robcat2075

Now they're lopsided again. it's taking longer to go up than to come back down. I recall Jeff Lew saying on his DVD "You will live in the curve editor". I had one AnimationMentor mentor who didn't mess with curves much, but that's because he was keying everything on every single frame. I really think that's the hard way to do it.

My bouncing ball up there... that took me 36 hours to get there. But that was my first time. Once you "get it", you can do these right in a few minutes. That last one is much better. But consider ... the first and second fall from the peak to the ground are both 10 frames even though they are not falling from the same height. The last hop and the next to the last hop are almost the same duration even though they do not reach nearly the same height. Note that falling from 2X a height does not take 2X the time. Gravity is complicated.

or not

I'd prefer a contest somewhat more focused than "anything". It needs some boundary so that there is some possibility of comparing apples to apples when we vote on them.

Certainly the ball is not slowing down as it approaches the ground and that is an improvement! - how can the ball take the same amount of time for high bounces as low ones? It can't. - your Y channels look like very flat half ovals rather than parabolas which is packing too much motion into the frames around the ground contact. Gravity has a way of adding about the same percentage of drop to every frame and a parabola in the Y channel mimics this well. This is different than lingering at the top and then racing to the ground. The acceleration of gravity is a very particular look that you want to get to recognize. -render these without motion blur so it is easier to see the location of the ball at each frame.

The assignment was to have a rigid ball bounce several times, no squash, no stretch, no personality, so that's what I tried to do first and at first the ball just moved from left to right as it bounced. The assignment required no more than that. It didn't even require that the ball move horizontally, it could bounce in place. I recall a stated time limit of 5 seconds so i was working to fit in that. the contacts and the peaks both. Not that those looked right at first, but fortunately I already knew about what channels did (not very well explained in the AnimationMentor lecture. Their first lecture on Bouncing Ball was easily their most confusing.) so i started tinkering with the shape and size of the vertical channel until the bouncing looked right. This is where i discovered the whole parabola thing and how to make an object accelerate all the way until it hits the ground, like gravity does. (i'm not sure the lecture even mentioned gravity. I got the sense that some of these animators doing the lecturing had never been in a science class. ) I also recall looking at some other students' WIPs and thinking theirs didn't look right, like maybe the ball was dropping faster than it rose up. Then I looked at mine again and realized I had the same problem! That's where I discovered the symmetrical parabola thing and how just a little deviation from that would give a very different feel to the flight of the ball. For example a ball that fell faster than it rose looked like a basketball being dribbled by an invisible hand. At this point I had a keyframe on the horizontal channel for every one on the vertical, but i noticed the result was pretty much a straight line and if i deleted the intermediate Cps I'd get a perfectly straight line. I realized that the horizontal momentum of a ball is rather independent of its vertical motion, so i could leave my up and down bouncing alone and just change the horizontal channel to make the ball bounce in place, or ping pong, or bounce out of the frame, or reverse when it hit the side of the frame like it hit a wall... then I realized how easy it would be to add a few blocks in the scene and have the ball appear to bounce off them. But I made certain to full fill the minimum terms of the assignment and that's why I cut the animation loop to start at the top of the ball's first arc. The ball falls and bounces several times...assignment done... the ball getting hit by the block and deflecting off the wedge, I cut and pasted that onto the end. But all this channel editing stuff is putting the cart before the horse. Bad mechanics is like alcohol, you have to recognize you have a problem before you can fix it. If you look at your bouncing ball and don't see that it's slowing down before it hits the ground, how will you know that that needs fixing in the channels? Teaching that "observation" thing is probably the hardest. Shhh! Don't tell anyone about the parabolas.

that looks wonderful, Jesse! Congratulations on getting that done!

Here are the translation channels from :

Here's how I'm currently set for names for credits. I'm going to use a "Real Name (forumname)" format... If your credit is missing or you want it different... now's the time to speak up! I would like to have the ???'s filled in If your credit is here as you want it to be, no response it needed. (in no particular order) Bruce Monahan (wedgeeguy) Chris Dailey (??? mouseman) (dblhelix) Sebastian Pfeifer (thefreshestever) Gerry Mooney (Gerry) Glenn Watson ?? (Glenn) John Lemke (johnl3D) Matt Campbell (John Bigboote) Nancy Gormezano (NancyGormezano) Paul V. Harris (fae alba) ???Tim Benefield (tbenefi33) Jason Hess (thejobe) Tim Roberts (number) Gene Thompson (TheSpleen) Robert Holmén (robcat2075) ???Jim DeVivo (jimd) Mark Skodacek (mtpeak2)

Start here http://www.hash.com/forums/index.php?showforum=36 the "Cornell Box" discussion should be your first stop. Radiosity is a very time consuming process suited more for stills than animation.

i just get a 320KB file

I agree with Rodney that Bouncing Ball is a complex challenge hidden in a simple task.

No, that was for you Gerry. I just get a dead link when I try to download the 32-bit v16. Where is the live link at?

Are you sure you were using the 32-bit version?

"Alarms should go off..." IlidrakeBounceChannelB.mov

the number one thing that bothers me is that the mouth is too complex. Too many bumps. I think you'd be way ahead with a simple horizontal mouth opening that can clearly do the graphic shapes of smile and frown and "oo" and "ah" and "ee". Also it's rather similar to "Bishop", the AnimationMentor character. There are several near-Bishop characters that people are using at the 11 second club and I think they look grim. Do something that is definitely not-Bishop but still graphically simple.

That has some classic problems. On my to-do list is something called Bouncing Ball Boot Camp. In the mean time, go to the very beginning and do a non-squishy bouncing ball. A rigid sphere about the size of a basket ball falling from about six feet. Make that bounce at least four times. Important concept: Falling objects do not slow down if they are approaching the ground. Falling objects do not slow down because they are approaching the ground. Falling objects do not slow down when they are approaching the ground. Falling objects do not slow down. This is central to bouncing Ball and all Body Mechanics after that which is why Bouncing ball is a good place to start.

I especially liked the auto entry. "Bile Flavor"?

Never mind, I'm about to make one myself.

That I've already tried but couldn't detect any linking happening.

Most of the people on the forum now have just basic animation chops. Something that is skill building would be in order. Long scenes, dialog, those are advanced topics.

A revisit of the "sequential art" (comic books, graphic novels) contest might be interesting. -One or two "page" limit. maybe a standard comic book page format. -maybe a required theme -all A:M of course, except you could use a paint program for lettering, word ballons and assembling all the frames into one page that would emphasize modeling and posing and storytelling and not so much animation.

And it only needs to be the front side, the display side. The back won't be seen.

Or similar handheld device with large screen? It doesn't need to be exquisitely detailed, I need it to "show" one of the entries that is a bit non-standard.

I can't detect that the .sinfo file is ever loaded by A:M and there's no way to manually load it. I suppose this is in a tech talk somewhere.