Track mattes are the primary method by which you can use the alpha or luminance information of one layer to set the transparency of another layer ( Figure 3.37 ). This is actually the normal way to apply a matte channel in many compositing applications, which don't have the concept of the alpha channel so firmly integrated into the overall pipeline as After Effects.
Figure 3.37 A basic alpha track matte setup: The alpha of layer 1 is set as the alpha of layer 2 via the highlighted pull-down menu. The small icons just to the left of the layer names help remind you that this relationship has been set up, and which is the color layer and which is the matte.
The perceptual difference between an alpha channel and a track matte isn't, for the most part, too difficult to grasp. In both cases, you have pixels with a value (in 8-bit color space) between 0 and 255, whether they are color or grayscale alpha pixels. That any image would be interchangeable with the alpha channel shouldn't be too shocking; even if you're using the luminance of a full-color image, After Effects simply averages the 0 to 255 value of the three color channels into one value. The principle doesn't even change in 16-bit color—it's just the same range of values, with finer increments, being sampled.
You set a track matte by placing the layer that contains the transparency data directly above its target layer in the timeline and choosing one of the four options from the Track Matte pull-down menu:
- Alpha Matte: Uses the alpha channel of the track matte layer as if it were the alpha of the underlying target layer
- Alpha Inverted Matte: Does the same as Alpha Matte but inverts the result, so that the lighter areas of the alpha are transparent and the darker areas are opaque
- Luma Matte: Uses the luminance data of the track matte layer (the relative brightness of the red, green, and blue channels combined) as if it were the alpha of the underlying target layer
- Luma Inverted Matte: Does the same as Luma Matte but inverts the result, so that the lighter areas of the alpha are transparent and the darker areas are opaque
By default, the source layer for the track matte (the upper of the two) has its visibility turned off when you set the track matte, which is almost always what you want. This sets the track matte with a single click to the Track Matte pull-down menu. If you need to adjust the source after setting it, you can always temporarily turn it back on; just remember to turn it back off when you're finished.
Why They're Useful
Track mattes are not the only way to apply alpha or luminance data to the transparency of a layer, but they're the clearest and most straightforward way. They help you out of a lot of jams in which creating a selection via other means would be inconvenient or in which mask or matte features constrain you.
For example, it's not possible to track a mask in After Effects. But it is possible to apply the mask to a track matte instead, and then to track that layer (instead of the mask itself). Chapter 8, "Effective Motion Tracking," discusses this in detail.
In Chapter 6 you will learn of the many ways to pull a key, including hi-con mattes. Some of these methods apply their results directly to the alpha channel of the target layer, but others cannot. Additionally, specific operations such as blue-screen keying can change the color of the source (automatically removing blue spill from the foreground); applying the key via a track matte to a duplicate clean source instead is an effective workaround if you want the matte but not the despill.
Why They're Occasionally Tricky
Unlike with parented layers (described in Chapter 2, "The Timeline"), selecting a track matte layer does not lock it to the current layer. In fact, setting a track matte means that no matter which layer is next higher in the stack, that layer is the source of the matte. Thus, accidentally moving a layer in between a layer and its track matte can cause easily solvable but, nonetheless, disastrous results.
When you duplicate a layer with a track matte activated (Ctrl+D/Cmd+D), After Effects automatically duplicates it above the matte layer or two layers above the current layer. If you duplicate the matte layer at the same time, the duplicate will also move up two layers, so that all layers preserve their proper track mattes ( Figure 3.38 ). That's good. What's bad is if you forget to duplicate both layers, because the track matte remains active in the duplicate layer, even if it has been duplicated to the top of the stack, and you can't see that it's active.
Figure 3.38 Selecting and duplicating the layers from Figure 3.37 creates two new layers that leapfrog above the previous layers to maintain the proper color/matte relationship in the source and duplicate layers.
If the layer to which the track matte is applied already has an alpha channel, then the new selection area created by the track matte will be opaque only in the areas that are opaque in both mattes. So applying a track matte in this situation is like making a subselection of the current selection.
Also tricky to work out with track mattes is the render order: Sometimes adjustments and effects that you apply to the matte layer are applied to the target matte, and sometimes you must first pre-compose to get the applied effects and adjustments.
And what happens if you apply a track matte to a track matte? It's actually hard to say; sometimes it will work, sometimes not. The user interface does not prohibit this behavior so you can try it, but a better idea is probably to pre-compose the first instance of track matting and apply the second track matte to that nested composition.
The next chapter, "Optimizing the Pipeline," looks in depth at solving issues related to render order, and you'll begin to use After Effects as a visual problem-solving tool for these issues.