AE 3D & continuously raster?
Lenie Ramos
Rick Gerard
Fredo Viola
Why do you have to have it continually rasterizing when dealing with nested comps? Let's say you want to add the layer's motion blur to a 3d comp; when you unclick continually rasterize the comp flattens out.
A client asked me to animate a logo and have a building in the logo build from bottom to top. So I made some 3d bricks that would interlock and comped them so that I could move them around more easily. I clicked the 3d button of the comp, put it at best mode, and then when I clicked the continuously rasterize button all the dimensions appeared. Great! Later I decided I needed to click on motion blur, but to do so I had to deselect continually rasterize, and in doing so I lost the 3 dimenstions....
Help! Was precomposing the wrong thing to do to marry the brick's 6 panels?
Rick Gerard
You can figure out the why and when easily when you understand the rendering order. Basically AE converts vectors to pixels first, then looks at position, applies masks, and then applies effects and then take those pixels and mix them with the next layer in the composition.
See if this helps. Take a 2D layer and drag it so that only half the layer is visible in the comp. Now pre-compose the layer and call it test. Now scale the “test” layer to 25%. You’ll see that only half of the layer is visible. If you collapse transformations, the entire layer appears. The collapse transformations option causes AE to look at all original pixels in their new position before anything else happens. Without collapse transformations applied, AE will apply any new transformation or effect only to the pixels that have been rendered in the original frame. Pre-composing without collapse transformation enabled produces exactly the same result as rendering the layer and importing that footage. The same thing applies to the layer when it’s in 3D space. When you collapse transformations, AE looks ahead to the original pixels before applying any other transformations.
Maybe a better term would make it easier to understand. “Use the original pixel information, the original position, and the original mask info before applying new position information” is probably too long. I hope this helps. Working efficiently in any compositing application demands that you completely understand not only what is happening, but when.
One more look at the original question also indicates a little misunderstanding. Continuously Rasterize and Collapse Transformations use the same button, but are not exactly the same thing. The correct question should “Why do you have to collapse transformations to use 3D properties in the original layer?” Now you know the answer.