Bouncing Ball critique

[Pesto]

Sorry I know it is boring but I wanted a critique before I move on to giving the ball some personality. Any and all opinions welcome. Also, I am using 3% motion blur. What are people's thoughts on MB? When I didn't use any the animation looked kind of "jumpy". When I used more (>5%) parts of the ball looked fragemented. Is it a question of finding just the right %?

 

http://www.youtube.com/user/Pesto642

[thefreshestever]

nice... the motion blur looks fine... i think you should squash the ball a little less, looks too much for me.

[Pesto]

Thanks for the reply. I agree with the squashing...I was trying to exaggerate it alittle more than normal. I agree a ball that bounces like that would not squash nearly as much. Thanks for the input.

[robcat2075]

I can tell you that the squash is outta whack and there's something going wrong around the apex, but for crits you should post a quicktime .mov with frameburn so it can be examined more closely. There's no way to single frame through a youtube movie unfortunately.

[nixie]

I would also add an extra bounce or two at the end, it rolls to early.

[Pesto]

Robcat - Good idea....but how do I show frame burn in AM? When you say squash is out of whack I assume you mean becasue I exaggerated it. As for the apex...the height of the bounce...having something wrong...can u elaborate? Not sure what u mean.

[Pesto]

Here is a new version...less squash and stretch, two more smaller bounces.... It's in QT so you can scrub thru it.

 

Thanks

Ball_bounce_BRAND_NEW_FINAL_MB3__new_bounces_smaller.mov

[CreativeAustinYankee]

Not bad, but you still need a few more bounces at the end before it rolls away. The height of the bounce is directly related to the "springiness" of the ball material.

 

For most of this, the ball looks pretty springy. My suggestion is to work out a simple formula. If the ball bounces this high at Point A, the percentage of that needs to be ? at Point B, and ? at Point C and so on. Do this with distance (movement across the screen) as well.

 

Nice start,

 

Steve P.

[jzawacki]

It could be my eyes, but the second bounce looks like the ball is going trough the ground. Maybe only because of the strait line and not enough bulge on the ball to show the amount of displaced rubber? Dunno.. ignore me..

[thefreshestever]

hmm... i stepped through it frame by frame and my question would be why does the ball stretch before it hit´s the ground? when it hits the ground it squashes, then stretches when it leaves the ground and then goes back to normal position, that´s right so far, but before? maybe i´m wrong and this is a common technique i don´t know about, but it seems wrong for me...

i agree with the others that it should bounce one or two more times at the end....

[ypoissant]

I see several issues. First, the squash and sretch is way too exagerated. The stretch before it hits the ground should not be there especially if you are using motion blur.

 

The motion is not right. The first bounce is 1/2 the height from where the ball first falls. So, the second bounce should also be 1/2 the height of the first bounce height and the third bounce should be 1/2 the height of the second bounce, etc. The acceleration when the ball falls and the deceleration when the ball lifts is almost linear. There is very little sense of acceleration or deceleration. The distance the ball travels when it falls is proportionately equal to the square of the number of frames. So the distance the ball trvels down between frame 2 and 0 is 4 times the distance traveled between frame 0 and 1. And the distance the ball travels down between frame 3 and 0 is 9 times the distance traveled between frame 0 and 1, etc.

 

The timing is not right. Assuming that you keep a 1/2 height bounce height, the timing between each floor hit should be 70% the time of the previous floor hit. I counted 10 frame from the fall to the first hit and this is equivalent to one half bounce so if it was a full bounce it would have taken 20 frames. That means that the second floor hit should be 14 frames away. It is currently 20 frames away. Then the third floor hit should be 10 frames away, then 7 frames away, then 5 frames away, then it gets complicated because we start getting half frames so the next one would be 3 frames away and then 2 frames away and then less than 2 frames can't be perceived except for some blurring. BTW, this 70% is not just arbitrary. It follows the law of gravity. The length proportion of the bounce is equal to the square root of the height proportion and square root of 50% is 70%.

 

The deceleration at the end is too short. The ball should continue to roll and decelerate much more slowly.

 

Once you have those issues straightened, then you can start playing with the timing and the motion to convey a different feeling but IMO, you need to get those straight first.

[Rodney]

why does the ball stretch before it hit´s the ground? when it hits the ground it squashes, then stretches when it leaves the ground and then goes back to normal position, that´s right so far, but before? maybe i´m wrong and this is a common technique i don´t know about, but it seems wrong for me...

 

This 'stretching before the contact' is a pretty common technique used in animation and is championed (at times and certainly not universally) by such animators as Richard Williams. Williams seems to have a special interest in it for character animation but also cautions against the use of it indescriminately. Where Williams suggest it not be used is where form or function suggests rigidity that would not allow for squash and stretch. The effect must work within the confines of the animated world the objects move through. Consider a head for instance. The jaw might squash and stretch considerably whereas the cranium would squash very little if any at all. I'm a fan of the 'stretch to contact' technique.

 

On of the reasons to use stretching in the first place is to allow the eye to register the movement of an object about to change direction. Through stretching the eye is allowed to see more of the object before and after the contact and register the change.

 

My animation books have not yet arrived so I won't be able to point to specific references but there are sections devoted to the use of stretching prior to contact in several places in Williams's book 'The Animator's Survival Kit'. Again, the warnings are there too that caution against creating 'floopy' animation (if I recall the word Williams uses for it correctly).

 

Yves points out the combined use of motion blur with stretching in this animation. Both are mechanisms designed to achieve the same thing; illusion of rapid movement. Both motion blur and stretching are shortcuts that can help achieve the illusion of speed. As Yves suggests if you have one you probably don't need the other. Where timing and spacing is used effectively or an object needs to be seen as solid you may not need to use squash or stretch at all.

 

It can be noted in this animation that the ball stretches for no apparent reason and there is no 'stretch to contact'.

With 'stretch to contact' the idea is to have the object stretch and make contact with the ground 1 frame before contact/squash is generated. In this animation the stretch would have to be considerable to make this happen.

[Pesto]

Thanks everyone for their input. I will redo the animation taking all these things into consideration. Thanks for the help!