bouncing ball exercises

[itsjustme]

I took a very quick look so far...I didn't see the problem until I set my project to 30 frames per second. The project should be 24 frames per second. See if that clears things up, Rodney.

 

I'm still going to go over it to see if there is any error due to the moving of the "pivot" bone, so there still may be an issue to be dealt with.

[robcat2075]

An intermediate exercise between the basic bounce and the obstacle course might the "light ball, heavy ball" test.

 

That's really a terrible name, since light balls do not necessarily bounce more than heavy balls. It would be better called "bouncy and not-very-bouncy balls". The object is to correctly show balls of different qualities.

 

Something else to try is a ball that takes more than one frame to squash and stretch itself off the ground back off the ground. I'd envision this as a ball larger and more massive than a soccer ball.

[Wildsided]

Sounds like a loada balls

 

Thankyou, thankyou, I'll be here all week. Remember to tip your waitress

[itsjustme]

An intermediate exercise between the basic bounce and the obstacle course might the "light ball, heavy ball" test.

 

That's really a terrible name, since light balls do not necessarily bounce more than heavy balls. It would be better called "bouncy and not-very-bouncy balls". The object is to correctly show balls of different qualities.

 

Something else to try is a ball that takes more than one frame to squash and stretch itself off the ground back off the ground. I'd envision this as a ball larger and more massive than a soccer ball.

 

I'll throw these together as well. Thanks, Robert!

[Rodney]

David,

Last week I spent some time with your collection and had a lot of fun messing around with them.

By way of feedback there were some things I wanted to discuss with you but I admit I have risen above my level of competence in that I cannot effectively communicate what I am seeing. I tried with a few animated sequences and failed miserably.

 

What I am looking at now is an approach to the theme of these bouncing balls from a different perspective entirely. Time will tell if I'm up to that challenge either.

 

This is just to let you know that your project files here are greatly appreciated.

[itsjustme]

This is just to let you know that your project files here are greatly appreciated.

 

Not a problem, Rodney. I hope they will help people avoid the numerous mistakes I make...I'm learning quite a bit from them myself.

[itsjustme]

Okay, I managed to get back to this...balls with different levels of bounciness. I'll put together the one for the ball that takes more contact frames next. Included in the ZIP are the models used, Choreographies, renders (24 frames per second) and notes. I didn't get a chance to render a final render, but if I made mistakes that need correcting, I'll add it for the update.

bouncing_ball_exercises_03_with_notes.zip

[robcat2075]

Okay, I managed to get back to this...balls with different levels of bounciness. I'll put together the one for the ball that takes more contact frames next. Included in the ZIP are the models used, Choreographies, renders (24 frames per second) and notes. I didn't get a chance to render a final render, but if I made mistakes that need correcting, I'll add it for the update.

 

Hey, David, I think those all look good. It was fun to watch each ball individually. The squetch versions were tastefully restrained.

 

Personally, I'm doubtful that a ping-pong ball is the least bouncy of all. Maybe your needs some air.

[itsjustme]

Hey, David, I think those all look good. It was fun to fun each ball individually. The squetch versions were tastefully restrained.

 

Very cool. I thought I might have gone a little much on the squetch of the tennis ball.

 

Personally, I'm doubtful that a ping-pong ball is the least bouncy of all. Maybe your needs some air.

 

I had my doubts as well. However, since I've never actually tested it I figured whoever put together that chart I used would know better than me. It wouldn't be the first surprising thing for me to find out. I'll have to double-check that info...it also wouldn't be the first error I've found in reference material if it turns out to be incorrect.

[robcat2075]

chart?

+

Edit: I found the link in your page.

 

A baseball more bouncy than a ping-pong ball? I think he's off on that one.

 

But I don't have either one to test.

[itsjustme]

chart?

+

Edit: I found the link in your page.

 

A baseball more bouncy than a ping-pong ball? I think he's off on that one.

 

But I don't have either one to test.

 

He may be...I'm pretty sure he actually bounces a baseball on that page (unless I'm not remembering correctly), but he didn't actually show the others. There is bound to be some more examples on Youtube...I'll take a look late tonight.

[robcat2075]

Ping-pong may have changed substantially recently...

 

From the always-reliable Wikipedia:

 

After the 2000 Summer Olympics in Sydney, the International Table Tennis Federation instituted several rules changes aimed at making table tennis more viable as a televised spectator sport.[12] First, the older 38 mm balls were officially replaced by 40 mm balls in 2000.[6][13] This increased the ball's air resistance and effectively slowed down the game. ...

 

Variants of the sport have recently emerged. "Large-ball" table tennis uses a 44 mm ball, which slows down the game significantly. This has seen some acceptance by players who have a hard time with the extreme spins and speeds of the 40 mm game.

 

 

When I was a boy we used 38mm balls that went by too fast to hit... and we liked it!

[itsjustme]

Wellsir, I think I misunderstood the chart...that'll teach me to assume I know what I'm looking at. It appears to have been the "coefficient of restitution" (speed up / speed down) and not the height of the bounce that it compares. Yeah, I'm feeling bone-headed at the moment.

 

First, I did a very rough check of two ping-pong ball bounce videos on Youtube. I approximated

as about a 72% bounce using the second drop. Then,

gave me a rough estimate of 83%. I thought about just averaging those two numbers and make it 77.5%, but the angle of the camera and the interlacing of the video both made some of my figuring suspect. I looked around a little more and found this page which states:

 

The international rules specify that the game is played with a light 2.7 gram, 40 mm diameter ball. Generally, it is the most-used ball. The rules say that the ball shall bounce up 23 cm when dropped from a height of 30 cm thereby having a coefficient of restitution of 0.88.

 

So, 23/30...which would be 76% of the drop height for the ping-pong ball.

 

This page has also has the bounce heights for a golf ball (73.6%) and tennis ball (50.6%).

 

I'll do a little more research tomorrow and correct all of the ball bounces. Thanks for catching that, Robert!

[itsjustme]

Here's the updated version of the 03 exercise with corrected bounciness on everything. I also reduced the squetch a little on the tennis ball, added some links and explanation to the notes and a final render (it could have been higher quality, but I decided to tweak it a little to get a faster render).

bouncing_ball_exercises_03_update_with_notes.zip

[itsjustme]

Here's the 04 exercise...an exercise ball with additional contact frames (24 frames per second). Included in the ZIP are the Choreographies, model used (with material), stills, renders and notes.

bouncing_ball_exercises_04_with_notes.zip

[robcat2075]

Hi David, sorry, i meant to get to this sooner.

 

Here are some notes on the squash. Let me know if this explanation makes sense.

 

HowMuchSquetchH150.mov

[itsjustme]

Here are some notes on the squash. Let me know if this explanation makes sense.

 

Makes perfect sense, Robert...several things I didn't take into account. Thanks for taking a look at it. I'll post an updated version in the next few days.

[MMZ_TimeLord]

I found my AM Bootcamp Project while looking for something else last night... Here are my two favorite results from that project.

 

One is a two, non-alive, ball bounce against a perpendicular surface.

 

TwoBallsBounce.mp4

 

Next is a two, alive, ball interaction sequence.

 

TwoBallsInteract.mp4

 

If anyone wants the project I've included it here. Some of the results, to me, could still use lots of polish.

 

AM_Bootcamp_Level1_v16_.prj

 

Enjoy!

[Rodney]

Excellent analysis Robert. Very nice.

[robcat2075]

Something to try... what's the longest the ball can contact the ground and plausibly bounce up again?

[robcat2075]

I will note that if you think my theory of squetch through too much there's an apparent problem:

 

If the ball leaps back up in the air because it is rapidly unsquashing itself... how does a rigid, non-squishy ball like a bowling ball bounce at all? Why doesn't it just stop dead when it hits the ground?

 

One excuse might be that it does squetch a tiny bit and unsquetches so fast that it is able to fling its mass back off the ground. But I dont' know if that has any reality in physics.

[Rodney]

You've given us a lot to ponder here.

 

I stared at that element of your critique for awhile and admit that I didn't follow it fully. I caught the important point of it in that you correctly identified the disparity in acceleration throughout. I think you've mostly got it right on the deceleration after contact as well but, as you state, we have to consider what the ball is made of in order to know more. I confess that you've got me thinking about the likelyhood of stored energy being able to accelerate an object. It certainly can happen but I think it would be due to the stored energy in the object the ball is making contact with?

 

I like how you talk about the compression of atoms as that helps to keep the idea of the object maintaining the ball's mass throughout. Some of the energy is transferred to the object the ball hits and the remainder is retained by the ball. In the case where the ball is made of fragile material it can lose part of it's mass but then we'd see those chunks of the ball separating from the ball and dispersing. The whole idea of Energy Transfer is ever present even where there is no obvious effect on the compression/extension of the ball.

[robcat2075]

I confess that you've got me thinking about the likelyhood of stored energy being able to accelerate an object. It certainly can happen but ...

 

Consider the case of hipster tennis players who swat a ball laying dead on the ground and are able to get it bouncing up. They are squashing it down and releasing it suddenly. It's the un-squashing of the ball that gets its center of gravity moving upward, fast enough to carry it into the air.

 

 

 

I think it would be due to the stored energy in the object the ball is making contact with?

 

A trampoline would be an obvious case of "the ground" storing some energy and giving it back to the ball.

 

But for most grounds, like concrete or a bowling alley, it's probably very tiny. As tiny as the bowling ball's was... so i have no idea where all that bounce is coming from!

[robcat2075]

Something else I meant to say...

 

In this theory of squash and stretch, squash and stretch is a form of overlapping action.

 

The ball squashes because the bottom of the ball has stopped at the ground while the top tries to keep moving down.

 

The ball stretches because the top has been forced into the air while the bottom lags behind because of its inertia.

[Rodney]

In this theory of squash and stretch, squash and stretch is a form of overlapping action.

 

Nice thought!

 

Brian Lemay seems to get a lot of flack from the fringe quarters but as a teacher he has a lot of insight into animation. If we were to adapt and put into practice all of his online exercises (in CG) we could reasonably expect to be much better animators. Sadly, I've never purchased any of his books but they are on my 'must get someday' list. He's got enough stuff online (many are examples from the courses teaches) that could keep an animator animating for years.

 

http://www.brianlemay.com/Pages/animations...ook/squash.html

 

What I like about this particular write up is how the idea of squash is related to a character's walkcycle with an emphasis on overlap and weight.

[NancyGormezano]

Something else I meant to say...

 

In this theory of squash and stretch, squash and stretch is a form of overlapping action.

 

The ball squashes because the bottom of the ball has stopped at the ground while the top tries to keep moving down.

 

The ball stretches because the top has been forced into the air while the bottom lags behind because of its inertia.

 

I think I regard squash & stretch (stretch in particular) as to not really being based strictly on physics (conservation of energy, mass, elasticity, PE, KE, force, gravity, etc). The concepts of squishy squashy are used to give personality to the ball. Not reality.

 

So IMO, trying to define it in terms of reality is weird. Any strobed photo of any ball (of most any ball type material) will show that there really isn't much stretching going on - but yes there is some compression on impact. Balls/objects especially don't stretch on the way down. Probably don't deform on the way up all that much either beyond their original shape

 

Regurgitating what I believe might be Keith Lango's thoughts: The concept of exaggerating stretch came about with 2D animation - the ART of shapes in transition being drawn at 10-12fps. The stretch image of a ball (or character) is more like a smear frame, or drawn breakdown/inbetween image or as you say possibly "overlap, ease stylistic behavior" : the image shows where we've been, where we are, and where we're going. It gives the character/ball personality, and also lets us not have to hand draw as many frames.

 

To combine toon behavior in CG, where we can have faster frame rates, accurate textures, shading, lighting, very human models can produce some mighty muddy looking uncomfortable stuff when combined with SQ& ST. If one is going for reality - then probably best to stick to the accurate physics. If you're going for stylistic - it's open to interpretation. And the more the models/environment are presented realistically, the harder it will be for people to suspend belief and accept the style. If you go for looney toon, ANYTHING goes, as long as it's consistent with the toon universe.

 

So I don't think a discussion of moleculer, inertial behavior has any bearing on stretchy squishy squashy toon or even stylistic behavior.

[robcat2075]

At the beginning, we're not doing "personality balls" we're learning to plausibly move mass on screen and the ball is our simplest case to practice with. Personality comes later.

 

Poor body mechanics is the hallmark of poor character animation and good body mechanics starts with being able to plausibly move masses on the screen.

[NancyGormezano]

At the beginning, we're not doing "personality balls" we're learning to plausibly move mass on screen and the ball is our simplest case to practice with. Personality comes later.

 

Poor body mechanics is the hallmark of poor character animation and good body mechanics starts with being able to plausibly move masses on the screen.

 

I think? I agree: exaggerated squash & stretch is not plausible for realistic motion, realistic character animation, body mechanics. As soon as you start squetching, you add personality, interpretation and distorts physical reality. Likewise, molecular physics has nothing to do with toon, stylized motion (in response to your critique of David's squetching).

 

Animation gets muddied mixing toon type behavior & accurate physics. It then starts to become a "taste" issue, where there are many choices.

[robcat2075]

All this talk of squash and stretch ans overlap has very little reality with actual balls. Real balls hardly ever squash and stretch and overlap to the point we see it like we do in cartoons.

 

However, other things in real life DO squash and stretch and overlap very noticeably, but those objects are very complicated to maneuver for a beginner and present too many intersecting issues simultaneously.

 

Take a look at this guy's rhumba. He doesn't get body mechanics and he can't move mass plausibly on the screen. There's no way to even usefully critique that because there's just too much going wrong from the very first frame to ever just tweak it in to rightness. It's painfully obvious from watching that and his other 11sec entries that he doesn't know how to move mass onscreen without it looking fake and awkward.

 

I'll bet you a dollar he's never done a decent bouncing ball. He really needs to be starting out with a far simpler case and a develop a good eye for moving simple things around before he leaps to full human bodies.

 

As is true of us all. So we start out with a rigid bouncing ball just to get "falling" right and then move to a ball given an exaggerated combination of floppiness and elasticity so we can practice with managing a deformable shape in a very simple case. Then we move on to a little bit more.

[NancyGormezano]

However, other things in real life DO squash and stretch and overlap very noticeably, but those objects are very complicated to maneuver for a beginner and present too many intersecting issues simultaneously.

 

Take a look at this guy's rhumba. He doesn't get body mechanics and he can't move mass plausibly on the screen. There's no way to even usefully critique that because there's just too much going wrong from the very first frame to ever just tweak it in to rightness. ....

 

As is true of us all. So we start out with a rigid bouncing ball just to get "falling" right and then move to a ball given an exaggerated combination of floppiness and elasticity so we can practice with managing a deformable shape in a very simple case. Then we move on to a little bit more.

 

I am not disputing that it's a good idea to start with a bouncing ball, and then introducing squash & stretch at some point.

 

And I do agree that in real life - deformable, elastic bodies, humans will stretch and squash realistically, and the body parts will move with "realistic overlap" to satisfy the reality of the physics of motion, energy.

 

Realistic and even plausible 3D/CG is about moving parts (head, torso, arms, etc) in some sequence, over some arc/path, over some time in some plausible way. Deforming realisticly modeled, or even stylized humanoids in a excessively, exaggerated toon type way is a hard sell. It looks odd. However, deforming them in a realistic way can usually result in something more plausible.

 

On the other hand, 2D, traditionally has been about transforming drawn whole shapes, silhouettes (not parts), not necessarily (and not usually) in a realistic way. Excessive deformation of more "toon" proportioned shaped characters is a much easier sell.

 

What I was suggesting is that explaining exaggerated Squash & stretch of a bouncing ball in terms of molecular energy is not appropriate (as in your critique). Nor as you suggested above:

 

The ball squashes because the bottom of the ball has stopped at the ground while the top tries to keep moving down.

 

The ball stretches because the top has been forced into the air while the bottom lags behind because of its inertia.

 

It muddies the concepts. Explaining it as a stylization, illusion, exaggeration to help sell the movement, compensation for frame rates, persistance of vision even, would be more "plausible", IMO.

[robcat2075]

I think I said "imagine" it as made of separate molecules. I'm pretty sure I didn't mention molecular energy.

 

I could have said imagine it as separate pieces of pigskin sewn together, or imagine it as separate plastic pellets melted and vacuformed together in a ball shape or imagine it as separate spoonfuls of jello in a containing plastic bag...

 

All of those parts, even the molecules, have inertia and momentum.

 

When the bottom of the ball contacts the ground, the molecules or pigskin or pellets or jellos at the top of the ball do not instantly halt when bottom does because they are only remotely connected to the bottom. Their inertia makes them continue down.

 

Neither do they instantly fling themselves down just because the bottom of the ball has contacted the ground. The inertia that makes them continues down after the bottom of the ball has stopped also prevents them from suddenly zooming down faster than they had before (and there's no force in this bouncing ball situation that is even trying to make them do this).

 

Of course all these molecules or skins or pellets are connected to each other (not so much the jellos) and those connections get stretched until they can't stretch any more and the ball has to stop squashing and start returning to its original shape. If that return is fast enough, the molecules or pellets or skins at the top get enough momentum to not only carry themselves higher but also eventually yank their loosely connected fellow traveling molecules or pellets or skins at the bottom of the ball with them.

 

All of this is to get the animator away from thinking of the model as a rigid shell that deforms itself when it is near the ground simply because we have a slider to do that or because there's a diagram in Preston Blair that says that and start thinking of it as an real object that he needs to show is shaped and moved by the forces of inertia and momentum (and obstruction from other objects like the ground).

[NancyGormezano]

All I've learned from your post is that either you're not able to get my point or else you are not willing to get my point. You win.

[robcat2075]

Your point is that explaining the deformation in terms of real-world mass muddies the concept because it's really about stylization. I disagree.

 

I think my way of explaining it terms of real world mass gives the student the best chance to take out of this exercise insight that can be carried forward to the more complicated cases of moving things around.

 

Most character animation doesn't involve things bouncing. It always involves moving masses around. I want the new animator to start thinking about those masses as soon as possible. Once you understand how they move you will know where to exaggerate that motion for stylized effect.

 

At AnimationMentor I saw a lot of fellow students NOT get body mechanics and I think it was because school was dealing in abstracted rules like "it looks better for objects to move in arcs" rather than real explanations like "objects move in arcs because the have mass and can't change direction instantly" or "a hand moves in arcs because it's swinging on the end of a bone and not moving by itself." (After two quarters the faculty knew they weren't getting body mechanics across, but they were people who intuitively understood motion and not quite able to explain how they understood it.)

[Rodney]

I don't know how applicable this is but what appealed to me in Robert's use of the terminology he used (i.e. atoms) is how that augmented my perception of the ball in several ways: 1) in comparison to a plausible or relative reality 2) consistency in the object's solidity.

 

In the first case, I got an immediate sense that the ball in question was 'real'. Not really real but within the framework of the scene. I began to sense, from it's movement alone, what it consisted of within the presented 'reality'.

 

The second relates to the first in that as long as that particular reality is consistent I will not question the laws of physics within the context of the scene. This is that willing suspension of disbelief we are all trying to achieve in our animation.

 

These things allow even the most cartoonish application of the bouncing ball to make sense to audiences as the ball will not be seen as suddenly changing what it is made of for no good reason. This consistency (I believe referred to as the Animation Principle of Solidity) is appropriate in this context and while complicating the matter, real world physics can play an important role in this. Why it is specifically important in CG is because the principle was first given as "Solid Drawing" and in CG, drawing isn't not something we are likely to see as we are animating on a computer screen. Therefore, animators tend to bypass this principle entirely because they see no way to apply 'Solid Drawing' in CG.

 

I believe this is why thinking in terms of atoms (cartoon or otherwise) resonates with me.

 

Hope that makes sense.

[MMZ_TimeLord]

I have to agree with robcat in that a lot of animators fail to see or grasp the significance of "weight" or "mass" in a character's motion.

 

They may have a very technically correct motion, but no feeling of mass, flexibility (joints) and inertia of other parts of the character.

 

When someone gets hit in the stomach with a fast moving cannon ball, they don't just fly backwards. The torso bends in response to the impact, dragging the other two halves of the body with it as it is pushed away.

 

When a heavy character walks vs. a skinny character. Lumbering vs. lithe and limber.

 

Bouncing balls give a VERY basic tool to understanding mass. I can animate two balls exactly the same size, texture and color. If I then change the way they are both animated, you SHOULD be able to tell which one is more flexible, light and bouncy vs. the one that is more stiff, heavy and resistant to bouncing.

 

I struggled with the "non-living" example I posted earlier for quite a while before I began to 'get it'.

 

It's very basic, simple and an excellent starting lesson for anyone who is beginning animation. You don't just put a beginning driver behind the wheel and hand them the keys. They need basic concepts and rules before ever starting that engine.

[NancyGormezano]

Your point is that explaining the deformation in terms of real-world mass muddies the concept because it's really about stylization. I disagree.

 

I think my way of explaining it terms of real world mass gives the student the best chance to take out of this exercise insight that can be carried forward to the more complicated cases of moving things around.

 

Most character animation doesn't involve things bouncing. It always involves moving masses around. I want the new animator to start thinking about those masses as soon as possible. Once you understand how they move you will know where to exaggerate that motion for stylized effect.

 

At AnimationMentor I saw a lot of fellow students NOT get body mechanics and I think it was because school was dealing in abstracted rules like "it looks better for objects to move in arcs" rather than real explanations like "objects move in arcs because the have mass and can't change direction instantly" or "a hand moves in arcs because it's swinging on the end of a bone and not moving by itself." (After two quarters the faculty knew they weren't getting body mechanics across, but they were people who intuitively understood motion and not quite able to explain how they understood it.)

 

I musta have missed the physics class when a bowling ball turned into a bean bag on it's way down to meet the floor due to gravity. If this helps you guys understand mechanics, forces, conservation of energy, mass, laws of motion, Potential, Kinetic Energy, vectors, etc, as well as body mechanics & physiology...then who am I to question reality?

[Rodney]

I musta have missed the physics class when a bowling ball turned into a bean bag on it's way down to meet the floor due to gravity. If this helps you guys understand mechanics, forces, conservation of energy, mass, laws of motion, Potential, Kinetic Energy, vectors, etc, as well as body mechanics & physiology...then who am I to question reality?

 

Nancy, you are the perfect person to question our reality.

 

 

Added: It's interesting that I just plunked down my $4.95 for KineticSoA's webinar on distortion and smear in animation. It's almost like the release was perfectly timed to go along with this discussion for me. Our perception of reality with regard to squash and stretch is a fascinating study. Mike does a good job with his take on it and , if you can see well enough through the bad images on the video playback, posts some nice examples from live action film as well as classic and modern day cartoonery. It's fascinating to see the distortion and blurs that appear in still frames of fast action where to a strictly logical approach to animation they should not be. Although there is a lot of room for interpretation, with regard to putting distortion into an animated scene it's clear that there is an entirely logical approach and practicality to adding squash and stretch into a scene.

[itsjustme]

I did overdo the squetch on that last ball. I was thinking sort of like a barely inflated exercise ball with a lot of "give" to it.

 

The problem in the frame that had the extreme difference in stretch/height was caused by my incorrect adjustment of the movement of the pivot (and not thinking about the top of the ball)...which makes the ball move higher when moved from the bottom of the ball to the center.

 

I'm hoping to get back to this sometime this weekend to make a corrected version.

 

 

----------------------

EDIT

----------------------

 

I just noticed the caption on Nancy's image...it was an exercise ball (in my mind barely inflated), not a bowling ball that is shown.

[MMZ_TimeLord]

itsjustme,

 

Not sure if it would help or hurt, but in the project I posted, there is a ball rig that really helped me. I made it, I admit, and it's probably limited. It did however let me rotate the ball while keeping the squetch direction constant. i.e. - the ball could be squetched a certain direction WHILE it was spinning.

 

Hope it comes in handy for someone.

[NancyGormezano]

I did overdo the squetch on that last ball. I was thinking sort of like a barely inflated exercise ball with a lot of "give" to it.

 

The problem in the frame that had the extreme difference in stretch/height was caused by my incorrect adjustment of the movement of the pivot (and not thinking about the top of the ball)...which makes the ball move higher when moved from the bottom of the ball to the center.

 

I just noticed the caption on Nancy's image...it was an exercise ball (in my mind barely inflated), not a bowling ball that is shown.

 

David: Please do not think I was critiquing your squetching at all! I like, prefer TOON, interpretive animation with squishy, squash. I like your animation in all it's imaginative glory.

 

But if one is going for realism, my point is that there is basically NO (to minimal) stretching at all on any ball, no matter what it is made of, as it nears the ground. And if it's barely inflated, it won't bounce much either. It will deform on the bottom when it hits the ground. I only posted that silly image with my overly squetched caption to serve as an illustration that stretch & squash especially of a ball is toon behavior and can not be accurately explained in terms of real world physics. But some people like to use imaginative analogies to explain the behavior. I could have said tennis ball/ping pong ball/ eye ball turns into bean bag as it gets closer to the ground.

 

Real world captures of bouncing balls are shown in this post. Theres no stretch and there's no noticeable squash (need to be closer up). Pity tho. Reality sucks.

[Rodney]

Richard Williams had an image like this in his book that is oft presented in animation lectures.

For a long time that picture was a fascination for me. (Results: We should take some care in interpreting what we are seeing)

 

We see an interesting thing in the images you've posted here and they are worth considering above and beyond the idea of squash and stretch for what we see and do not see:

 

 

While we cannot be sure what camera was used or what it's particular settings were (we can assume 24fps) we can note that there are many missing 'images' from this tennis ball's bouncing sequence. IMO when considering squash and stretch the most important images from this sequence are not there... they are missing; the contact frame and those frames immediately before it and proceeding it. Note that this represents perhaps the most exaggerated squash and stretch/blur/distortion we will ever see... the action so quick that the ball is completely removed/blurred from the scene! If played back in sequence our brains have no other choice but to fill in the missing details and we assume a contact was made. There had to be a contact even if the movement was so fast that we could not see it.

 

As you've stated concerning 2D, the squash and stretch is added mostly as a means to fill in the gaps of frames that cannot otherwise be seen. The blur of a fast moving object being one of the primary things that moves to fast for us to see.

 

Now, if you are meaning to say that tennis balls do not squash upon contact... (please forgive the terminology but) that is something of a stretch to me.

 

Although we may not see it, when a bowling ball makes contact with another object in a scene, something is going to give way (or break).

 

What is interesting about Richard Williams's book is that he spends considerable effort suggesting squash and stretch is something that isn't needed but then goes on to fill up the rest of his book with examples with squash and stretch in them. While I get his point and gladly receive it, his internal conflict (or perhaps it is more his fascination with the exploration of that theme?) makes him harder to read. It's almost like he's got a theory that he himself does not believe.

[NancyGormezano]

Now, if you are meaning to say that tennis balls do not squash upon contact... (please forgive the terminology but) that is something of a stretch to me.

 

read my post more carefully:

 

It will deform on the bottom when it hits the ground.

 

I have no idea what speed the camera(s) were taking images. It was not meant for use in any animation. The pictures came from reality based physics websites that had nothing to do with using it in an animation. They weren't concerned, nor had probably ever heard of stretch & squash. The images were used to demonstrate the path of a bouncing ball

 

Here's one of the websites where I got one image.

 

S&S was manufactured by a bunch of animators smoking dope.

[Rodney]

 

Edit: Drat embedding isn't working for me.

 

Here's a fun video sequence with a golf ball against steel.

Very unfortunately, the steel plate moves off screen so we don't get to see all of the #@$%! contact frames!

[itsjustme]

itsjustme,

 

Not sure if it would help or hurt, but in the project I posted, there is a ball rig that really helped me. I made it, I admit, and it's probably limited. It did however let me rotate the ball while keeping the squetch direction constant. i.e. - the ball could be squetched a certain direction WHILE it was spinning.

 

Hope it comes in handy for someone.

 

The rig I'm using has similar capabilities, but you may have something in there that would be better...I'll take a look at it. Thanks, Jody!

 

 

I did overdo the squetch on that last ball. I was thinking sort of like a barely inflated exercise ball with a lot of "give" to it.

 

The problem in the frame that had the extreme difference in stretch/height was caused by my incorrect adjustment of the movement of the pivot (and not thinking about the top of the ball)...which makes the ball move higher when moved from the bottom of the ball to the center.

 

I just noticed the caption on Nancy's image...it was an exercise ball (in my mind barely inflated), not a bowling ball that is shown.

 

David: Please do not think I was critiquing your squetching at all! I like, prefer TOON, interpretive animation with squishy, squash. I like your animation in all it's imaginative glory.

 

But if one is going for realism, my point is that there is basically NO (to minimal) stretching at all on any ball, no matter what it is made of, as it nears the ground. And if it's barely inflated, it won't bounce much either. It will deform on the bottom when it hits the ground. I only posted that silly image with my overly squetched caption to serve as an illustration that stretch & squash especially of a ball is toon behavior and can not be accurately explained in terms of real world physics. But some people like to use imaginative analogies to explain the behavior. I could have said tennis ball/ping pong ball/ eye ball turns into bean bag as it gets closer to the ground.

 

Real world captures of bouncing balls are shown in this post. Theres no stretch and there's no noticeable squash (need to be closer up). Pity tho. Reality sucks.

 

Sorry for the misunderstanding, Nancy...I was just making clear that the last animation was of something other than a bowling ball.

 

I know that squetch is an animation contrivance to give both a smear/blur frame (which would show up on film depending on the camera shutter speed) and a feeling of weight and does not accurately reflect reality. I'll have to go back and read some of the discussion that I missed.

[NancyGormezano]

 

Yes, that would be relevant if we were animating at what ? 1000 frames/sec? Oh yeah. Interesting video. I've seen it before.

 

here's my previous post in case you might want to read it, see where I got the images (yes I'm delusional)

 

I'm done with this conversation, it's becoming useless.

[Rodney]

I'm done with this conversation, it's becoming useless.

 

One of the many unfortunate side-effects of exploring the bouncing ball exercise is that you quickly find out that people don't want to explore the bouncing ball exercise.

[Rodney]

Here's something this latest discussion has got me thinking about.

While admittedly overly simplistic it represents an approach to the bouncing ball that I haven't previously considered.

It is therefore highly exploratory in nature (by me).

 

The idea is to consider each ball as a container and then make a determination of what the interior of that container consists of; a solid, a liquid or a gas. We may not actually know what may be underneath the outer shell but know that it can be presumed to favor one of the three. From there we can get a sense of how that object will react when it comes in contact with other objects in a scene.

 

Note that while the inner material is important we cannot discount the outer container. For instance, the exterior may be made of solid matter while the core is in a liquid state (think of an freshly laid egg here). A bubble may have a thin layer of liquid that only momentarily can contain air. Another object might have a soft exterior with relative inner solidity (think of the human body's outer tissue with its mostly hidden interior support framework).

 

This works well also if you consider balls of similar size, shape and exterior. A fully pumped up basketball can be expected to bounce higher than one partially drained. A flat basket ball can be expected to deform more easily. Fill that same basket ball with solid matter and it should squetch less.

 

Submitted for your consideration.

[NancyGormezano]

Submitted for your consideration.

 

Oh yes. I've reconsidered & decided: me being useless, ie. disruptive, off topic, can be fun.

 

Perhaps your chart needs to augmented? Instead of considering classifications of gas, liquid, solid, one needs to be considering specific gravity (a better measure of the relative densities of these balls, (eg ice solid is less dense than water liquid), as well as considering what material the ground is made of (specific gravity of feathers, sand? water? concrete, etc), and the surrounding atmosphere (eg, in a vacuum, under water?) as well as on which planet?

 

Probably ambient temperature, material temps are in there as well...

 

I'm pretty sure we don't need to consider IR signature, but never can tell with these things.

[Rodney]

Nancy,

If you are willing to expand on those subjects you know I'd be locked to the topic. (Check with David though... it's his topic)

 

Perhaps your chart needs to augmented?

 

Given it's current state, I'd say it is required to be augmented.

 

Instead of considering classifications of gas, liquid, solid, one needs to be considering specific gravity (a better measure of the relative densities of these balls, (eg ice solid is less dense than water liquid), as well as considering what material the ground is made of (specific gravity of feathers, sand? water? concrete, etc), and the surrounding atmosphere (eg, in a vacuum, under water?) as well as on which planet?

 

Would most definitely be of interest but perhaps to the newbie of a lesser priority?

 

Probably ambient temperature, material temps are in there as well...

 

Did you just say 'math'?

 

I'm pretty sure we don't need to consider IR signature, but never can tell with these things.

 

I was messing the other day with automated cool, warm and hot zones (via contraint to groups) so sure... why not!

And don't forget to throw in the albedos for good measure to make sure we capture the whole environment.

 

As always, I thank you for your considerate insights.

[NancyGormezano]

Nancy,

If you are willing to expand on those subjects you know I'd be locked to the topic. (Check with David though... it's his topic)

 

As always, I thank you for your considerate insights.

 

ugh...You took me seriously? Hopefully you're pulling my leg now.