Collapse Transformations and Continuously Raster problem

[Roland R. kahlenberg]

I don’t see the point in nesting the AI layers that are of the runner. Just leave the Timeline structure as it is in the first image. Nesting is overated/abused – use it to your advantage, not to your disadvantage; as the case seems to be here.

Unfortunately, I can’t tell what’s wrong with your setup without looking at the AEP BUT what you have in the Timeline for first image looks like how I’d set it up myself. However, I am curious why you aren’t using AE’s 3D lights and shadows?

HTH
RoRK

Intensive AE & Mocha Training in Singapore and Malaysia
Adobe ACE/ACI (version 7) & Imagineer Systems Inc Approved Mocha Trainer

[Scott Gaskin]

Thanks Roland,
The reason I would prefer to pre comp the footage is because there are going to be another 7 runners added to the scene and it’s going to be a pain to manage with so many layers. I presumed this is what pre-comping was used for. Just buggin the hell out of me why it’s doing what it’s doing. As regards to 3D lights, I did try this but have to admit ive not had too much experience with using the lights and didn’t look right. I applied the cast shadows to it but it didn’t cast the floor shadows of the runner. Should I be using more than one light?
Thanks for your help
Scott

[Walter Soyka]

[Scott Gaskin] “The reason I would prefer to pre comp the footage is because there are going to be another 7 runners added to the scene and it’s going to be a pain to manage with so many layers.”

You might look into Zorro the Layer Tagger [link] to help with layer management.

Precomps should be used primarily when an element needs to be reused, or when you are forced to by an Ae requirement. Using precomping solely to reduce the number of layers in a timeline can complicate animation, as you can’t make changes live in context in the main comp view.

Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events

[Scott Gaskin]

Thanks Walter,

I really thought that one of the main purposes of pre-comping was to help manage work flow like this particular example. I have seen several tutorials where it is done for this reason.

As far as Im aware, I havent done anything too complicated in my original comp. There is some parenting of layers, postion and rotation keyframes so I not sure why its causing the problem.

May it be a layer order stacking problem? I have read about the render order and that the way certain effects and tranformatioms are added can cause some problems. Dont understand how this would be the cause though as I’ve not added any effects to the main character, just transformations so am stuck as to what the cause is.

Thanks

Scott

Thanks for your response.

[Walter Soyka]

[Scott Gaskin] “I really thought that one of the main purposes of pre-comping was to help manage work flow like this particular example. I have seen several tutorials where it is done for this reason.”

Yes — but I agree with Roland that pre-comping here is probably not a help to workflow. You’d reduce the number of layers in your main comp, but you’d have to jump across 8 different comps (the main comp and each of the runner comps) to animate, and you won’t be able to see your work live in context (because ETLAT [link] can’t give you live views).

[Scott Gaskin] “As far as Im aware, I havent done anything too complicated in my original comp. There is some parenting of layers, postion and rotation keyframes so I not sure why its causing the problem.”

It could be the rotation or placement in Z-space. If the runner is centered in your precomp, but out at the edge of the main comp, you will get different camera perspectives on 3D objects, because the objects are placed differently relative to the comp cameras.

Imagine placing a 3D cube in your Ae scene. From perfectly straight-on (when the cube is in the center of the frame), it looks more like a flat square. As you change the viewing angle (or move it out to the sides of the frame), you see more of the sides of the cube.

If this is the case with your runner, it is technically correct to see two different results in the different comps, but of course will make animation quite difficult.

Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events

[Jon Bagge]

Regardless of whether you use precomps or not, your main problem is with Collapse Transformations. In most cases you don’t want to use it since it changes the AE render order. It causes transformations of the subcomp (transforms applied to the subcomp layer inside the main comp) to be applied to the layers inside the subcomp before effects and masks inside the subcomp.

For your type of project you can really only use Coll Trans if you don’t use any masks or effects at all to either the layers inside the subcomp or the subcomp itself. (at least not without causing massive grief)

If you keep everything in one comp like the others have suggested you can use Cont Rast on the Illustrator layers so long as you realize any effects on these layers will render after transforms. This is reverse of how it normally is. It will also render masks after transformations which makes them hard to use.

I have done quite a few project of somewhat similar nature to what you’re doing, and this is my method:

I use subcomps like you do. I make the subcomp about twice as big as it needs to be (or more). This means in the main comp I scale it down, and I can move it around and scale it up/down a bit while remaining sharp.
If you use Illustrator layers, scale the layers up so they fit this larger subcomp, and enable Cont Rast. This will make them sharp inside the subcomp. DON’T enable Coll Trans on the subcomp itself.
With this technique you can also add effects to the subcomp without running into trouble. You can also use adjustment layers and whatever else inside the subcomp.

For shadow, duplicate the subcomp layer inside the main comp, and use the second comp for shadow. You can then add tint effects, blur, and so on.
Or use 3D to cast shadows.

The general rule is: Don’t use Collapse Transformations on subcomps unless you have a very good reason. If you must, there are severe limitations you have to be aware of.

Bit of a long explanation, but hope it helps.

————–
https://www.jonbagge.net
Jon Bagge – Editor – London, UK
Avid – FCP – After Effects

[Scott Gaskin]

Thanks for this Jon. Your suggestions sound like a good plan.
I have however since gone back to AE and pre-comped my illustrator layers and kept the pre-comp within the main comp and the problem seems to be solved even with Coll trans applied. (I have no effects added to the layers) Its when I bring the precomp into another comp that caused the problem! Confusing! I’m obviously doing things the wrong way but hopefully learning!

Thanks for your help

Scott