[Steve Roberts]

I think you need something such as Conoa 3D or a 3D app — something that maps to a 3D cone better than AE.

Yep, taking an object along a diagonal path, then applying polar coordinates should get you a spiral. In a 3D app, to get a seamless spiral mapped onto the cone, you could map the spiral from the top using flat projection, rotated P=-90. In Cinema 4D, anyway.

[Steve Roberts]

… and Jon, of course.

[Steve Roberts]

… and Serge, too!

[Steve Roberts]

There was some jitter in the trees, but what can you do?

It was suggested by a colleague that I recreate the whole scene in CG.

We had a good laugh over that.

[Steve Roberts]

You must use Flash, Director, HTML, a programming language (Visual Basic?) or something else to create interaction. AE, Cinema 4D or MAX will not do it.

[Steve Roberts]

Not sure. Can you download the Wimpy FLV Player to play the FLV outside of Flash, just to help with troubleshooting?

[Steve Roberts]

To make movies small enough to load well within Flash, I recommend:

1) Swift3D, or
2) rendering to Flash format within Cinema 4D, or
3) using AE to render a SWF. Read the AE Help on “SWF” before doing that.

Any of those three will *not* be interactive. They will only be movies. To get interaction, you will have to import them into a Flash project, and do the programming there. For more information, check out a Flash forum.

[Steve Roberts]

In general, when deciding whether to precomp sections or not, you should determine how much your sections interact.

“Chemical Burn” would probably work well with precomping, since each text animation stands on its own. You would make the text animation in a composition (“nested comp” or “sub-comp”), then drop that sub-comp into the main comp, then slide the camera around, watching one sub-comp play out, then moving to the next.

(By the way, “precomping” and “nesting comps” get the same result, but nesting is what you do when you make the inner sub-comp then drag it into the main comp. Precomping is when you make the layers in the main comp, then group them together into one layer. Precomping is often done when you’re making things up as you go along, while nesting is a little more organized. Most people use both methods when it suits them.)

I think Pulp Fiction (I love that one) might have been done as a series of shots (cut between), where each shot might have been a separate comp. Within a shot, I think everything was in one comp, since the elements interact closely. Separating into sub-comps within a shot might make integration between text elements too difficult.

The Lion’s Roar looks like a series of sub-comps playing out. You can tell what’s a sub-comp when you go from shot to shot, or from scene to scene. When it moves to a new scene, it’s a new sub-comp. Probably.

I’d start learning how to do this by setting up a comp where words are all laid out nicely, then make the words just pop on to complete the sentence. Then I’d drag that into the main comp, and lay out a long line of text at 90 degrees to it. The letters of that line would pop on one ata a time. I’d lay this out using the AE “front” camera, just to get an overall view.
Then I’d make a new camera, and place it so the first animation plays out in front of it. Then I’d pull it back and rotate it 90 degrees to reveal the long line of text as well.

For the next scene, I could use a new comp with the same background as the first, and do something similar with a mix of sub-comps and elements in the main comp.

My point is: experiment with simple animations. Maybe you want to rotate the camera, maybe you want to rotate the two elements parented to a null. Sometimes rotating elements is easier, since you can lay out the whole thing from a bird’s-eye view (front camera) and watch the whole thing play out from that view. If you rotate the camera, it can sometimes get hard to keep track of which way you’re looking. That would work better when you’re cutting from one comp to the next, with only one camera move per comp.

Hope that helped.

[Steve Roberts]

Yep, apply a negative mask expansion so they don’t intersect.

The only way to fiddle with 3D intersections is by using Alpha Add mode to improve the look of the place where two layer edges meet. Transfer modes can be displayed by hitting F4. This toggles “switches” and “modes” in the timeline.

[Steve Roberts]

Buying an AVCHD camcorder is just asking for trouble. Read the earlier posts.