of creases or ridges in this mesh design. Soon I will have all the mouth controls configured and will move on to the rest of the rigging. I wanted the mouth to be one of the high performance pieces of the model so I could get a lot of mileage out of the mouth and eye expressions. Thanks for the input! Bill Gaylord

Wahoo! The small mouth and new mesh design work great! Here's how the extreme of the pucker looks in the smile/purse extremes: Smooth! I think this is a keeper. Bill Gaylord

Here is the smallmouth variety of Marshmallow... The mouth slit is just visible, but should give you a sense of relative size. I think this looks more appropriate and should be much easier to rig. Also added another spline ring around the mouth that should help quite a bit. Well, round two of rigging a mouth! Bill Gaylord

The wide mouth design is creating challenges. For most mouth movement it has worked out OK, but an extreme combination of pursing and puckering the lips looks a bit ugly in my opinion. Let me know what you think. I'm seriously considering redoing the mouth mesh. Here are some pictures to illustrate. The first is the neutral mouth. The second and third show the purse/pucker. Note the ridges that form in the extremes. Here is an early thumbnail sketch of the marshmallows. This design has very small mouths. Any opinions on whether I should make it small as in the original sketches or extra wide as in the current design or something in between? Smaller would behave better, but give me your votes! Or mesh design suggestions that might make the wide mouth work better in the extremes. Bill Gaylord

Fixed the link. You should be able to download a "zipped" copy of the project file. Enjoy! Bill Gaylord

I can't download the project, so I'm wondering if the link is working for you? Bill Gaylord

Tried some experiments and came up with this: GerryEye Project You can check out the project to see the details. I wound up assigning a bone to each semicircular spline, instead of using Smartskin. The in-between bones have orient like constraints such that rotating the main eyelid control bone also rotates the intermediate bones by a proportionate fractional amount. This gave the smoothest results. There is a bit of distortion, but it looks very like the normal bunching of a real eyelid. For some reason I am unable to adjust bias in a percentage pose--have experienced this before and have reported it to Hash support. Hope you find this useful. Bill Gaylord

Yep, that's how it works: Since the lids are fixed bowl shapes I assign a bone to each lid and center the bones origin at the very center of the eye. That is the way I get the lids to follow the curve of the eye. I then use a set of additional bones, constraints and percentage poses to build a set of controls for actual use in animation. I would highly recommend Raf Anzovin's "Rigging a Face" tutorial CD. He presents a way to rig the eyelid controls that is easier to use in animation. It takes a bit of time to learn but is worth the effort. What I would recommend for your model is to use one bone each for controlling the top and bottom lids in your model that pivot at the center of the eye. Then use a Smartskin relationship for each of the upper and lower lids. Use the bone rotation to control the rotation of each of the semicircular splines in the moving part of your eyelid. Make sure you go a bit farther than you expect to give yourself enough range of lid motion. Once the Smartskins are keyed, you will have four main control bones that control the four lids. Then you can add constraints and percentage poses that in turn control these bones to give you a nice set of animation controls. Bill Gaylord

Since the lids are fixed bowl shapes I assign a bone to each lid and center the bones origin at the very center of the eye. That is the way I get the lids to follow the curve of the eye. I then use a set of additional bones, constraints, and percentage poses to build a set of controls for actual use in animation. I would highly recommend Raf Anzovin's "Rigging a Face" tutorial CD. He presents a way to rig the eyelid controls that is easier to use in animation. It takes a bit of time to learn but is worth the effort. What I would recommend for your model is to use one bone each for controlling the top and bottom lids in your model that pivot at the center of the eye. Then use a Smartskin relationship for each of the upper and lower lids. Use the bone rotation to control the rotation of each of the semicircular splines in the moving part of your eyelid. Make sure you go a bit farther than you expect to give yourself enough range of lid motion. Once the Smartskins are keyed, you will have four main control bones that control the four lids. Then you can add constraints and percentage poses that in turn control these bones to give you a nice set of animation controls. Bill Gaylord

I like the eyebrow action and left eye squint along with the hand gestures as the "green" guy pauses to reconsider...like he's working out some calculations. Works quite well. Bill Gaylord

I posted the wireframe of the Marshmallow guy's eyes, plus some notes. The main concern I would have using that approach is that in your model you can see much of the back half of the eye. The approach I've used is the same as David Rogers' Washer model (in his book "Animation:Master 2002 A Complete Guide"). John Henderson's Kapsules probably use the same kind of eyelids. They are basically two hemispherical bowl shapes that pivot about the center of the eye. In your case the back half is visible, so the flaw in this design will be visible. I hide this by positioning the corners of the eyelid intersection near the surface of the face. You might have the start of a better solution for your model that would avoid this flaw. Not as simple to set up, but not all that difficult to set up in a set of percentage poses. I use a slightly modified version of Raf Anzovin's method of rigging eyes (in his "Rigging A Face" tutorial--available on CD at Anzovin Studio). His approach would work well for your design, too. Really like your character design! Bill Gaylord

Thanks, Gerry! Here is a picture with the wireframe that should at least give you some hints on how the eyes are constructed. For a detailed tutorial on how I build the cartoon eyes with a resizeable pupil check out: Cartoon Eye Tutorial To make the lids, lathe a sphere a bit larger than the eyeball, then delete one pole plus all but two of the spline rings south (or north--depending on which pole) of the "equator". Then you sort of tuck these two spline rings up inside the lid so the surface curves inward at the equator. By adjusting the diameter and postition of the tucked-in rings, you can get very close to a perfect outer hemisphere shape without having to tweak any bias. These lids are basically simple bowl shapes...nothing complicated. Hope this helps! Bill Gaylord

For the past few months I've been overwhelmed with my "paying" job working for BellSouth as a technical consultant. The tree animation project has also been neglected for quite some time. The marshmallow is a bit of a diversion that is helping me learn a lot about rigging a face for animation (albeit a simplefied one). Once I get this little guy rigged and cloned, I should be able to crank out a series of shorts sort of like "Petey and Jaydee", but with marshmallows. Bill Gaylord