Pose 2 Pose with ONE SLIDER controlling a Layer with Multiple Sliders (And therefore: keyframes)

[Wouter Dijkstra]

The way I did it now (if, else), there is no animation between the keyframes…. And I want to have a morph between the shapes. Not just “different images”.

[Filip Vandueren]

Can you make a joystick and slider to control your different joystick and sliders ?

[Wouter Dijkstra]

As it seems to me Joysticks and Sliders always gives for each ‘frame’ one slider. So it will give me 6 sliders again.

[Walter Soyka]

A lot of Ae rigs like this use time-remapping to let you choose a particular frame from a comp. When you animate a controlling slider from the pose in frame 1 to the pose in frame 3, it passes by the pose in frame 2.

To have the mouth shape interpolate between two arbitrary mouth shapes, you have two options that I see:

1) Build your mouth shape such that every pose of the mouth can be created by adjusting the same set of properties. Then we can use valueAtTime to get the value of each property at the outgoing and incoming pose, and have Ae interpolate between them. You’ll have to use the new Master Properties features and some expressions to “push” the pose value into the rigged comp.

2) Animate all 30 transitions by hand (1-2, 1-3, 1-4, 1-5, 1-6, 2-1, 2-3, etc.) on the same timeline and use time-remapping expressions to “pull” the correct sequence from the rigged comp.

Can you share the rigged mouth shape comp you’ve built?

Walter Soyka
Designer & Mad Scientist at Keen Live [link]
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
@keenlive   |   RenderBreak [blog]   |   Profile [LinkedIn]

[Wouter Dijkstra]

Hi Walter,

I was as a matter of fact experimenting with ValueAtTime, but still it only gives me the ‘exact’ value (1 frame) and I am not able to make an animation between the values controlled by one Slider. This specific Slider I have called “control”.

I pasted the following code in each of the 6 sliders:

x = thisComp.layer(“control”).effect(“Slider Control”)(“Slider”)

if ((x>=1) && (x<2)) {[valueAtTime(framesToTime(0))];} else
if ((x>=2) && (x<3)) {[valueAtTime(framesToTime(3))];} else
if ((x>=3) && (x<4)) {[valueAtTime(framesToTime(6))];} else
if ((x>=4) && (x<5)) {[valueAtTime(framesToTime(9))];} else
if ((x>=5) && (x<6)) {[valueAtTime(framesToTime(12))];} else
if ((x>=6) && (x<7)) {[valueAtTime(framesToTime(15))];} else 0

I attached the RIG in this post.

13327_mouthrig.zip

[Filip Vandueren]

Then what happens if you animate your slider from 5 to 1: do you want it to pass through setting s 5,4,3,2,1 (similar to the behaviour that Walter described for Time-remapping) or do you want it to tween between the settings of 5 and 1.
The latter can not be achieved by just 1 slider (unless you use hold-keyframes,n which defies the purpose ???? )

you would need 3 sliders:
fromPose
toPose
amount.
and animating these 3 needs its own housekeeping, because fromPose and toPose should use hold keyframes and not change in the middle of when you’re animating slider 3.

Then ach of your original 6 sliders would need something like

[this is pseudo-code, won’t work yet)
a=valueAtTime(fromPose);
b=valueAtTime(toPose);
linear(amountSlider,0,100,a,b);

[Wouter Dijkstra]

Thanks Filip. So I would need 3 sliders. It makes sense what you are telling me! Thanks for this. But I suppose 3 sliders doesn’t give me de ‘fewer housekeeping’ that I was looking for. If this is the only solution, I suppose I have to stick to keyframing 6 sliders.

I was hoping for a very nice and fancy way using only one slider. Would it according to you be possible like this –>

Slider value goes from 1-10

1 till 2 is the same as 100, 0, 0, 0, 0, 0 to 0, 100, 0, 0, 0, 0, 0

It seems like it should to me. On the other hand all my tries gave me only absolute frames and no in betweens. Shouldn’t it be possible to create something like this with the linear expression that you are using as well?

[Wouter Dijkstra]

Walter:

1) Build your mouth shape such that every pose of the mouth can be created by adjusting the same set of properties. Then we can use valueAtTime to get the value of each property at the outgoing and incoming pose, and have Ae interpolate between them. You’ll have to use the new Master Properties features and some expressions to “push” the pose value into the rigged comp.

This solution seems very close as well, but I’m not sure what the New Master Properties give me for options to make this. I would be grateful if someone could maybe elaborate this solution more. It seems to be the thing I almost had, except this will probably give the in-betweens as well.

[Wouter Dijkstra]

just to add an alternative that would also make me happy:

The control slider (just one) should function more or less like a button. Pasting 6 keyframes every time. If it would work like this there would be in-betweens as well. But still the question remains… how can I do it.

[Filip Vandueren]

This works:

this is the expression that each of your six sliders should get, change the first line accordingly.
I have used a way of interpolating that will take into account the easing on your control-slider, so if you use eases or even hold-keyframes, that will all be reflect in the timing of the 6 controlled sliders.

myIndex = 1; // number of the slider.

sl=thisComp.layer("Null 1").effect("Slider Control")("Slider");

if (sl.numKeys>1 && time>sl.key(1).time && time<sl.key(sl.numKeys).time) {

nk = sl.nearestKey(time);
prevK = nk.time<time ? nk : sl.key(nk.index-1);
nextK = sl.key(prevK.index+1);

fromValue= (prevK.value == myIndex) ? 100 : 0;
toValue= (nextK.value == myIndex) ? 100 : 0;
if (prevK.value<nextK.value) {
linear(sl.value, prevK.value, nextK.value, fromValue, toValue);
} else {
linear(sl.value, nextK.value, prevK.value, toValue, fromValue);
}
} else {
sl.value == myIndex ? 100 : 0;
}