Sunday, May 9, 2010

Akem Singh - ball and tail pt2 - Critique

Here's a follow up on a previous feedback post:

Hey Jean,

Thanks for the feedback but i'd like a little more information since its kind of hard to follow in texts form. Maybe little sketch's or doodles would make it easier to describe what your saying?

Yes please, tell me when I'm not clear enough. I need to make sure that the critique makes sense!

Anyway, I'll try and break down the bits that I didn't understand:

"when it lands on x10, the way the tail curves, you could have it a bit more to the left (so that the tip is not right above the ball), but then on x11 the tail wouldn't be up anymore, but pointing down"

Okay, this part I get! :)


"(imagine a whip action, or a flag waving and you're looking at it from the top)"

I'm having trouble relating this to a tail on a ball?

I meant this:
Imagine when you crack a whip on the ground, after you swing it up, when you throw it down, that downward motion, that last bit where the tip is point up and then goes down. Or take your hand, straighten the fingers like for a salute, but the ball of the hand on a table, not but the hand down flat on the table. But do this so that it's not laying flat at once, but you go from the ball to the tip of the fingers. One bit at a time rotates until the tip is down. Your ball tail move should be like this when it hits the ground. It should curl over the ball, it should curl down to the ground.

"As the ball goes up it would be more like x16 where it follows the path of action"

Could you describe a bit more about following the path of action?

Putting aside all the overlap and good stuff, if you have a tail attached to a ball, it's going to follow where the ball is going. The ball is pulling the tail. Wherever the ball is moving, that his path, or path of action: straight roll, bounce up and down, bounce left to right, etc. So the ball will follow that path and because it's a soft tail and bendy, it will follow the path pretty closely as opposed to being just still and hanging in there for the ride.

"and after the apex curl up, so not down like on x20. More up but not as much, maybe half of that"

Sorry but this has me totally confused! Hmm, to me it does look like its curling up already at x20. How can I make it go up
half as much without bringing it down?

This goes back to the tail following the path of the ball. So at the apex, meaning the highest point of the bounce, the tail will be curled less down, because it needs to stay on the path of the ball. See the red lines.

"as the ball falls the tail gets pulled again along the path of action and ends up like on x26"

Please explain this path of action thing as it relates to a ball & tail. Seems quite important in getting this to look right.

Check the pictures above for the path of action explanation. In your clip, on x26, the tail is straight and point up, right on the path of the ball falling down.

"one thing I wouldn't do is the direction change like you have it on x55, where it goes down and the up again. That looks more like a voluntary change vs. a straight physics overlapping action"

Don't understand this part. Where does it change direction at x55? Do you mean after x35 where it hits the wall and then spins in the other direction? Is this wrong?

Areas like x48 or x55 do look wrong if you are going for physics and even on a character level it looks odd. This being an overlap exercise the tail has to follow and be pulled by the ball, so if the tail is moving downwards, clockwise, it can only change the downward direction if the ball forces the tail to do so.

"I would keep it rolling beyond x70, it comes to a sudden stop there."

Would the ball "roll" at all since it has a tail attached to it? I couldn't quite picture what it would do but I guessed it would keep bouncing a little more and then come to a rest with the tail overlapping over the ball a little?

That's exactly it. The ball's bounces are getting smaller so the tail will have to come down and not stay up and at the end the ball can roll a bit and take the tail up, over the ball, like you said.

I think I made the mistake of not making the distinct between "a voluntary change vs. a straight physics overlapping action" and assumed there was suppose to be some action propelled by the ball having some character or motivation of its own?

Yes and no. It's true that in those bouncing ball assignments with tails the balls suddenly don't rotate anymore. :) It's just a ball taking the tail with it. But for the sake of the exercise, it's okay. The ball can have a motivation of its own, but the tail not, it's just straight physics at this point. Unless you go one step further and treat the ball and tail like a little animal and have the tail move on its as well. Like a little dog. When they move around you have straight physical tail overlapping action mixed with voluntary swinging and stiffening and all kinds of other movements.

"Or just cut the shot there."

I'd like to keep animating it coming to an end properly than cheating myself out of learning how to do that :) I think it might help me learn how to make an object looks like its coming to rest properly with overlapping action.

Also can you give me some more advice about what the tail should be doing between frames 44 and say 70 to make it look better? I'm guessing there should be a lot less movement but still overlapping some how?

Yes again, you're right. The overlap will get smaller and smaller but the ball will still pull it along. You can always take a piece of string and attach it to a ball and then film that for reference.

Sorry if it seems like i'm asking for too much detail but i'd really like to understand and get the hang of overlapping action as I think it will improve my animations alot!


No need to apologize!! That's exactly what I'm here for and what this workshop is all about. You can never ask for too much detail!

I hope this is clearer!


No comments:

Post a Comment