[Dan Ebberts]

It is possible, but it’s probably the most complicated kind of triggering you can do with expressions. Because expressions have no memory, your expression would need to look at every previous frame up to the current time and calculate how many times the triggering condition (mask expansion going from above -231 to below -231, for example) has occurred. Then use that number (or some multiple of it) as the seed for seedRandom() to generate a new random position.

[Dan Ebberts]

What time expression are you referring to?

[Dan Ebberts]

You could set up an expression for your mask path, where the height is based on the slider value. This example is more complicated than it needs to be, because I wasn’t sure whether your path goes clockwise or counter-clockwise so the expression allows for either:

s = effect("Slider Control")("Slider");
p = thisProperty.points();
if (p[1][0] < p[3][0]){
  p0 = p[3] - [0,s];
  p1 = p[2] - [0,s];
  newP = [p0,p1,p[2],p[3]];
}else{
  p0 = p[1] - [0,s];
  p3 = p[2] - [0,s];
  newP = [p0,p[1],p[2],p3];
}
createPath(newP,[],[],true)

[Dan Ebberts]

Expressions don’t have memory, so you’d probably need to have separate master controls for each precomp. It sounds like a job better suited to a small script though.

[Dan Ebberts]

I think this is what the After would look like:

a=text.sourceText.value;
if (textIndex > 1 &&  a[textIndex-2] == "a") {
  b=a[textIndex-1]; //check letter after the A
  (b=="a" || b=="e" || b=="k" || b=="l" || b=="q" || b=="x" || b=="z") ? 100 : 0;
}
else 0

[Dan Ebberts]

This is an interesting (and tricky) one. What this should give you is that variable “t” will represent how long it’s been since the active layer has changed. You would then use this value to drive your animation. I’m not sure that gives you exactly what you need, but it might give you an idea of how to get there.

myIdx = 0;
t = 0;
for (i = 1; i <= thisComp.numLayers; i++){
  if (i == index) continue; // skip this layer
  L = thisComp.layer(i);
  if (L.enabled && time >= L.inPoint && time < L.outPoint){
    myIdx = L.index;
    break;
  }
}
if (myIdx > 0){
  myActive = thisComp.layer(myIdx);
  myTrigger = Math.max(myActive.inPoint,0);
  for (i = 1; i < myIdx; i++){
    if (thisComp.layer(i).enabled && thisComp.layer(i).outPoint <= time){
      myTrigger = Math.max(myTrigger,thisComp.layer(i).outPoint);
    }
  }
  t = time - myTrigger;
}
t

[Dan Ebberts]

You were pretty close. I think this will work (as long as your text is only one line):

a=text.sourceText.value;
if (textIndex > 1 && a[textIndex-1] == "a") {
  b=a[textIndex-2]; //check letter before the A
  (b=="a" || b=="e" || b=="k" || b=="l" || b=="q" || b=="x" || b=="z") ? 100 : 0;
}
else 0

[Dan Ebberts]

It’s hard to tell exactly what you’re after here, but it can get pretty complicated if you need the wiggle to start where the last rotation ended and vice versa. Play with this and see what you think:

bamm= thisComp.layer("Audio Amplitude").effect("Both Channels")("Slider");
rotateRate = 150;
threshold = 60;
f = timeToFrames(time);
rotateFrames = 0;
wiggleFrames = 0;
for (i =  0; i <=f; i++){
  bamm.valueAtTime(framesToTime(i)) > threshold ? rotateFrames++ : wiggleFrames++;
}
tW = framesToTime(wiggleFrames);
tR = framesToTime(rotateFrames);
w = wiggle(7,50,1,.5,tW) - valueAtTime(tW);
tR*rotateRate + w

[Dan Ebberts]

Actually, if you’re applying it to a point control, you’d want to change the last line to this:

fromComp(L.toComp(p.pointOnPath(offsetPct,time)))

[Dan Ebberts]

This should work:

sc = thisComp.layer("SCALA_SECONDA_PARTE").transform.scale;
sc[0] > 100 ? sc - [15,15] : sc