[Lucia Bastenhof]

I’ll try some tweakings tomorrow. But what’s strange is that the rest of the expression works perfectly when applied with the first linear…

[Lucia Bastenhof]

Hi,

I made a little rendering so that you can have a glance at it.
Here it is : https://fr.youtube.com/watch?v=IJWqxMeX5ls

I followed a tutorial by Maltaannon, to make a radar, and an object that becomes visible only when encountering the radar.
I then tried to make my own one, and to make it really like I wanted, I needed to make the radar line by a different way, in a precomp.

The line A I talked about is the tutorial one, and it works perfectly of course, that is to say the object appears when asked.
The line B is my own radar line, and I try to adapt the expression.

It seems to be more difficult than I thought, but if I learn something, that’s OK. That’s why I’m still trying to solve it.

As you can see in the video, the object remains visible quite a long time, and it reacts like with a wiggle on the opacity. But what’s good now (with the first changes you suggested) is that it appears at the right moments, when encountering the radar.

With the values you suggested in your last message, I made some attempts, changing :
– the rotation values (at first,0 -> -360) to 540 -> 180, and to 360 -> 0 ;
– the values of the linear in the expression following the new ones, trying 90,-270 and -270, 90.
But the results are worse : the same wiggling reaction, and it appears only one time instead of 3 (the 2nd time only).

Do you understand what’s wrong with all that ?

Thank you.

[Lucia Bastenhof]

I don’t understand quite well. You write : “The second parameter to linear() needs to be less than the third parameter. “. But 0 is more than -270 ?
Did you want to say : “The first parameter to linear() needs to be less than the second parameter.” ?

Anyway, I tried it, but the returned values remain different :
– time1 : A=-30 , B=-150
– time2 : A=-270 , B=90
– time3 : A=-193,2 , B=13,2
I learn the order is important, time2 is different, but there remain 2 deep questions :
1/ Why are the returned values different ?
2/ The result is half good, half wrong. Basically, the expression is written on the opacity parameter of a layer-object, and this object appears now, indeed, when encountering the scanner-line.
But its behavior remains strange : its opacity is on and off like with a wiggle. ?

For full information, here’s the entire expression :

===========================
sldval = degreesToRadians(180-linear(comp(“Aiguille PC”).layer(“Aiguille”).transform.rotation,-360, 0,-270, 90));
object = transform.position;

x0 = thisComp.layer(“grid”).transform.anchorPoint[0];
y0 = thisComp.layer(“grid”).transform.anchorPoint[1];

rayon = length(thisComp.layer(“grid”).transform.anchorPoint,object);

xA = x0 + rayon * Math.cos(-sldval );
yA = y0 + rayon * Math.sin(-sldval );

aiguille = [xA,yA];

dist = length(aiguille,object);

if((dist > 0) && (dist < 10)){ 100; }else{ 0; } ================================== The source parameter is coming from another comp ("linear(comp(“Aiguille PC”).layer…..) but I don’t think that it’s the cause of this behavior.

Do you have an answer to one of those 2 questions ?

THANKS !! 🙂

[Lucia Bastenhof]

I had written Comp instead of comp… :-/

Thank you Dan.

[Lucia Bastenhof]

(it tells me that “undefined is not an object”… ?)

[Lucia Bastenhof]

Thanks for answering, Dave.

[Lucia Bastenhof]

@ Darby : Thanks for the tip ! 😉

@ Frank : I needed 2 masks because there is an animation of opacity. But the anti-aliasing leaves a line between 2 masks just touching each other.
But finally, I don’t need any mask at all ! 🙂

[Lucia Bastenhof]

Dave, I first tried your method (a white solid with the edge coming just in the middle of the comp – Track matte) but it appears that the same problem as with the masks occurs !!
Looking at the white solid in solo mode at a big scale, it appears that there is a transparency on 1 pixel !!
But there are pixels remaining everywhere !! 😉 I’m really astonished !

In my case, I found a solution by changing radically the precomp size, in order to make it the half of the symetrical comp. It solves my problem, but this lack of accuracy is frightening for future projects.

🙂

[Lucia Bastenhof]

Darby, I didn’t see your question about blending modes sooner.
The blending mode is “Normal”, but indeed, in the precomp, the original image is the matte of another image.
I changed this to see what happened in the symetrical comp : the black line remains. Switching the Render Quality button makes it disappear, but the quality of the whole image becomes bad.
So, the matte in the precomp isn’t the source of the problem.

In this case, I don’t need masks finally, but it could happen on another project, and I can’t imagine there isn’t a way to make disappear this line (without damaging the quality of the image).

If you have any idea… 😉

[Lucia Bastenhof]

I tried what you said, Darby, and indeed, the mask isn’t accurate. Trying to fix it, I’ve noticed that it succeeded by switching the render quality button (with the dotted line).
Glad to find that, I went to try it on the masks (already done) of my symetrical composition.
As a result, the black line becomes white, and of course the quality of the whole photo is suffering.

In this case, it doesn’t matter (thanks to Dave ! 🙂 , but I’m sure the problem will come again in another project. That’s why if you have any idea about how to fix it, I’ll listen to it thoroughly.

Thanks ! 🙂