A composition consists of layers stacked on top of one another in the Timeline panel. When the composition is rendered—either for previewing or for final output—the bottom layer is rendered first. Within each raster (non-vector) layer, elements are applied in the following order: masks, effects, transformations, and layer styles. For continuously rasterized vector layers, transformations occur before masks and effects.

Transformations are changes to those properties grouped under the Transform category in the Timeline panel, including Anchor Point, Position, Scale, Rotation, and Opacity. What you see in the Layer panel is the result of the rendering before transformations are performed.

In a group of effects or masks, items are processed from top to bottom. For example, if you apply the Circle effect and then apply the Magnify effect, the circle is magnified. However, if you drag the Magnify effect above (before) the Circle effect in the Effect Controls or Timeline panel, the circle is drawn after the magnification and isn’t magnified.

After a layer has been rendered, rendering begins for the next layer. The rendered layer below may be used as input to the rendering of the layer above—for example, for determining the result of a blending mode.

If a composition contains other compositions nested within it, the nested composition is rendered before other layers.

If the Collapse Transformations switch is selected for a nested composition, then the transformations for the nested composition are not performed until after the masks and effects for the containing composition are rendered. This allows the transformations for the nested composition and the containing composition to be combined—or collapsed—and performed together. The same is true for vector layers that are not continuously rasterized.

Collapsing transformations can, for example, preserve resolution when a layer is scaled down by half in a nested composition, and the nested composition is scaled up by a factor of two in the containing composition. In this case, rather than performing both transformations and losing image data in the process, one transformation can be performed—doing nothing, because the individual transformations cancel each other.

If transformations are not collapsed, a nested composition that contains 3D layers is rendered as a 2D image of the 3D arrangement, using the default composition camera. This prevents the nested composition from intersecting with 3D layers, casting shadows on 3D layers, and receiving shadows from 3D layers in the containing composition. The nested composition is also not controlled by the containing composition’s cameras and lights.

If transformations are collapsed, the 3D properties of the layers in the nested composition are exposed to the containing composition. Thus, the nested composition can intersect with 3D layers, cast shadows on 3D layers, and receive shadows from 3D layers in the containing composition. The containing composition's camera and lights can also control the nested composition.

When a closed mask (with mask mode other than None), a layer style, or an effect is applied to a nested composition with collapsed transformations, the layers in the nested composition are first rendered on their own, then masks and effects are applied, and then the result is composited into the main composition. This rendering order means that the blending modes of the nested layers are not applied to any underlying layers in the main composition, and that 3D layers above and below the collapsed layer cannot intersect or cast shadows on each other.