[Mike Clasby]

It’s a render order thing. Remove the mask, then precomp the solid with the numbers effect on it (Layer>Precompose (move all attributes)). Then mask that precomp layer.

It’s weird but about half off the problem in AE can be solved by precomping, OK, maybe a slight exaggeration.

[Mike Clasby]

The only way I know is to duplicate the comp, then swap in the new layers by Alt-Click dragging them from the Project window onto the selected layer, one by one for each of the 16 layers. This should be faster, but not one click.

You’ll end up with effects, keyframes, expression, everything.

[Mike Clasby]

Good question, I will use this.

Here’s a way that’s not too bad.

Overview:

1) Create each wall with a Dan Ebbert’s Grid expression.
2) Create floor and roof similarly.
3) Drag those comps into the main comp and run a script, nabScript’s create Cylinder.jsx, to create a cylinder (will explain later).
4) Move the floor and roof into place.

Details:

1) Create each wall with a Dan Ebbert’s Grid expression.

Here’s a nifty Dan expression, as a test I used 9 dupes of a single video (640×480) to get finished wall of 1920×1440. That three rows and three columns of the same video. I’m assuming you’ll use different videos for each one.
Make the 9 layers 3D, the put this expression on position (Copy the expression, Alt-Click the Position Stopwatch, Paste, click outside the box):

numRows = 3;
numCols = 3;

row = Math.floor((index – 1)/numCols);
col = (index – 1)%numCols;
x = col*width + width/2;
y = row*height + height/2;
xOffset = (this_comp.width – numCols*width)/2;
yOffset = (this_comp.height – numRows*height)/2;

[x,y] + [xOffset,yOffset]

To apply the expression enmass, Select the expression in that first one you did, then “Edit>Copy Expression Only, select the other layers you want the expression on and Paste.

So if your individual wall is made up of more videos, change the number of rows and columns as needed (lines 1 and 2 in the expression). You should have a video wall. The expression assumes the layers are at the top of the stack, so if you add a camera to look at things, put it at the bottom.

2) Create the floor and wall similarly. Because my videos were 4×3, you’ll need 12 videos with four rows, (to fit the walls right, you’ll see) so change line 1 to:

numRows = 4;

3) Drag those comps into the main comp and run a script to create a cylinder.

Add the wall comp, floor and ceiling into the main comp, make them 3D and turn on the Collapse Transformation switches).

I choose a script to make a cylinder because it’s the only one I know that will work for this. A four sided cylinder is that same as the walls of a room. Here’s an earlier post on downloading, installing and running the script:

https://forums.creativecow.net/readpost/2/915391

Be sure to get createCylinder, not createCube (that uses solid’s and is not what you want).

To run the script, select the four walls (not, the floor and ceiling) and File>Run Script>craeteCylinder.jsx. The first parameter corresponds to the number of layers per level, so make that 4 to get 4 walls; the second to the Radius (rayon) of the cylinder, so make that half the width of the walls, or 1920 divided by 2 or 960 . You can change the radius on the slider ( E on the Null reveals).It creates a cylinder with the wall parented to a CONTROL layer (Null) so you can rotate/reposition the cylinder.

Click on the eyeball of the wall facing you so you can have one end open.

4) Move the floor and roof into place.

First add a camera if you haven’t already and move it (C rotates thru the camera orbit tools) so you can see the cube from afar.

Select the Ceiling layer and grab the little BLUE handle (Z will appear) and slide it up until its where you want.

Do the same for the floor, sliding it down. You can also change the Z position manually as normal.

Now parent the Floor and Ceiling to the Null as the walls are already. If you want to rotate the room, rotate the Null. The also works for moving the room, by changing the Null’s position.

This should be what you’re after.

Again, thanks for asking the question.

[Mike Clasby]

I think it would be easier (not easy, but easier) to make the whole canvas or screen one large Solid, then animate the individual mask, by setting mask shape keyframes, the first a square, the second, down the timeline that lizard. You’ll have to add mask points on that second keyframe to get around the outside of the lizard. Put those first four vertices on the tips of the lizard toes, then add mask points in between to match the outline. To rotate one lizard at the end, doubleclick a mask point and you get a little transform box like in photoshop to move and resize, rotate the mask. That way the mask shape will morph from square to lizard.

When you have what you like, then duplicate the mask and move those two keyframes to the new starting and ending spots. A quick test looks like it’ll work.

Mask shape changes are notorious for quirky behavior, so checkout “Smart Mask Interpolation” in the Help files. There’s a lot you can do by choosing the correct first vertex.

[Mike Clasby]

This should work. Here is an earlier thread on how to spread out layers in 3D space with a slider. I set it up for z space, but you could add sliders for x and y spread. Here’s the thread:

https://forums.creativecow.net/readpost/227/8455

Just add new sliders, one named xSpread, and another ySpread, and one for zSpead, then add this expression to position:

xOffset = thisComp.layer(“Leader”).effect(“xSpread”)(“Slider”);
yOffset = thisComp.layer(“Leader”).effect(“ySpread”)(“Slider”);
zOffset = thisComp.layer(“Leader”).effect(“zSpread”)(“Slider”);
x = thisComp.layer(“Leader”).position[0] + (index-1)*xOffset;
y = thisComp.layer(“Leader”).position[1] + (index-1)*yOffset;
z = thisComp.layer(“Leader”).position[2] + (index-1)*zOffset;
[x, y, z]

The expression will distribute the layer from “Leader”‘s position, so if you change that line to the name of the layer for your Maya locator (I’m lost about Maya), then it should work. You’ll be able to control their spread with sliders.

Alternatively, there is a neat tutorial on crowd generation with Trapcode’s Particular, search here and maybe Google, or look at Trapcode.

[Mike Clasby]

There are several ways to do what you want, I think the easiest is with this script. Basically it’s just select the layers you want in the cylinder, then run the script, putting the number of layers to be in the cylinder in the first box and the radius in the second, details on installing the script and running it,here:

https://forums.creativecow.net/readpost/2/915391

It creates a Null with slider for radius and so you can move/rotates the cylinder, just rotate the Null.

There are several other methods, read the whole thread for other options.

Before this script I really liked that setup, “sphereoscarus”, here:

https://forums.creativecow.net/readpost/2/915236

Then just swap your layers in, bit now I think the top one is easiest.

[Mike Clasby]

Don’t really know what you want, there are so many homemade versions of Buffy and Charmed openings at youTube; I dunno what was the original.

[Mike Clasby]

Just render out that 1:06 worth, then when you Import, right click the footage in the Project Window and Interpret Footage>Main, then choose how many times you want it to loop (down near the bottom), twenty it sounds like, now the footage will loop twenty times and you won’t need to render out the 20 second plus clip.

You could also put your little loop in a comp of it’s own (or precomp them), then put that in the main comp and to that precomp ;layer, Layer>Enable Time Remapping, then add a loop expression to Time Remapping, (Copy the expression, Alt Click the Time Remapping Stopwatch, then Paste):

loopOut(“cycle”)

This will cause those 16 frames to loop endlessly, not rendering needed. It loops those two time remapping keyframes.

At some point you’ll need to drag that right edge of the clip to the far right as when you apply time remapping, now you can extend the clip to the right.

[Mike Clasby]

Well you make a Luma Matte. Add a Black Solid, put Ramp on it, adjust the start and end of ramp so it goes from black to white where you want the area to be brightened, white for the brightest, then with the footage on the bottom, select Luma for a track matte. But if you use a Linear Ramp, the line of brightness won’t have any curves to conform to the contour of the body that’s masked off. And I’m assuming levels is applied to the masked layer below the Black Solid.

And you’d need another copy of the layer below.

1)Black Solid with Ramp (eyeball will be off when the layer below set to luma matte)

2)Footage with mask, Levels… and TrkMat set to Luma Matte

3)Footage

[Mike Clasby]

Do the masking on a duplicated layer, and set the first mask to Add (default) and the second mask to “Intersect”. Then apply levels to that top layer with the masks. It should work.

Here is a great tut on matching lighting that has a few neat tricks that might be of help.

https://library.creativecow.net/articles/coots_brendan/greenscreen.php

Andrew does a similar thing here:

https://www.videocopilot.net/tutorials.html?id=38