Trying to Create a Wiggle Effect that STAYS Where It Moves To (Instead of Always Bouncing Back)

Question

For making stop-motion animation, I'm trying to alter the AE Wiggle expression (or essentially create a Wiggle knockoff) that jumps to a slightly altered position, but then stays at that position until a new randomized Wiggle coordinate comes in. So far, it isn't quite working and would love some input if anyone else out there can figure out how this would work better!

Why I'm Doing This: The idea here is that realistically when shooting stop-motion with an actual camera, if a paper cutout accidentally gets bumped into a slightly new position, it will probably stay there until bumped again, X number of frames later. (It won't, in other words, jump to a new spot then immediately bounce back, as the Wiggle expression has it do by default.) So I'm trying to make a more realistic-looking stop motion effect.

The Current Setup with Expressions: My approach has been to use valueAtTime to have After Effects read previous position values. If a new wiggle coordinate is given, the wiggling layer's position accepts that value and moves there. But if no new coordinate is given (and the randomizing function spits out a zero instead), the position is told to stay put (by accessing the previous position coordinates with valueAtTime(time-thisComp.frameDuration) and taking in those previous coordinates as the new coordinates).

I can't seem to get this to work more directly with Wiggle, so I've set up a Null with multiple sliders to mimic (but alter) what the Wiggle expression does.

Here are the first two sliders in the Null that I've called "Wiggle Sliders":

Slider Control > Slider (In "Wiggle Sliders" Null):

x = Math.round(random(Math.round(random(Math.round(random(Math.round(random(Math.round(random())))))))))

wa = 40*random(x)

neg = Math.round(random())

if(time==0 || wa!=0)
{
if (neg==0)
{
val=wa
}
else
val=-1*wa
}

else
{
val =thisComp.layer(index).effect("Slider Control 2")("Slider").valueAtTime(time-thisComp.frameDuration)
}

Slider Control 2 > Slider (In "Wiggle Sliders" Null):

val = thisComp.layer(index).effect("Slider Control")("Slider").valueAtTime(time)

Slider Control 2 just takes in what Slider Control currently has, so that Slider Control can then access Slider Control 2's previous value (because I can never get it to work when just one Slider is trying to access its own previous value).

The first Slider Control value is then accessed by the x-position of the layer I want to wiggle, adding or subtracting this amount from its original position value (and then uses two more sliders, 3 and 4, in the Null that do the same thing for the y-position):

Transform > Position (In "Wiggle Sliders" Null):

val = [transform.position[0] + thisComp.layer("Wiggle Sliders").effect("Slider Control")("Slider"),
transform.position[1] + thisComp.layer("Wiggle Sliders").effect("Slider Control 3")("Slider")]

Problems with Execution: Generally this seems close to working...but then the sliders glitch a lot. The slider values at first successfully hold on a random value that's been spat out (ex: 0.3 or 5.7 or -9.4), but then often a few frames later they revert back to zero (which they are not supposed to). Sometimes they revert back to zero after ten frames, sometimes after two frames, sometimes right away.

Also, even when the slider values hold as they should, the position value of the wiggling layer does not. So far, it goes to the new position, then one frame later always jumps back to the original position (even if the inputted slider values are telling it to stay put and not go back to the original position for, say, another twenty-five frames).

I'm certainly no expression expert, but this setup seems like it should work to me, so I'm awfully frustrated that it isn't working so far. Thanks in advance for any advice!

Dan Ebberts · Accepted Answer

How about something more like this (this one would be for position):

minOffset = [-10,-10];
maxOffset = [10,10];
minHold = .5;
maxHold = 3;
seedRandom(index,true);
t = inPoint;
idx = 0;
while (t <= time){
  idx++;
  t += random(minHold,maxHold);
}
seedRandom(idx,true);
offset = random(minOffset,maxOffset)
value + offset

If you need random jumps in position and rotation to be synchronized, you could add a slider with an expression like this:

minHold = .5;
maxHold = 3;
seedRandom(index,true);
t = inPoint;
idx = 0;
while (t <= time){
  idx++;
  t += random(minHold,maxHold);
}
idx

and then this for position:

minOffset = [-10,-10];
maxOffset = [10,10];
idx = effect("Slider Control")("Slider");
seedRandom(idx,true);
offset = random(minOffset,maxOffset);
value + offset

and this for rotation:

minOffset = -5;
maxOffset = 5;
idx = effect("Slider Control")("Slider");
seedRandom(idx,true);
offset = random(minOffset,maxOffset);
value + offset

Mylenium · Answer

AE has no way of remembering values and that's what it basicalyl boils down to. If a "dumb" change with posterizeTime() and such doesn't work, you can't possibly avoid delving into all this integration stuff to figure out what the last value/ position actually was as explained in Dan's own classic article:

http://www.motionscript.com/articles/speed-control.html

Mylenium

Sign up

To post, reply, or follow discussions, please sign in with your Adobe ID.

Sign in to Adobe Community

To post, reply, or follow discussions, please sign in with your Adobe ID.

Scanning file for viruses.

This file cannot be downloaded