I posted my first bootcamp project in the New Users Forum. I guess I still considered myself a new user and had to gather courage to leave the security of that forum. Dhar suggested that since the bootcamp was "down the hall to the right (students), and not in the New Users Forum, that I should pack up and get my XXX down here (these Gunnies are tough here). Anyway, I packed my seabag, pulled on my boots and here I am.


Project 1: make a cycle of a ball bouncing in place.

[attachmentid=20781]

Project 2: drop two balls of different materials/weight, let their bouncing come to a stop

[attachmentid=20782]

JohnnyB

Alright, you made it, now grab a chair and let's see what you got.<grrrrrrrrr (the sound of a 16 mm projector>

OK, not a bad start at all, you got your squetches in the right places, the ball looks elastic/rubbery. Very good.

But the bounce isn't convincing, is it? That's probably because the speed at which the ball falls and bounces back up are all uniform. In other words gravity doesn't seem to play a role in the bounce. The way gravity affects a falling ball is that it increases in speed as it falls, hits the ground at maximum speed. Then as it bounces up it goes at full speed then slows down as it reaches its highest point then slowly accelerates back down, and so on and so forth.

BTW, recommended reading for this exercise (and the rest of the bootcamp) is Williams, Richard. "The Animator's Survival Kit"

Hey JohnnyB, welcome to boot camp soldier!
(you'll hopefully find this area of the forum just as friendly as the new users area, it is after-all the same forum, so you'll have a lot of the same people watching your progress and helping you along the way)
I know Rodney checks up on all the areas of the forum, so you need not worry about losing his help and advice!
There are also quite a few animation mentor students lurking the forums, who tend to comment on student boot camps who are of tremendous help.
So you made a wonderful decision to move your camp site to here!

Alright, on to critiquing......
Dhar basically hit it on the nose, these balls need some gravity factored into their bounce. (if you don't have the "Animator's Survival Kit" another great resource is to just browse previous students abc threads, that helped me tremendously, there's a lot of information in those things). Also besides the unity in the bounces, they also tend to reach the same height at the top of the bounce....example in your first project- the ball bounces to the same height every time and then all of a sudden stops...you might try putting in a progression
where each bounce gradually gets lower toward the ground. Just think if you were to bounce a ball once and let it settle to a stop on it's own, you would notice how every bounce doesn't bounce as high as the previous one (you could even try it right now if you want, that would help, observation is one of the best resources).
Your Squash and Stretch seem to be very good on your first project however, that looked really nice!

(I'll comment later on project 2, for now I must go to bed), but good job and keep up all of the hard work!

Teresa

How's this? I built flat-topped parabolas with pointed bottoms (bounce points) between parabolas, then I decayed the height and length of the parabolas over time. Is that the idea?

[attachmentid=20801]

It looks great to me Sarge

JohnnyB

Now might be a good time to think about compression in your films. As I write this I am waiting for yours to download. When you render to a file, try changing the setting under quick time movie to serenson video 3. It will reduce the size of the file and alo people to download it faster.

As far as the animation goes(see it downloaded while I was typing)
The contact frame on the first bounce is only one frame. I would be tempted to have the first point of contact on the ground, the next squished and maybe one other before it leaps into space again.

Also the parabolas are a good idea, but the first drop takes 15 frames, then the first ascent takes the same 15 frames,then 12 down,12 up,13 down,9 up,10 descending, 8 up, 8down etc. Take the first descent. The ball is falling to the ground. That is gravity is acelerating the ball at about 10mss. So the frames should be slightly further apart until the ball makes contact with the ground. Then the ground stops the motion, wich results in a squash pose as some of the energy is absorbed, usually over 3 frames. Then the ball acelerates away from the ground (further apart again) until it reaches the apex. Since there is no way it could have the same amount of energy as it started, the ball doesnt travel to the same height, or take as long as it did on the way down.
So if it took 15 frames to drop, then it probably would bounce up for 13 or 12 frames, then fall for 11 then up for 9/10 etc etc.

I would decay the height and length of the parabolas over time, but make the apex of each nearer the following bounce point.
Chris

Much better JohnnyB, but watch that first drop. The ball looks like it's descending at a constant velocity. Make it accelerate as it reaches the ground and you'll have yourself a convincing bounce.

Hey Johnny,

Nice improvement, the spacing is much better and the S & S looks good. You seem to have a little too much hang time. I think you are decelerating at the top over too many frames and the center of gravity (CG) should really only hang at the top for one frame. You can almost take out every other frame at the top and make the crest one frame and I think it will reduce the floatieness. Just a reminder, the number of frames going down should roughly equal the number going up, reduce the number on the next up cycle to show decay. Nice job!

Bruce

Spacing

OK guys. How's this?

Teresa- The bounces diminish over time. (Don't even look at my project 2.)
Dhar- I've adjusted the first drop to accelerate on the way down.
Chris- It's compressed with the serensen video 3. and I added 2 frames on the ground (one for landing, one for squashing, and one for take-off. I speeded up the down side of each bounce by shifting the peak slightly to the right.
Bruce- I diminished my hang time so it looks less like a glider.

Boy! all of this just to make a ball bounce I think it was easier learning to dribble when I was 6.

It does get better every time. (I think!)

[attachmentid=20827]

Yes, it definitely gets better every time! You're doing great! It just takes practice, practice, practice!

Notes (on your last post):
-on the first bounce, you have the ball contacting the ground for 3 frames....try to only have it touch those first two and remove that third one, because after that squash the ball is going to leave the floor and bounce up with almost as much force as it came down with.
-as it is right now, on the ball's trip back into the air (after the first bounce) all the increments are equal again. Maybe try having similar increments to how the ball comes down, but opposite (if you know what I mean). I imagine this looks like a diagonal straight line on your graph...try to have it be curved so the bounce is sorta more of a parabola.
(actually these two comments apply to all the bounces)
You've greatly improved the fall, it looks good!
Anyways, can't wait to see it with those little tweaks!
Great job!

Like this?

[attachmentid=20838]

Hi JohnnyB,

I think this is doing well.

The thing that sticks out to me is that the last huge jump to the ground on every fall. It's almost as if some invisible hand has swatted the ball down at the last instant rather than let the ball fall on it's own.

Fine tuning your motion graph would fix that.

I made some marks on the movie about that and some other notes.

[attachmentid=20843]

I just tracked the center of the ball. If we tracked the leading edge we'd probably find some other interesting motion. Sometime at the top of a bounce , the top edge of the ball looks like it's stuck in space, like it's hit a ceiling. It's hard to get every thing moving exactly right.

But overall this is a good bounce exercise.

Wow! your marks and notes really helped me visualize the bounce and relate them to frames. You guys go to fantastic lengths to help. That was really cool! How'd you do that?

So, I realigned my parabolas. Better?

[attachmentid=20844]

Everybody seems to agree that the ball should accelerate (not be linear) on each descent, and fall faster than it takes off. What you're saying is that it shouldn't jump so far in one frame.

Project 2: drop two balls of different materials/weight, let their bouncing come to a stop

[attachmentid=20857]

I love it when Robert makes critiques like that! It's so helpful!

JohnnyB your exercises look grrrrrrreat! You've got it now!
The only little very minor minor itsy bitsy comment I have (which has nothing to do with the bouncing) is why do the balls roll to one side at the end? If the balls were bouncing across the screen that would make more sense, but as it is now it's a bit out of place. Other than that little crit I have nothing else to say....I think it's time for the next exercise!!!

Thank you! Thank you! Thank you!

The roll. It just seems more natural to me. Whenever I've bounced a ball (in the real world), it rolls away when it's stops bouncing and I have to chase it. I guess it's just my thing.

Nice job!

Thanks to all of you that gave me so much help and support through those first two bounces. Here comes my next one, then I'll be on the road for the next three days. See ya when I get back.

Exercise 3: Bounce two not alive balls of different materials across the screen (watch your arcs)

[attachmentid=20891]

Two things I can think of that might improve this is
1) Try rotating the axis of sqash and stretch with the direction of the ball slightly. Right now the squash and stretch are prfectly vertical and should be tilted in the direction of motion.

2) beware of texturing the balls this if used as a learning process should first be done with nothing that show rotation of the ball. Once you add texture to the ball you have to add rotation to make it work. The smaller ball appears to have some near the end put not enought while the basket ball has nothing.

I would take the textures off first for this exe then if you wish try another one with textures but rotate the ball. It really adds to the exe.

In the animation of the two balls moving accross the screen, the x velocity should remain pretty much constant. The balls also look like they are being thrusted downwards rather than gradually gaining velocity from the acceleration of gravity.





Check out this website.
http://www.awbenson.com/tutorials/Carrara_...ball-bounce.htm

Even though it's not in A:M, it still demonstrates the fundamentals of squishing on impact.

I'd say PF Mark and Matt have made good comments.


-the direction of the stretch is very important to show where the force is aimed. If it's aimed straight up... what's making the ball move sideways?

And a big sign of this will show in the curve for vertical motion. If the apex of the parabola is squished closer to one of the ground keyframes, then the up and down motion won't be equal, and they really should be the same within each hop. There may be cartoon situation in which you'd depart from that, but here in bounce class we want to master generic first.

Also the blue ball seems to float down at first, then bounce heavily. The heavy part isn't bad, it's that first float that seems odd.


Edit: I just tracked the orange ball. Several part seem to evenly spaced and/or linear, particularly the sections I marked with a line. This is part of what creates the "un natural" look.

[attachmentid=20967]

Hi Johnny!

I think most of the major points have been covered by everyone else for this exercise. It's looking great, but strongly I agree everyone else that you need to point the ball in the direction of the arc. Also I assume that you're trying to make the ball coming from the right seem heavy. If so, try lessening the squash and stretch, or even delete the S&S entirely and see what the result is.

Keep it up!
-Josh

I think I've incorporated everyones suggestions in this next attempt:

[attachmentid=21002]

Hey guys! What's the difference between a not-alive ball and an alive ball?

JohnnyB

drop a ball with a tail, let its' bouncing come to a stop

[attachmentid=21018]

Now I've got whiplash! Reporting for sick call.

JohnnyB

Really loved that Bounce Across with the blue and black balls.
The tail on the other hand... um... er... a bit stiff?

You've been creating really nice, expressive animation.
I hope you are getting involved in TWO!
You look like you are up to some serious (even comical?) character animation.

How do I get involved?

Launch a PM or email to Rhett Jackson (Jandals) and he should get you on your way.

Starting Monday (tomorrow) it looks like there is going to be some very serious animation going on.

Follow through: move a ball with a tail back and forth across the top of the screen, focus on the follow through of the tail.

[attachmentid=21065]

Rodney: I've tried to make the tail less stiff (rebuilt it actually).

JohnnyB

John,
You've got some nice touches in there.
I especially like the 'push off' as the ball* 'swims' toward the wall.

*You must have hit success here as I immediately felt the character and instinctively wanted to refer to 'the ball' as 'he'.

Some minor strangeness/stiffness still present in a few spots but overall... very nice.
Nice use of color to accentuate mood too.

Tail bounce: bounce 2 balls with tails across the screen.

[attachmentid=21172]

You're really getting creative with this. I love it

Hop: take a juicebox or Pixar lamp and have it hop across the screen, remember follow through.

[attachmentid=21503]

Clever composition. Things happning in your office?

Ok, so now that you thought of follow thru, it's time to think about balance. Does the lamp look balanced when jumping? Is that how you'd jump? With your body so far forward I'd expect you'd fall flat on your face, no?

Refer to the Survival guide on page 213 to get an idea about the dynamics of jumps.

Yeah! It happened in my office. I was the guilty one that left the juice box on the floor. Cost me a nice Pixar lamp.

Oh well... How's this?

[attachmentid=21544]

The diagrams on page 213 show the jumper leaning far into the jump then bunching up and leaning back for the downward thrust. It's tricky translating that to a Pixar lamp.

But you can translate that! Even though the anatomy of the man and the lamp are different the basic stretch and crouch poses are the essential element.

[attachmentid=21545]

The first thing I notice about your lamp jumps is that it's not fully extending before it leaves the ground. That's a must. The first jump leaves the desktop in the crouch position. Impossible!

Do this as an experiment... go outside and try jumping with your knees bent. Don't let them straighten out. How well does it work? How about if you let your legs push all the way untill they are straight?

How's this?

[attachmentid=21563]

I love the creativity of this hehe.

One thing I noticed is in the initial jump of the lamp off of the table. It is not falling at an accelerated rate which gives the landing a "floaty" sense. It's like the lamp is floating to the floor. You lose the illusion of weight. In fact in all the following jumps except the quick jump right after the lamp directs the juice box, I'm not getting the sense of weight. If you accelerate the landing part just a little on each jump, I think it will show more weight.

Also,

You should add a little bit of squash to the juice box at the bottom of it's hop. It will look better.

Finally,

When the lamp jumps up (I assume to try to get back on the table), you should make the impact more pronounced by stopping it cold on impact with the table or you could even lift the table a little maybe with a little follow through of the lamp head and then accelerate it back maintaining the increased acceleration until the impact with the ground which causes the lamp to break apart.

It's coming along real nicely!!!! Keep up the good work!

I hope this helps

George

Great tips George! Thank you. It gets better with each new suggestion I work out. It's actually getting to look like I visualized it. How's this?

[attachmentid=21569]

Well, compared to the drawings... is the lamp really extended when it leaves the desk? Is it really crouched up at the top of the jump? Is it really extended again when it touches the ground?

Can anyone really jump with their knees bent that much? It can not be done. That is why those jumps look odd still.

Also consider where the motion is going in that first leap. It's flinging its weight both up and forward at about a 45 degree angle from the base, but then in the first frame off the desk it has only moved directly upward, not forward at all. What force made it change direction? Momentum would make the object leap off in the same general direction it was pushing. That's another reason the jump looks odd; the audience can't see any force applied that would make it change.

Thanks I'm glad I could help

Looking much better!!!!

I love that juice box

I don't know if you're purposely doing this or not (my hunch is you're not) but there appears to be spline drift as the lamp takes off and lands. Now it also looks like it may be sliding due to inertia of the land but my suggestion would be if you didn't do this purposely and you know how to adjust the animation through the splines mode of the timeline, to go into the spline view and look at the different movement keys (I.E x,y,z translation and x,y,z,w rotation).

If you are in spline view mode and in the properties open your animation keep expanding till you get to the "transform,rotation and transform,transform. If you now click on these, it will only show the spline of the one you click on instead of all the splines. It is here where you can really refine your animation.

I learned from Jeff Lew's DVD (if you don't have it, it's a must have) to always check the spline arcs to see if you have any curves or extra keys that could be affecting the animation (I.E slowing down a movement too soon or too late, spline drift, etc)

If you are not familiar with the spline mode of the timeline, I highly suggest you learn this. It's my last tweak I always do. If you get the arcs just right, your animation will show it big time.

It's really shaping up! Good Work

George

How's this? The all-new improved HoppinJohn2.


[attachmentid=21917]

John,
That is much improved.
Robert may break out his tracking dots (if he doesn't I'll try to do that tonight).
That would help perfect the shot.

The part that stands out to me is the base of the lamp.
Right now it stays horizontal throughout the sequence.
If you were to concentrate on that base alone and tilt it appropriately you could add some nice anticipation... followthrough... overshoot.

You've come a long way from your first take! Keep it up.

Like this?

[attachmentid=21923]

For checking arcs, you might use the "ArcTracker" located here...it is very cool. I reposted it on July 20th. It was made by Ben de Leeuw and John Aquino.

I, of course, am still trying to unload all those worthless dots I was stuck with after the dot com crash.

[attachmentid=21926]


Johnny: when I talk about "evenly dividing the weight", a 60/40 split would probably be more realistic than a 50/50 split.

David
I downloaded the ArcTracker, but I can't figure out how to use it. I imported it as a model into my choreography window where it seems to be a single bone.
John

Wow Robert! Thank you!

Now I have to figure out how to track that center of gravity through the arcs. I downloaded the ArcTracker that David mentioned but haven't yet figured out how to use it.

You know the help and support of this community makes the AM program worth far more than Hash charges. This is like a college education in animation. I had no idea what came with that little box I bought at a computer show with my last few bucks (honey I just had to have this program ).

I hope that others who follow me will find my bootcamp thread. It's so full of information, from the first dots on my bouncing ball through this hopping lamp analysis movie.

Thank you again!

It wouldn't be too hard to "translate to" constrain a null to both the light and the base. Set the enforcements to 50/50 or 60/40 or whatever and it would float at an appropriate midpoint.

Not that you would be keyframing that null; it would just be a visual guide of where your c.o.g. was at any moment.

That pixar-style lamp is probably one of the few characters that that could be rigged on. Most have shapes too complex to find the c.o.g. so easily. Hmmm....

I am clueless about how to "translate to" constrain a null to both the light and the base, and how to set enforcement to 60/40.

Sounds like a question for the rigging forum.

OK! Got my null constrained (60% base 40% lamp shade. Tracked the arcs of the null with the arc tracker (See Page 3 near the bottom. Thank you David). I think it's looking better.

[attachmentid=22031]

I did pose my problems to the rigging forum and got a quick reply from Mike Fitz:

Hey, I think that should be pretty easy.
But since I don't know what the lamp looks like, here is a basic suggestion, if I'm reading what you say correctly.

I wouls place 2 bones in the model. One at the base of the lamp, the other at the location that would be at the location Robcat mentioned.
Then add a null.
in an action window, or pose window, right click on the null and assign a new translate to constriant to the base bone. then do this process again and translate it to the other bone.
the null will find the exact middle location between the 2 bones.


Me: What's a null?

While in "Bones" mode in the modelling window, right click, and navigate the drop down menu
till you see add new null.
It's basically a Bone that you can not assign any control points to.
Its used as targets to aim at, targets to grab for pertinant points on your rig, like as a foot target, hand target, eye targets. Stuff like that.

Mike Fitz
www.3dartz.com

wadya think?