Controll a radial wipe with a Null

[THExDUKE]

Hello Community,

currently I am working on an automotive project and I was trying to find a way to control the blend effect "radial wipe" by the position of a Null object.

The scene contains a Null, which runs along a curve. Beside some other things the Null does, I would want to create a radial mask, just like the radial wipe. But my poor expression skills haven´t brought me further.

The path is not a full circle (277.1°) ! - I attached a solid to the null to kind of aim-constrain to it. That showed me some rotation values starting from -67.5° going to 0 to 180 and then to -150.4° (in sum 277.1°). My idea was somehow to tell the radial wipe that it´s 0% start is my -67.5° and the 100% is the -150.4° but I fail in doing so.
Then I thought that maybe instead of the negative values I should take the 277.1° but that didn´t work as well.

Then I tried to construct a radial wipe like mask out of solids, but it doesn´t work. I could hook those nicely to the null but the overlaping pieces just can´t be masked. Im out of ideas and asking for help and maybe someone is here to give some hints.

Thanks for reading and the help in advance.

Cheers

[Dan Ebberts]

That's an interesting conversion. Maybe something like this:

r = (your rotation value);

if (r > -90)

  linear(r,-67.5,180,0,100*247.5/277.1)

else

  linear(r,-180,-150.4,100*247.5/277.1,100);

Dan

[THExDUKE]

Hey Dan,

thanks for the help.

What I have right now is this:

I created a solid with the radial wipe effect. This will be my synced mask, which I want to use for each other element to be masked. I called it RotationMask

In the effect settings of the radial wipe, I hooked your expression into the "Completion" attribute (don´t know the english term for it, cause I have it in german here )

This is the code there so far:

a=thisComp.layer("RotationMask_VALUE").transform.rotation;

r = (a);

if (r > -90)

  linear(r,-67.5,180,0,100*247.5/277.1)

else

  linear(r,-180,-150.4,100*247.5/277.1,100);

The layer RotationMask_VALUE is a second solid, which is aiming to the Null - thats where I get the angle information and will be invisible because it only gets the values. The expression there is this:

target = thisComp.layer("mouse layer"); // set this to the layer to aim at

p = position - target.position;

radiansToDegrees( Math.atan2( p[1], p[0] ) )

(Mouse Layer is the Null, btw)

Right now, the setup doesn´t work as wanted. Plus I have the problem of alligning the start point of the radial wipe to the Null´s initial position.

Any ideas?

Thank you very much

Erce

[Mylenium]

I don't follow. You can't calculate a surface normal or tangent just from a single origin point. Hence it would be impossible to have the radial wipe respond in the manner you seem to desire. You have to include a look-ahead or look-behind point sampled from the path to have a stable reference for the "edge" of the wipe. Conversely, that would be required to determine from which sector the angular functions actually return their values. You know, a full circle is 2 x Pi, not just Pi and different angles produce the same values. Once you clean up that mess, everything else will fall into place - your crooked angle values could likely easily be normalized to sensible ranges because once you actually calculate the inclination, it can simply be added/ subtracted.

Mylenium

[THExDUKE]

Hey Mylenium,

So you say that I have to get rid of my "value" solid and instead of that I would want to put a "look ahead" (whatever and however...) to the Null to get the proper angle. But it would also then require a point FROM where (center of rotation) it does measure, right? Also you talk about a full circle hopefully knowing, that my "circle" isn´t full? Im not sure, like I said...I have poor skills in scripting and this is a tough one for me including not even knowing all the possible tricks. So "mea culpa"

Thanks

Erce