[Steve Bentley]

It might actually be easier (at least in getting it more naturalistic) to use a 3D camera and make the layer 3D as well. The problem with scaling things is that scaling is logarithmic. If you zoom (but with the scale parameter), the perceived speed of that “zoom” gets faster and faster – when you scale something to 200 percent it’s not twice as big, it’s four times as big; when you scale to 400 percent its now 16 times the area of the original. Whereas with a camera you can move it in and out and so it acts more like what a real camera (with a prime lens) would do.

Ignore all this if you are looking for that “Battle star/attack of the clones” zoom-and-find-the-hero kind of guerilla news reel stuff. Scale can work fine for that kind of thing.

[Steve Bentley]

Yes and no.

All normal AE solids layer and shapes are 2D in nature, – even when you check the 3D box, its just a 2D layer tilted back in 3D. All paths you draw in AE are drawn on the flat surface of these layer types. So technically you can’t take a few nodes on the path and push them into the Z.

However there are work arounds.

You can use TrapCodes 3D stroke (and Form and Mir). This doesn’t really make a 3D path but it warps the shape of the path you have drawn and gives the illusion that it’s tilted back in 3D. It is camera aware so it really feels like you can move around the path in 3D. It even thickens the line so when you look at it, from what would be the edge on the layer, it still has thickness. The catch is any manipulation of the points along the path are still restricted to the original flat plane x,y coordinates – with one exception: you can bend the path around an axis. While the overall path shape remains the same, the bending will curl the lines into what looks like the layer’s Z space. Keep in mind that if you curl the lines around in a “C” shape, you can’t really put another layer in between the ends of the C’s “jaws”. It’s still a single flat layer in the layer stack so anything above the path layer will appear in front and any layer below will appear behind – again, with one exception: you can use two versions of the path layer – one in front and one behind other layers and mask the depth of the line so the illusion of behind or in front of other layers gets really persuasive.

Then there are other plugs ins like BOA That have taken this further and get closer (in fact, dead-on in some cases) to an actual 3D line.

Then there’s Element 3D, that is actual 3D geometry, albeit in a separate program that is hosted by the AE interface. The Element render engine is doing the work to generate actual 3D stuff but the layer that you put the Element effect on is still a flat layer so again the “anything below and anything above” rule still holds as far as the layer stack goes, and you’re back to a flat render of a true 3D object. Element does have depth masking though so you can get pretty close.

Finally, there is C4D. It does what Element does but in a completely separate program and you can either import the renders as movies or you can import the C4D project and have AE render them for you among your stacked layers – but again that render is just a flat card in the stack.

These last two options are good because you can transfer spatial data from AE to either of them and position your 3D paths exactly where they need to be in Z space.

There’s probably a few other options (like Zaxwerks) that others will mention that I haven’t thought of here.

[Steve Bentley]

Depending on the problem you can pull the final render back into AE and render out at 29.97 to match your playback specs.

[Steve Bentley]

Perhaps what you are seeing is the mesh once its separated. With the triangles together the phong shading creates a smooth transition between polys (that are less than a set angle of difference). As soon as you separate the polys the phong tag is rendered pointless and you see each triangle as a flat surface and any reflections are severely harsh.

[Steve Bentley]

I assume the smaller inner comp is because it only fills that many pixels of the outer comp? Or are you really blowing up the inner one in the outer comp and hoping the CR button will fix and sharpen?
The CR button can only do so much – too small a composition from illustrator tends to get soft if blown up too much.
As for page turns or other 3D effects, you can always precomp your inner comp to make it the size you need and CR that, and then precomp again and then add the page turn on the outer most comp.
If you are trying to keep the resolution at its best in the outer 3D comp, as long as the inner comp is 2D you can CR it within the 3D comp and it will rasterize for each frame so no matter how close you zoom in 3D it will stay sharp (within reason).

[Steve Bentley]

You can paste the copied vector paths into an empty mask (or any mask) at specific keyframes and the shape you paste on that key frame will be the shape the mask morphs to. You can do this all the way down the timeline.
If the paths have a different amount points from each other the paths may “chainsaw” as they change shape.

[Steve Bentley]

Do the nulls childed to the jet packs have a external comp tag on then in c4d?

[Steve Bentley]

There isn’t a frame rate difference between nested comps is there? Like 29.97 in the outer master comp and maybe 30 in the inner? Or footage at a diff frame rate from the comp?

[Steve Bentley]

If it does end up to be an alpha issue, you could try changing the alpha to premult or straight.

[Steve Bentley]

To keep the eyes lips and ears from disappearing among the foliage, we always keep these as separate models (but tied to the same rig). This also helps with poly counts – these areas usually have higher poly concentrations but when you do the surface mapping with mograph or particles the mapping is often based on the source object poly count so it can be tough to get the kind of distribution you want when the body has one kind of poly concentration and the lips have a higher one.