Dynamic Parenting?

Hi,

I'm wondering is there any kind of Dynamic Parenting available inside Motion?
Track Point behavior is almost, but it's not based on current time and current place of source so target shape will jump to that place depending how much the source have been moved before behavior starts to affect. Not ideal!

Thanks!

Try a Link parameter behavior for Position parameters, or clone layers. Not really clear what exactly your goal is.

Hi,

I made quick example to clear up what's I'm looking after. Hope this will help

Video Link
Motion File Link

I'm not entirely sure I've understood the effect you are trying to achieve but I would have thought you could do it if you made a group of the two text layers and applied the link behavior to the group.

Does that answer it?

Here is key framed version what I'm looking after. Please check this!

What I'm looking after is simply current time and current position based linking.
When Link behavior starts, linked layer will follow where ever source goes, but when Link behavior ends, it should stay on that current place on that current time without jumping around any kind of offset. Just like you travel with bus

EDIT:
Sorry about blinking MOV. Motion seems to hug so much memory or does use GPU so Quicktime will drop some frames...

Ah, I may know what's happening. Make sure you have NO keyframes that run ata the same time as the Link behavior. Go to the first frame of the Link parameter behavior, then set the Offset to adjust properly.

I've not seen the Link behavior do that all on its own before.

Adjusting offset will correct starting point, but when Link behavior ends, it will pop again. It's like you getting from London to New York and as soon as you get out of airplane, you just get transported back to London

Video

Keyframe the offset.

Hi,

It won't work because when behavior ends, it won't affect any more and also if I have to keyframe to match position, there is no point to add link behavior to it and I could just key frame position directly. I like idea behind behaviours. It's key frame free experience with easy to experiment different timings and such. Too bad this time it's not possible to accomplish what I'm looking after, but I sent feedback for Motion team so maybe some day we will see easy current time and position based linking.

Thanks for all!

Personally if that was ALL you were doing, what is in the demo, I'd keyframe.

Step 1 - Three keyframes.
Step 2 - Done.

Cause the example is way, way to simple for any sort of behavior or tracking or anything. 3 keyframes and you're done.

Or you're going for something way more elaborate?

Yeah, sure that demo was only to illustrate what I'm looking after very simple way.

Here is another example (still not too complex, but I hope you'll see my point)
Video

Here I want red ball to always move with dark "magnetic bar" and this all is made by key framing. All bars and red ball it self.

If I want to change how all moves, I have to animate magnetic bar and red ball, when if there would be "dynamic parenting behaviour" all I need to do is animate only magnetic bar and red ball will just follow without any animation fix. With this example I used linear interpolation for all key frames to keep all simple, but If there would be custom ease in/out with bezier, that will cause some "slipping" between magnetic bar and red ball.

Sure it's doable using key frames, but at least for me there is good place to have that kind of behaviour.

Image

OK, now I see what you are trying to do!

It's pretty simple to achieve the linking and unlinking effect you are talking about if you use the Scale parameter in the Link behavior.

When the Scale is at the default of 1 it will have full effect, but if you set it to zero then it will have no effect. Therefore all you need to do is keyframe the Scale value at the points you wish.

if you didn't want to keyframe at all you could use a Ramp behavior on the Scale parameter and control the dynamic linking that way.

BTW in your example you will probably also be wanting to use the Offset value as well, as Ben pointed out, in order to correct for the position at the zero Scale state.

To test this out, I set a zero Scale keyframe at frame 77 and another with a value of 1 at frame 78. I then set an X Offset of -278 to match the position. I'm sure there's a neater way of handling the offset but you will be able to find that if you look at how you are setting your other values.

Here's the revised motion file:



The interesting thing about the Scale parameter is that you can also use negative values which opens up some really useful possibilities. Alternatively use values greater than 1 to get a larger effect to the behavior. This is the closest Motion gets to some form of expression type behavior and with some ingenuity you can get a lot of value out of it.

Thanks for reply.

I tested that, but there was weird jump when I changed scale value... Anyway, even it would be possible it's seems to be way too complex and not too handy as I expected it to be. If I have to add custom behaviors and animate offsets every time it has new target to follow, it's like animating with key frames and pointless to use behavior at all.

Like I said, I did sent feature request about it so I hope we will see easy to use current time and position based parenting behavior some day soon... (Just pick source parameter to follow and follow that as long as behavior range last and stay there when it ends).


Thanks for all!

I really don't see your difficulty. I sent you the revised project, you should be able to see how easy this is to achieve. What's not clear to you?

Thanks Simon to taking your time to help me understand how Motion works.

Yes it sure works, maybe I was misunderstood how behaviors should behave and wanted it to be more simple and straight forward than they are. I was thinking all behaviors are like Move and Fade in/ out and easy to move around time and scalable.


Thanks again!