[Dan Ebberts]

A scale expression like this might work:

pct = effect("Slider Control")("Slider")/100;
val = value;
if (numKeys > 1 && time > key(1).time){
	d = (time - key(1).time)*pct;
	val = valueAtTime(key(1).time + d);
}
val

[Dan Ebberts]

Hard to be precise without seeing any of the details, but I’d create a position (not anchor point) expression for the text pre-comp where you calculate the current y position of the center of the text, subtract that from the y position of the center of the circle and add the result to the current value, roughly like this:

yT = textCenter[0];
yC = circleCenter[1];
yAdjust = yC - yT;
value + [0,yAdjust]

[Dan Ebberts]

An expression can’t do anything other than alter the value of the property hosting the expression.

[Dan Ebberts]

Here’s a simple example that will need more complexity if there are multiple lines, or if the number of characters changes:

curChar = text.sourceText.value[textIndex-1];
val = 100;
t = time - thisComp.frameDuration;
while (val > 0){
  prevChar = text.sourceText.valueAtTime(t)[textIndex-1];
  if (curChar != prevChar) break;
  t -= thisComp.frameDuration;
  val -= 10;
}
val

[Dan Ebberts]

v = toComp(anchorPoint)[0]; 

should give you the x position of the layer itself.

[Dan Ebberts]

I think the first test is messing things up. Try it this way:

if (MENU  == 1 || MENU  == 2 || MENU  == 3 || MENU  == 4){

[Dan Ebberts]

I think something like this might get you started. Set nLayers to the number of text layers participating (they need to be contiguous) and set firstLayer to the index of the first participating text layer.

gap = thisComp.layer("controls").effect("Gap")("Slider");
nLayers = 4;
firstLayer = 1;
curEdge = 0;
wTotal = gap*(nLayers-1);
for(i = firstLayer; i < firstLayer+nLayers; i++){
  L = thisComp.layer(i);
  r = L.sourceRectAtTime(time,false);
  w = r.width;
  wTotal += w;
  if (i == index) myOffset = r.left;
  if (i < index) curEdge += w + gap;
}
x = thisComp.width/2 - wTotal/2 + curEdge - myOffset;
[x,value[1]]

[Dan Ebberts]

Well, you can’t have a 3D path, so the best you could do is to break it into a bunch of segments, with one path layer required for each connection between 3D layers. Then you could make each path layer 3D (but the path itself would still be 2D), but then you have to worry about positioning it in world 3D space and orienting is so that it connects between the 3D layers and also presents itself to the camera in a way that maximizes its visibility (so that it’s not edge-on to the camera). Lots of work…

[Dan Ebberts]

It should work fine with the smet layers as 3D, but not the line layer.

[Dan Ebberts]

I think you’ll get that message if you’ve made the line layer 3D.