phatso

Vern's post brings up a point that can't be repeated too often. You notice that there's a bit of reverse anticipation on the left part of at least one spline. This is because A:M fits smooth splines thru control points. You cure this by changing the interpolation at the control point in the middle of the animation, where the action has to transition from still to moving. If you want acceleration from zero, click on the control point, right click, and select Zero Slope. If you want an abrupt start to the motion - which would be appropriate where the rabbit is getting hit on the head - click on the control point and peak it.

Aren't hooks mentioned in TOaM? You make one by hitting "A" to add a spline, but instead of clicking the left mouse button you click and drag, then terminate by clicking the right button. This attaches a spline to another spline without making a control point. It's useful when you need lots of splines in one part of a model but not another, or when you're attaching things at angles as in attaching a branch to a tree. As you know, you can't make a patch out of just any combination of control points. Hooks are a help in this direction. Which brings us to 5-pointers. Very often, when a 5-point patch won't make, you either have duplicate control points sitting on top of each other (move each one to check, then Undo to move it back) or a problem with flipped normals (shift>1 to see normals, shift>1 to hide them again). Sometimes you have to manually select the 5 CPs, then hide everything else, then Group those 5 CPs. This usually works when nothing else will. The problem usually resides with splining errors and not with A:M.

Or you could cheat: rotate the cylinder 90 degrees before applying.

Is the point of confusion over the definition of keyframes or how they are made? Since you do stop-motion, you must surely know that every pose occurs at a specific frame number, the next target pose occurs a specific number of frames later, and the in-between poses are spread out according to where you came from, where you're going, and the number of intervening frames. The difference with stop-motion is that you can't jump ahead to establish a keyframe, because you'd never get the in-betweens to match up perfectly. But with hand animation you can. In the days when cartoons were hand-done, for each motion the animatior would make a start pose and an end pose, specify the number of frames between them, and hand the drawings off to flunkies to make the in-between frames. With computer animation, that in-betweening job is what the computer does. So: put the models in their beginning places at frame zero. Make that a keyframe (something many people forget to do.) If something is to start moving at frame 0, you don't have to do anything more at that frame. Advance the timeline cursor until the next frame where some part of the action will change. In the case of the dropping ball, it would be the frame where the ball hits the rabbit's head. How many frames do you want between where the ball starts dropping and where it hits his head? That's the frame you put the cursor on. Now, imagine your computer is a roomful of hand-animating flunkies. You need to move the ball down to where it's in contact with the rabbit's head, but until this time the rabbit hasn't moved. So you make a keyframe here with the rabbit in the same position he had at the start. That tells your roomful of flunkies to make all their inbetweens up to this point with the rabbit in the same position. This "hold until the ball hits" directive is what your animation didn't have. If you tell the computer (or a roomful of flunkies) that the model is standing on frame 0 and kneeling on frame 80, it will make inbetween drawings that move the model over that whole range. You need to have that intermediate keyframe that says, "At frame 0 the model is standing; at frame 40 he is still standing, at frame 80 he is on his knees. Make your inbetweens accordingly." BTW - if you know anybody else who has A:M and hasn't discovered the forums, have pity on the poor sucker and spread the word.

OOOh, I gotta run right out to the dealer and get one. Um, not. But I'd sure like to. Renoir said he knew he had a painting of a nude woman right when he got the urge to pinch her. I know a model of a hot car is good when I can feel the key jingling in my hand.

rus - takes your breast away? Was that a freudian slip? (Can you buy freudian slips at Victoria's Secret? ) nix - just tell everybody she's Greek. Greeks tend to have long torsos and short legs. Then your model is right.

The key word is "characters." A:M is designed first and foremost for organic modelling. Polygon-based programs do very well making robots, but you run into all sorts of problems when you try biological forms, and more problems when you start animating them. Just by being a spline-based program, A:M works the way skin, muscle, bones and fabric work. You will hit a speed bump right out of the gate, in that using splines is a lot different than using polygon edges. This takes some getting used to and you'll be frustrated at first. Keep at it. Once you master splines, you'll find that a lot of the work you had to do yourself in other programs is done for you automatically in A:M. How old is your "older" laptop? A:M comes on a disc which must be physically in the computer to boot up. This will become an advantage when you update your computer, as the disc is licensed to a user and not to a seat - thus it's portable.

Coupla other things you have to do: 1. At least in V13, the default time line is one second - that is, your picture will render 30 times unless you change it. set the end frame to 00:00:00 and then you'll just get one render. Otherwise, there will be lotsa useless copies. 2. Look at resolutions and other factors. In Buffers, you may have to choose whether to turn Alpha on or off (depending on which format you render to). In those formats which support Alpha, leaving it ON will result in a black background (unless you have something filling the whole frame).

Suggestion - I haven't played with toon renders much, but I think you can choose the outline color, object by object. If that's the case, the cage should be lightened so as not to compete so much with the mouse.

(more detail) When you create a choreography, the default view is the camera view - you're looking at things from the camera's viewpoint. While in camera view, when you select an object and move it, you are moving it horizontally on the "ground." That is, mouse left and right work as you would expect but forward and back move the object toward the camera and away from it. To cure this problem, as Robcat said turn on N or select another view. The obvious choice is front view, which you can select by hitting "2" on your number pad (NOT the "2" at the top of the keyboard, which shares the @ sign). Now you can move the object right, left, up and down. To move it in the Z axis, switch to a side view (numpad 4 or 6). If you're just starting to use A:M, you should get used to using the numpad to select views right away, and more keyboard shortcuts as fast as you can learn them. A:M is designed to be a "one hand on the mouse, one hand on the keyboard" program. I personally prefer to have two windows open when I'm doing a choreography. One shows the camera view and the other is front, right, left or whatever I've selected at the moment, which I use to move things around. You get a second window by clicking Window on the toolbar, then New Window. I like them side by side, which you can get by clicking Tile Vertically. Are you going to be moving the camera around the lab, with different angles and such? - Cuz if you are, 2D pictures won't work, you'll have to build a lab. How long before the project is due, and how many hours can you afford to spend? On the one hand, attacking a big project before you've mastered the program will frustrate the #&!! out of you. On the other hand, there's no better way to get good at it. Maybe tell your teacher what you're trying to do, and trying to learn modelling at the same time, and ask if the deadline can be extended. Then get started and come back here for advice when you get stuck.

I do use my real name. Just not here. Wells - you're in for some frustration at first. Stay with it, cuz after that you're in for a treat!

Gerry - well, see how much work I saved you? Geoff - I wasn't referring to any of the splines in the model. You can move the whole model up and down... let's see, I haven't done a knight backflip, but I think the inverse kinematics are arranged so you can move the hips and the whole model moves... otherwise use the model bone. In A:M, virtually everything can be assigned a spline and animated; splines aren't just for modelling. This is one source of frustration for newbies - complication, you know - but also one of A:M's great strengths. If you move the knight up and down (either by moving hips or the model bone, you'll have to experiment) the position of a control point will move up and down in the timeline. You set the frame counter to various frames and adjust the knight's vertical position; the resulting control points will make a spline that represents altitude versus time. (Don't forget to set a beginning control point at frame zero.) This is the spline I was talking about. Then, when you rotate the knight, you'll be making another spline that represents rotation. Set a control point at the beginning, set a control point where he starts spinning, set one where he stops spinning. You don't have to set one at the end. I'm no animation expert, but I can tell you that believable animation is all about Newtonian mechanics - forces, momentum, center of gravity, etc. That's kind of what I was getting at.

I assume you-re doing it in slow motion on purpose... Not a bad first try, but you might want to do another one from scratch, and take a physics approach - you know, conservation of momentum and all that. This will involve doing things a bit backward. Establish the basic arc of motion first and then go back and animate the arms and legs. Create an altitude spline on the timeline. This will determine the knight's overall vertical position at every frame. First: The knight, standing straight, goes down a ways and stays there for a moment. This means his feet will go thru the ground. Second, he quickly accelerates upward until the point where his feet leave the ground. Third: from the time his feet leave the ground until they touch down again, the spline should take the shape of a parabola. A good approximation is to adjust the bias handles so they run parallel to the spline. Fourth: his feet should go thru the ground again as he decelerates. Fifth: raise him up slowly until he is standing on the ground. At the transitions between first and second movement and the penultimate (Ooh I got to use a big word!) and last movements, set the control points to zero slope. The slope of the spline as he leaves the ground and as he touches down again should not change radically at those control points. All of the foregoing with the knight standing straight up and rigid, and coming down about where he took off. Next: Click the Rotate button. Adjust the rotation center so it's at his center of gravity (the hips). Make keyframes so the knight is straight up until his feet leave the ground, spinning backwards as he jumps, and standing straight when his feet hit the ground again. Between the time his feet leave the ground and when they hit again, the rate of rotation should be constant, meaning that the start and stop rotation control points need to be peaked. (I would have suggested rotation during jump-up and recovery periods, but then the feet would slide on the ground.) Then: Animate the feet and legs. Use zero-slope and peaked control points liberally to keep feet from moving except when he's in the air. This is the point where you need to tweak and tweak and tweak until it's believable. You'll have to have him lean forward a bit as he begins the jump, have the body arched back when he leaves the ground, legs pulled in at the top of the flip, body arched forward by the time he hits the ground again, and probably do a couple of balancing steps forward or backward at the end, depending on what your eye tells you about whether he's going to fall over. This is also where you correct for the fact that the rotation started and ended when his feet left and returned to ground - you will animate a bit of rotation into the jump and recovery periods, by tweaking the way he crouches. Next: Animate the arms. I'd guess they will look best if they go straight up by the time he leaves the ground, then windmill forward as he's spinning so that by the time he hits the ground again, they've come all the way forward and down and are behind his back a little, with elbows bent. Finally: Go back and tweak the shape of the rotation spline so the spin looks natural at every point. When he pulls in his legs and windmills his arms forward, that part of the spin will have to be made faster and the rest slowed down to compensate. Last but not least: Post the results and bask in glory.

Looks good, the gesture is very expressive. All you have to do is look at that and know about what he's saying. Please take pity on us folks with slow internet and post at lower resolution... as you get further along, these are going to become all-day downloads.

I love going to Will's site just to look at...well, you know... One of these days I'll even get around to ordering his DVDs. (I know, I know. "Phatso, you cheap &($#!" >WHACK

And syntheyes costs how much?

"Sorry, guys, they use relish for money." I completely understand. I rather relish money myself.

John, that was a terrible waste of Thyme.

Hmm. True - using a video file as a rotoscope isn't mocap, in that it doesn't do the work for you - but it does work and it is cheap.

...but then, MOCAP requires extra equipment. Could you be more specific?

About 30 of those 34 are an investment that will come back later when you get good at this. (...says Phatso, who is nowhere near mastering this himself, pretending he knows what he is doing.) In the meantime, it can't be said often enough: when you reach a point where you've got something kinda halfway right, SAVE IT! And then if you tweak it and it improves, SAVE IT! Hard disk space is cheap these days, there's no good reason to leave yourself exposed to possible loss of work.

Hmm. Veddy strange. You're quite sure you did open a choreography window, not a modelling window? It's easy to click on the wrong thing. Assuming you are in a chor window, the only other thing that occurs to me is that you may have accidentally clicked a button in the toolbar at the top. Clicking the little yellow person puts you in modelling mode, and the pose sliders won't work since they're choreography-related. A bit to the right are three choreography buttons: a muscle icon, a bones icon, and a megaphone. Click on the megaphone. If that doesn't do it, report back and somebody else will figure out the answer.

We live to be of thervith. Actually, so will you, when you master this stuff and then see others working on the same problems. Learning any complicated thing is like climbing a mountain, but this one comes with people at every level giving each other a hand up. Happy holidays to all, and needless to say, in 2008 may a fat, juicy, easy, lucrative animation project fall into everybody's laps.

Giuseppe would be perfect for something I'm doing. Is he in use elsewhere? Can I license him?

I knew A:M has features I haven't run across yet, but I didn't know Eddie could play the piano.