Context-click on any of the column headings and you can reveal or conceal all the different types of columns available in After Effects. Don't bother revealing them all—several are only occasionally useful, and After Effects provides you with a better way of finding them in context.

Figure 2.1 shows a typical minimal column setup that still gives you instant access to all of the columns, from which you can toggle in one step to each of the columns that don't currently display. From there, you have several options:

Figure 2.1. You don't need to vary your view much from the default to have it about as streamlined as you can get. Here, the Parent column is hidden but all columns other than Parent and Comment are available in a single left-click or via a shortcut.

With unused columns hidden, you'll have more space for the columns you need most.

Comments and Color

Ever try to make sense of someone else's project? Chaos, right? Make it easy on your coworkers: Color code your layers and use plenty of comments. Choosing a distinct color for each layer can provide intuitive visual clues. It can even help you as you work. How you handle colors is really up to you; Figure 2.5 shows an example of a timeline with several layers that are color coded by type.

Figure 2.5. A unique color has been assigned to each type of layer (and to the composition itself), making it easy to quickly distinguish between them.

The Comments column is probably the most neglected, yet it enables you to say a few words about the purpose and use of a layer, which is especially useful when creating templates. But even better than enabling the Comments column is to add a layer marker. Press the * key on your numeric keypad, then double-click the layer marker, and enter text in the comment field. Unlike the Comments column, which requires quite a bit of that precious horizontal space and might quickly get turned off, layer marker text is written right over the layer bar itself, so you have as much horizontal space as the layer has.

Note

Pay no attention to the other options in the Marker dialog; the Chapter and Web Links fields work only when you are writing to a medium that supports them, such as QuickTime or SWF. It's a nifty feature, but you're likely never to use it.

Solo, Lock, and Shy

Three toggles on the A/V and Switches columns pertain to how layers are edited and viewed as you work, and because I've seen users who should know better miss or forget them, I'm including a quick refresher.

The small circular icon in the A/V Features (Figure 2.6) column is the Solo switch. After Effects decided not to copy Photoshop this time around, so Alt/Option-clicking on the layer eyeball does not solo it—this switch does. The reason for doing it this way is that too many of your layer visibility options in After Effects are permanent in nature to have them inadvertently toggling on and off. Solo a layer and you see only it; send a comp with a layer toggled solo, and the Solo Switches menu in Render Settings determines whether it renders that way.

Figure 2.6. You probably already use the Solo switch, engaged here for the bottom layer; remember that leaving a layer solo in a subcomposition will display it that way in the main comp, a common gotcha.

The Lock toggle is really great in situations where you are certain you no longer want to inadvertently edit a layer (or you want to signal to other users of the project not to edit it). You can't even select a layer that is locked.

Shy is trickier to new users; toggle a layer to shy and nothing happens until you also activate the Shy toggle for the whole composition (Figure 2.7). Then suddenly, the shy layers disappear from view in the timeline, although they will still be visible in the comp itself, if they were previously. This is a way to temporarily reduce clutter in your timeline, but beware—if you're left scratching your head as to where that layer went, when you can plainly see it's still there, take a close look at the Index numbers: If they are no longer in sequential order, then the missing number belongs to a layer that is now shy.

Figure 2.7. Hey, why are those layers numbered 1, 3, 4, 6, 7? The track matte layers have been set to shy and hidden by enabling the Shy toggle at the top of the timeline (highlighted). Very useful for showing only the layers you're still adjusting, but don't forget about those hidden layers.

More than any other window in After Effects, the Timeline window is the place where keyboard shortcuts—lots of them—come into play. There's a heck of a lot you can do in the timeline via either a single keyboard shortcut or a single click of the mouse, where a beginner would use several clicks of the mouse.

Time Navigation

Most users learn time navigation shortcuts right away, while others are content simply to drag the time needle around all of the time, a policy I oppose on the grounds that it quickly becomes tedious. Here are a few handy navigation shortcuts:

• The Home, End, Page Up, and Page Down keys correspond to moving to the first frame, the last frame, one frame backward, and one frame forward. Shift+Page Up and Shift+Page Down skip ten frames backward or forward, respectively. Press Shift+Home or Shift+End, and you are taken to the In and Out points of the work area.

• Click on the blue time status at the upper left of the timeline to open the Go To Time dialog (press Ctrl+G/Cmd+G). Here, you can enter a new time. There's no need to add punctuation, so 1000 is ten seconds in Seconds mode or frame 1000 in Frames mode.

• To navigate an arbitrary but precise number of frames or seconds (say, 48 frames after the current time), in the Go To Time dialog, replace the current time with your increment, in the format +48, click OK, and After Effects calculates the increment for you.

But here's a gotcha: If you need to navigate backward in time 48 frames, you can't simply enter –48 in the Go To Time dialog. If you do, you're transported to negative 48 (frames or seconds, either of which probably moves the time needle right off the timeline). Instead, you must use the format +–48; that's right, you're adding a negative number, not subtracting. It's weird, but it works (Figure 2.8).

Figure 2.8. Setting this value to –48 would set the current time to that negative frame value, probably not what you want. But entering +–48 adds an offset of –48 frames, which might very well be useful.

Note

The add-a-negative method for offsetting values works in many number fields in the program. For example, you can subtract 30 degrees to Rotation by entering +–30, and you can multiply or divide as well, using the * and / symbols, respectively. This won't work, however, in certain number fields outside the timeline, such as those in the Composition Settings dialog, where it would occasionally be quite useful.

Timeline View Options

Now that I've barraged you with useful shortcuts, what about interacting with the view of the timeline itself? I'll assume you're familiar with the process of setting keyframes and move right on to some controls for zooming in and out of keyframe data:

• The ; key toggles between a view that's fully zoomed in on the current frame and the whole timeline view.

• The slider at the bottom of the timeline is handy for zooming in and out more selectively (Figure 2.9).

  

  Figure 2.9. Unless you have a mouse with a scroll wheel, this click-and-drag interface at the bottom of the Timeline window is probably the fastest method of zooming in and out.

• If you have a mouse with a scroll wheel, there's a cooler option: Not only can you scroll up and down the layer stack, and Shift-scroll left and right in a zoomed in timeline view, but Alt-scrolling (Option-scrolling) dynamically zooms you in and out of time. All of these operate simply by having the cursor positioned over the Timeline window.

Your time needle is actually capable of going where your Timeline window cannot: to time frames that are below the first frame of the composition or beyond the last frame. This needle disappearing act typically happens after you navigate the timeline in the Layer window and then return to the Composition or Timeline window. Have no fear, clicking where you want the time needle to be or using any of the time shortcuts will save you.

Note

The scroll wheel also zooms in and out of footage view windows when the cursor is placed over them.

You can, of course, view the timeline without seeing the associated Composition window; it's just tricky to unlink the two because making it easy would be confusing to beginning users. If you Alt/Option-click on the box to close the Composition tab, you close the Composition view while the Timeline tab stays active. This “headless” timeline is useful for making adjustments in one timeline while seeingthe results in another comp (Figure 2.10). I'll cover this more in Chapter 4. If you need to bring back the Composition view for your timeline, click the Comp button at the right edge of the timeline (Figure 2.11).

Figure 2.10. There's an easier way to preview one comp while continuing to adjust another related one, or several others. Activate the Always Preview This View toggle in the lower left corner of the Composition window. Now, whenever you set RAM Preview, that's the view that will load and play.

Figure 2.11. Clicking the Comp button reveals the Composition window for the current timeline; this is helpful if you have closed the comp view while leaving the timeline open.

Just above the Comp button is the comp marker bin, where you can click and drag out a marker to a point on your timeline. Markers are numbered sequentially, and you can also add whichever number you like at any time (even when a RAM preview is up) by pressing Shift and one of the numbers at the top of your keyboard (not the numeric keypad this time). These aren't quite as cool as layer markers (which you can add at any time with the asterisk key) because you can't label them, but they're sometimes useful anyhow. Holding down the Shift key as you drag the time needle snaps the needle to any comp or layer markers that are present.

Replacing Layers

Sometimes you need to replace the source of a layer with an alternate take or a different element in the Project window. You can easily do this by highlighting the layer to be replaced in the timeline and holding down the Alt (Option) key while dragging the new source to the timeline. An even easier method is to select both the existing layer and the new source and then press Ctrl+Alt+/ (Cmd+Option+/).

All of the settings from the previous layer, along with any applied effects, translate to the new layer automatically. Beware if you are replacing with source of a different size or aspect, as the mask and transform values may only be appropriate to the predecessor.

Note

If your replaced layer has a time that mismatches with the previous source and there's no simple fix such as extending the end point, see the section called “Playing with Time Itself” later in this chapter.

While we're on the subject, what do you do to neutralize a layer that is taking up a disproportionate amount of render time? I recommend that you not replace it in this manner; Chapter 4 covers some much more elegant methods for assigning proxies and pre-rendering embedded compositions.

The most common way that keyframes are used in an effects situation is to say, about a given value, “I want it to be this value at this point in the shot, and this value here.” By default, After Effects sets linear temporal transitions; in other words, you travel from keyframe A to keyframe B at a steady rate, then from keyframe B to keyframe C at a separate but still steady rate.

Many basic animations are greatly aided by adding eases. Experienced animators and anthropologists alike will tell you how rare it is that anything in nature proceeds in a linear fashion; nature is full of arcs and curves. A classic example of this is a camera push: In most cases, with a real camera operator, a push would start and stop gently, moving at a steady rate only at the middle of the transition.

Flips and Flops

That's all very well, but how do I flop a shot? Unlike many other compositing and editing applications, After Effects has little in the way of dedicated transform effects, so searching in the Effects palette for flop is not going to help you. For those unfamiliar with the term, to flop a shot is to invert it horizontally, on its Y axis (as opposed to flipping a shot upside down on its X axis). To flop, you toggle Constrain Proportions off and set the X value to–100 (Figure 2.12).

Figure 2.12. To flop a shot, be sure to uncheck Constrain Proportions before setting the X value to –100%.

Easy Ease

A quick way to deal with easing is to apply the Easy Ease keyframe assistants. To apply them, context-click on a keyframe. Now, look under the Keyframe Assistant submenu, or use the keyboard shortcuts. Clicking F9 applies Bezier interpolation, creating eases in and out of a keyframe, while Shift+F9 creates an ease into the keyframe only, and Ctrl+Shift+F9 (Cmd+Shift+F9) creates an ease out. For the camera push example, you would ease out of the first frame and ease in to the final frame, and you would add an ease in both directions if there were any keyframes in the middle (for example, if the speed of the push changed at some point).

To see the result, twirl down the property to which the eases were applied. Figures 2.13a and b show the difference between a default linear camera push and one to which eases have been applied. The keyframes have changed from linear to Bezier type, and rate of motion is described by a curve instead of a straight line.

Figure 2.13a. With no eases, the Speed: Position graph is flat, moving at a steady rate of 34.72 pixels/sec.

Figure 2.13b. Add eases and it arcs from 0 (at the beginning and end of the move) to 52 pixels/sec (at the center peak of the ease).

You must set Hold keyframes deliberately, by context-clicking on the keyframe and choosing Toggle Hold Keyframe from the menu or by using the shortcut Ctrl+Alt+H (Cmd+Shift+H). A Hold keyframe's value does not to interpolate to the next keyframe, which is indicated by the square appearance of the right side of the keyframe.

If you need a layer to reposition over time with no transitions (no keyframe in-betweening whatsoever) begin by setting the first frame as a Hold keyframe; all keyframes that follow it in time will be Hold keyframes, signaled by their completely square appearance (Figure 2.14).

Figure 2.14. A series of Bezier keyframes have been converted to Hold keyframes, evident by the square shape at the right of each keyframe and the speed of 0.0.

Keep holds in mind as a problem solver any time you see unwanted in-between information creep into an animation, which can particularly happen with Position animations.

Effects compositors are rarely, if ever, called upon to perform complex animations, but eventually your project may require it, and the keyframe interface of After Effects has long been one of its strengths. Rather than teach you to be a great animator in After Effects, though, I prefer to keep it simple and focus on what you need to know to keyframe effectively and to avoid common pitfalls.

When keyframing Position (or Anchor Point) data, there are in fact two ways you can adjust keyframes: temporally (on the timeline) and spatially (in the Composition window). To illustrate this, I'm going to use the most classic bare bones animation the world has ever known: the bouncing ball (Figures 2.15 to 2.18). This example will show you (rather than just tell you) what you need to know about adjusting spatial keyframes.

Figure 2.15. Under Preferences > General (Ctrl+Alt+; or Cmd+Option+;) the option Default Spatial Interpolation to Linear is off by default, producing this result when you create three or more position keys. After Effects also applies an automatic Bezier shape to any point that falls between two other points—the opposite of what you want for a bouncing ball.

Figure 2.16. Click the check box for Default Spatial Interpolation to Linear, restart After Effects, and you get this result: three keyframes that are linear in both time and position. Often this is they way you want moves to go, but it does not create the realistic arcs of a bouncing ball.

Figure 2.17. Hold down the Ctrl/Cmd key while clicking the middle keyframe to activate its Bezier handles, then Ctrl/Cmd-click on one of the handles to “break” them, enabling you to form the V-shape shown here, which naturally causes curved arcs to and from the bounce point.

Figure 2.18. Finally, temporal eases create a more realistic variation in motion speed. Selecting the first and last frame and pressing F9 causes eases to those frames and an effective acceleration in and out of the keyframe in between them. Look closely. The dots are spaced further apart (traveling faster) at the bottom and closer together at the top, where the eases are.

You'll probably never have to bounce a ball to create a realistic effect, so what can you possibly learn from the example? Let's recap:

So, even though there is an infinite variety of position animations available to you in After Effects, they will all be done with the basic constraints covered here. True, it can get a little trickier in 3D, but mostly it's just trickier to see what you are doing; the basic rules and constraints remain the same, with each Position keyframe having three basic transition types: Linear, Automatic Bezier, and Custom Bezier Shapes.

There's a bit of a gotcha possible when copying and pasting keyframe data. The key is to recognize which item is active when you copy and which is active when you paste.

By default, After Effects does its best to help you work this out. If you copy keyframes from a particular property, then simply paste them to another layer, After Effects pastes them to that layer's corresponding property. If that property doesn't exist, because it comes from an effect that isn't applied on the target layer, After Effects adds the effect containing the property to the target layer, but with only the property that you copied.

However, you can also apply keyframes from one property to another similar property. For example, if you want to apply a 2D Position keyframe to the position of an effect, such as the Flare Center position of a Lens Flare effect, you can do so, but only by selecting the Flare Center property in the timeline before pasting.

That's still pretty logical. Where people get screwed up is by not noticing, for example, that an effect in the Effect Controls window is active and forward when they attempt to copy and paste a whole layer. Figure 2.19 shows a classic example, where a layer does not have its properties twirled down, and so it's easy to miss that the Effect Controls window is forward and has an effect highlighted, so that's what gets copied, instead of the whole layer.

Figure 2.19. A classic example where you can get tricked: Five separate items are highlighted: the source comp in the Project window, the target layer, the Curves effect (in both the Timeline and Effect Controls windows), and the layer itself in the Composition window. If I click Delete, which item disappears? Hint: Check which menu bar is not grayed out. Further hint: The Effect Controls window.

Bottom line: If strange things happen when you copy and paste, immediately undo and try the copy operation again, selecting exactly the thing you want to copy and copying it with no intervening steps. Because even if the logic of how copy and paste works seems obvious, you can still get tripped if you don't pay attention. When something stops making sense in After Effects, there is almost always a good reason traceable to something you've just done (and can undo). This is more than you can say for much graphics software.

Twirl down properties that have keyframes with numerical values, and you'll reveal graphs that are useful for fine-tuning transitions, going beyond what is possible with just keyframe assistants, such as Easy Ease. You only have to understand what these graphs are trying to show you.

Figures 2.20a, b, and c show the timeline of the bouncing ball example, this time with linear keyframes, with automated eases (courtesy Easy Ease, F9), and with eases accentuated manually, the result of adjusting the curve handles. You will mostly use the Velocity graph to fine-tune eases. The easiest way to do this, especially if you're not yet at home with this graph, is to start with an Easy Ease adjustment. Draw the handles out further from the keyframe to extend the ease, or move them in closer to shorten it.

Figure 2.20a, b, and c. These three timelines show a progression, from a default linear animation (2.20a), to adding eases to all keyframes (2.20b), and finally to toggling the middle keyframe back to a linear keyframe by Ctrl/Cmd-clicking on it and hand-adjusting the other eases, to accentuate the first and moderate the second (2.20c). Do not feel the need to follow these as a “recipe for a bouncing ball.” Rather, they exemplify a progression of how you can use curves to refine an animation, and what that looks like.

Note

Take care not to extend handles so far that they overlap; this can cause curves to spike and buckle, leading to sudden jumps in the animation. If you see something strange while previewing, look for a sudden sharp jolt in the graph and back the handles off to smooth it.

The main thing to understand about the Velocity graph is that it is shows you the rate of change in keyframe values; the Value graph shows the change of the values themselves. The other important thing to understand, which applies to both of these graphs, is that they show relative values; 200% velocity makes a steep slope if the surrounding values are smaller, but is dwarfed by an added 600% velocity change.

You can drag the lower boundary of the graph view to enlarge it and adjust the viewable range by editing the upper and lower values, in blue. Checking the box next to the current rate resets the graph view.

Roving Keyframes

Say you need an animation to follow an exact path, hitting precise points, but you want the animation to progress steadily, with no variation in the rate of travel. This is the situation for which Roving keyframes were devised. Figure 2.21 shows a before and after view of roving keyframes being applied to the finished bouncing ball animation; the path of the animation is identical, but the keyframes are now evenly spaced. This won't come up often, but if it does, Roving keyframes will be very useful.

Figure 2.21. The middle keyframe was set to rove (by turning off the tiny toggle below the keyframe) and is now positioned in between frames 15 and 16; the keyframe's interpolation setting is no longer taken into account, and it moves to keep the curve created by the other keyframes smooth and steady.

Suppose you've been handed someone else's project and have to make sense of it; two tools will help you more than any others.

For a broad overview of the project setup, you can enable the Project Flowchart view; look under the Window menu or use the shortcut Ctrl+F11/Cmd+F11. This is it—the node interface option in After Effects (Figure 2.23), similar but far less interactive than the process trees at the heart of competing applications such as Apple's Shake, Eyeon's Digital Fusion, and Digital Domain's Nuke.

Figure 2.23. What node-based compositing application is this? Why, the decidedly non-nodal After Effects. You cannot interact with this view much, but along with being an occasionally handy way to dissect an unfamiliar project, this view shows that all compositing applications are nodal, when you get right down to it.

This view allows you to see how the objects in your project have been used and how they interrelate. The toggles at the lower left of the window show layers and effects, but these muddy the waters somewhat on a project of any complexity; for a project overview, toggle them off. The most intuitive view that fits well on a monitor seems to be Left to Right, which you can choose via the pull-down menu at the lower left of the menu. Whether you choose straight or angled lines is up to you, but you can clean up the view by choosing one of them with the Alt/Option key pressed.

Click on the plus signs at the left of each comp, and you can see the component source files. This is step one in dissecting the project: Understand what is used where.

Step two is to open the individual comp; you can double-click to open it directly from the flowchart. Turn off the Shy toggle at the top if it's on, highlight all of the layers, and press UU to reveal all properties that have been altered on all layers. Now preview the composition, stopping at any frame where you have questions and investigating which settings and animations apply to that point in time. If there is a nested composition, you can open it and investigate that in the same manner.

The most effective way to do this is to start with the top-level compositions, the ones that are not contained in other compositions, and work your way backward. While I wouldn't call this process speedy, if you approach it in this logical manner, it will diminish the sense of “Where am I?” that can come over users who come to After Effects used to a more node-based approach. It also opens new comps on tabs that array to the right of the currently open comp, helping keep the tabs in visual order.

Although every other shortcut pales in comparison with the überkey, there are several other highly useful shortcuts for dealing with keyframe:

Keep these tips in mind, and your keyframe edits should be rapid and painless.

To offset the values of multiple keyframes by the same amount, use rub text. If instead you edit one of the values, all of the keyframes will be set to that value—not what you want. Here are a few more tips for working with multiple keyframes:

There is no easy way to scale an animation spatially, but I can offer a method if you ever really need to do it:

The new animation automatically will be two seconds long with roving keyframes, making the cases in which this is useful even rarer.

But what the heck, maybe you can impress someone at a party with this trick—that is, if you invite other After Effects nerds to your parties.

There are a bunch of ways to edit the anchor point position, and the effect on existing Position data varies with each method.

The most straightforward method is to edit the anchor point with the Pan Behind tool (Y) right in the Composition window; doing so causes Position to offset to maintain its current appearance. For a similar result, if you edit the anchor point numbers in the timeline, it is like holding down the Alt (Option) key using the above methods: The Position data does not change.

If you want to edit animated anchor point data, you may have a difficult time in the Composition window. Activate the Layer window, and under the View pull-down, change the default of Masks to Anchor Point Path (Figure 2.25).

Figure 2.25. Masks, rather than the anchor point path, appear in the Layer window by default. Change that with this menu at the bottom of the Layer window.

When setting up (or, if necessary, dismantling) a parent-child relationship, you have a choice of methods. You can choose the target parent layer from the pull-down menu in the Parent column of the timeline. Or, you can use the pickwhip control next to the pull-down menu, clicking and dragging from the pickwhip icon of the child layer to the target parent layer. If you have several layers selected and want to parent them all to a single layer, either method will set them all to the same parent.

Parenting does an effective job of sticking even when you change layer order or you duplicate or rename the parent. You can select all of the children of a parent layer by choosing Select Children from the bottom of the context menu. You can remove parenting from all selected layers by choosing None from the pull-down menu.

Nulls and the Parenting Cure

Keep parenting in mind whenever you are trying to solve a problem that involves individualizing transforms. A prime example is the lack of separate transform channels in After Effects.

Position, Anchor Point, and Scale properties share a single keyframe for all three channels. Animators who are used to being able to keyframe these as separate channels sometimes find this a bit of a drag. There is a way to solve this problem with expressions sliders, but I'll leave that aside in favor of one that's easier to set up: a null axis hierarchy.

Null objects are quite useful, invisible, nonrendering layers that occupy only a single point in space, although they are based on layers with dimensions that show up in the Composition window (Figure 2.26). You can create a null by context-clicking an empty space in the Timeline window and choosing New > Null Object (you can choose it from the Layer menu, as well).

Figure 2.26. A null object in After Effects looks like the outline of a 100-by-100-pixel layer, but cannot show up in renders. The reason for showing the outline is to give you something to easily recognize and grab.

Consider an example of creating a hierarchy for 2D Position. This addresses a concern some animators have with After Effects: that it does not provide the ability to keyframe transforms on separate axes individually.

Figure 2.27 shows the basic setup, in which two nulls are created, one called X and the other Y; the layer is parented to Y, which is then parented to X. Ideally there would be a way to now lock the second (Y) Position value in X, and the first (X) value in Y, but there's not, so you must remember to adjust only the X Position value in X and the Y Position value in Y. If it's important to you to stagger keyframes between the two values, however, this is an effective-workaround.

Figure 2.27. You may never need to animate a layer on separate axes, but it's a good example of how a parent hierarchy can help you out. The foreground (fg) layer is parented to the Y control, which is in turn parented to the X control. The idea is to animate only the X Position value of X and only the Y Position value of Y, so that the timing of each channel can vary (which it cannot with standard keyframes).

Except for one thing: After Effects 6.5 has no curve editor, nowhere to see the animation data from two distinct channels together. So visually coordinating this edit may turn out to be a hassle. As a matter of full disclosure, I will add that I have rarely ever needed to set things up this way; I tend either to anticipate the problem by animating in a certain order to avoid it, or to use expressions to tie each animation channel to a slider (described in Chapter 10, “Expressions”).

A physical film camera typically has an angled shutter that can open in a circular fashion anywhere between a few degrees and a full 360 degrees (an electronic shutter does not behave the same, but never mind about that in this context). Theoretically, 360 degrees would provide the maximum amount of blur a shot could have.

The Virtual Camera Shutter

With a real, physical camera, there is a good reason for how long the shutter is open: Along with the aperture setting, it controls how much light is passed on to the film or video pickup. Low-lit scenes tend to be blurrier as a result of the shutter remaining open longer to gather more light. Other scenes may deliberately be shot with a slow shutter (and the aperture closed down to compensate) just to get some nice, streaky blur.

After Effects is unconcerned with light gathering and does not emulate the differing amount of light that would pass through a lens with a high or low shutter angle setting (nor when adjusting the Aperture setting on a 3D camera, discussed in Chapter 9, “Virtual Cinematography”). Therefore, you only get the desirable part—the blur itself.

After Effects, however, cheats by offering you double the blur overhead of a fully open shutter. The maximum setting for Shutter Angle in the Composition Settings dialog is 720, an impossibility in the real world (because a fully open shutter only has 360 degrees, or a full circle, to reveal).

A camera report may help you adjust this setting, although it's often good enough to carefully eyeball it by zooming in on an area where your background and foreground element both have blur and matching them (Figures 2.29a and b). If you do have a camera report and it gives you the shutter speed, you can calculate the Shutter Angle setting with the following formula:

Figure 2.29a and b. In 2.29a, the default Shutter Angle setting of 180 degrees is too heavy for this white solid masked over the front hubcap, which has been tracked to match the shot motion. In 2.29b, Shutter Angle was cut down by 50% to 90 degrees. In this case, there was no camera report to guide the decisions; they are based on blur that can be observed on the rest of the moving truck.

This isn't as gnarly as it looks, but if you hate looking at formulas, try thinking about the relationship this way: If your camera shoots at 24 frames per second, but Shutter Angle is set at 180 degrees, then the camera is exposing half the total time of each frame (180/360 = ½), or 1/48 of a second. If your camera report tells you that the exposure was 1/96 of a second, the Shutter Angle setting should be 90 degrees.

When Blur Is Needed Where None Exists

Motion blur is cool and automatic, if you animated the layer to receive the blur in After Effects. However, if the element comes in already animated but without sufficient blur—as can happen with 3D elements or when a scene is retimed to be faster—you need a different solution, because After Effects has no motion to sample to generate blur.

The painful way to deal with this is to apply directional blur to the elements in question. The problem is that that natural motion blur is rarely unidirectional.

In such cases, and when there's no possibility of reshooting or re-rendering, it's worth considering an investment in Real Smart Motion Blur by RE: Vision Effects (www.revisionfx.com). This plug-in can sample motion from other layers and enhance the amount of blur created by that motion. It's render intensive and can require some tweaking, but it can be just the thing for such situations. A demo version is available on the book's CD-ROM.

The Shutter Phase setting determines when the shutter opens; in other words, the blur doesn't have to start at the beginning of the frame, the shutter could open between frames. Here's the trick: the Shutter Phase is meaningful as a fraction of the Shutter Angle.

In other words, a setting of 0 means the motion blur starts at the beginning of the given current frame. A –90 setting with a 180 Shutter Angle equates the blur starting –0.5 frames before the frame or in between this frame and the previous one. This setting is often used with motion tracking, where you are matching the middle of a motion-blurred target (not one end of it).

Consider a Shutter Phase setting of –50% of the Shutter Angle when working with motion-tracked plate (background) footage; otherwise, leave it at 0. The amount of blur won't change, only the timing of it.

Ricochet

One place a –50% Shutter Phase setting is used is when tracking in an object that ricochets, such as a bouncing ball, so that you can see the motion in both directions at the frame of impact. The thing is that this effect doesn't work by default in After Effects 6.5, because it averages the surround frames (Figure 2.30a). You have to trick it into making the V-shape of the ricochet by adding a very slight rotation to the ricochet layer. This forces After Effects into sampling all of the frames necessary to create the ricochet (Figure 2.30b).

Figure 2.30a and b. These two images should look the same, but you have to force After Effects into getting the result in 2.30b by adding a slight bit of rotation. The other key is setting the Shutter Phase to –50% of the Shutter Angle, which causes the motion blur to be sampled right at the middle of the motion.

The thing to understand about how After Effects thinks about time is that, internally, it thinks in absolute terms, in seconds (not in frames that change in absolute time according to how many of them there are per second). If it were instead calculating time as number of frames total, or frames per second, it would have difficulty if you were to change the frame rate, but actually, at the deepest level, it doesn't care what the frame rate is; After Effects lets you worry about that, while it continues to think in terms of total time measured in seconds.

What does that mean? In real terms, it means that you can change the frame rate of any comp on the fly and the keyframes and other timing information will stay the same in seconds: The timing will not change even though the frames play back at a faster (or slower) rate. The keyframes won't even be nudged to the nearest frame; they will rest between frames, if necessary, to hold the same position in seconds. You can't place them between frames manually, mind you; it only occurs in service to absolute time (Figure 2.31).

Figure 2.31. This composition was changed from 29.97 fps to 24 fps. All frames following the first one now fall in between whole frames, and After Effects averages the data accordingly.

The same goes for footage (or a composition) that comes into a composition with an alternate frame rate; After Effects simply looks for the current frame in absolute time, whether it started at the same exact moment in time or not.

Sometimes you need to change the duration and speed of an incoming clip or subcomp; perhaps you need it longer and slower to cover the duration of the shot, or maybe it must be shorter and run faster to complete sooner. Your solution in both situations is the Time Stretch feature, provided you don't need the frame rate to change in the course of the shot (in which case only Time Remapping will do).

There are several ways to activate stretch, depending on how you want to set it up. Click on that little double-sided arrow along the bottom, between the columns and the Zoom slider (Figure 2.32). This reveals four columns that show the four basic ways you can set Time Stretch for a layer. You can stretch by

Figure 2.32. The double arrow (highlighted) reveals and conceals the Time Stretch features.

Click on the Duration or Stretch value to bring up the Time Stretch dialog (Figure 2.33).

Figure 2.33. You can set an alternate Duration or Stretch Factor in the Time Stretch dialog. Here you can also specify the pivot point for the stretch by choosing whether the In point, Out point, or current time is held in place.

Note that stretching a layer does not cause any keyframes applied to that layer to stretch as well; you can highlight a set of keyframes and Alt/Option-drag on the keyframe at one end of the set to stretch keyframes as needed, or you can pre-compose the clip prior to time stretching, guaranteeing that all of the animation stays lined up.

Frame Blending

There is, of course, a downside to lengthening a source clip (and to a lesser extent, to shortening it); frames repeat or skip, and unless the stretch value is an even multiplication or a division of 100% (say, 50% or 200%), the repeating or skipping occurs in irregular increments, potentially causing a herky-jerky motion.

When Frame Blending Won't Do

When it's critical to retain as much detail and accuracy in your footage as possible but frame blending is causing its trademark blurriness, you may want to consider another plug-in: Twixtor from RE: Vision Effects.

This plug-in transcends frame blending, sampling adjacent frames to realistically synthesize new ones by warping and extrapolating frames from the original sequence. Rather than simply blending adjacent frames, Twixtor tracks and calculates motion for each individual pixel. For a demo version, go to www.revisionfx.com.

Enable Frame Blending for the layer and the composition, and After Effects averages the frames together. This works not only on layers with time stretching but with any footage that comes in at a frame rate other than that of the composition.

There's only one problem: To say that After Effects “averages” adjoining frames to create the Frame Blending effect is really too generous; it actually just overlays them, making them appear blurrier than at 100%. Although, at Best quality, After Effects averages as many as five adjacent frames in each direction, it's not enough to prevent a disconcerting variation between blurry and sharp images at irregular frame rates.

So, use this feature only in cases where it doesn't count for much; where the element itself will help conceal the trick. For example, use it for something already soft in appearance, such as smoke or clouds.

Time Reversal

You can reverse the timing of a layer with a simple shortcut: Highlight the layer, and press Ctrl+Alt+R (Cmd+Option+R) to set the Stretch value to –100%. The layer takes on a stripy appearance to remind you that its time has been set in reverse motion (Figure 2.34).

Figure 2.34. The candy-striping is a clear indication that the Stretch value is negative. Keyframes no longer align with the beginning of the frame, a potential gotcha if you're adding animation.

There are some bizarre side effects of doing this, however, which may lead you to prefer using Time Remapping to achieve this effect (more on this soon) or to drop the time-reversed clip into a composition of its own. The first thing you notice is that the layer jumps in time, so that the layer now begins where the last frame had been. Moreover, everything to do with that layer is reversed: the In and Out shortcuts are inverted, keyframes appear at the end of each frame, making them look offset from the time needle, and moving forward in time in the Layer window moves you backward in the composition.

Time Stretching an Animated Composition

If you apply Time Stretch (or Time Remapping) to a composition, After Effects, by default, goes back to that composition's animation and recalculates its frame rate, slicing any keyframe animations to fit the new rate instead of adhering to the old one.

Be aware, therefore, that the frame values on any given frame will be different unless the underlying time values match. After Effects does not round off to the nearest whole frame; again, it looks at where the animation is in absolute time, even if that means it's between frames in the subcomposition.

You can override this behavior and force After Effects to round off, however. On the Advanced tab of the Composition Settings window of the nested comp, toggle on Preserve Frame Rate When Nested or in Render Queue (Figure 2.35) to force After Effects to look at only whole frame increments in the underlying composition. This would be a way to show footage at a film frame rate (24 fps) in a composition designed for NTSC television (29.97 fps). You may never need this feature, but knowing it's there and how After Effects otherwise handles pre-composed animations is helpful.

Figure 2.35. If you need a subcomposition to run at a different rate than the master composition, checking the highlighted box in the subcomp will prevent After Effects from resampling keyframes at the master comp's rate

You can do away with the Time Stretch feature set altogether if you are comfortable working with Time Remapping instead. This feature may require a bit of work to wrap your head around, but once you understand it, it has fewer pitfalls than with Time Stretch.

The philosophy behind Time Remapping is elusively simple: Time has a value, just like any other property, which means you can keyframe it, ease in and out of it, loop and ping-pong it, and do all of the other things you can do with any other keyframe data.

Figures 2.36a, b, and c show sample timelines that contain typical uses for Time Remapping in an effects situation. The presumption is that you're not going completely nuts rolling footage back and forth, ramping the frame rate up and down, although you can; I'm just considering the deep exploration of this ability to be off-limits for realistic effects work.

Figure 2.36a, b, and c. Simple, everyday things you can do with Time Remapping: speed up the source with eases (shown in 2.36a as Value and Velocity curves), progress through a series of stills selected from the moving clip using Hold keyframes (2.36b), and loop the source using a simple expression (2.36c).

Note

Beware that if you apply Time Remapping to a source layer that is longer than the composition in which you're applying it, there are two keyframes assigned every time you activate this feature, and the second one hides somewhere off the end of the timeline. Set a new keyframe at the (visible) end of the composition, click the Time Remap property title to highlight all three keyframes, Shift-click the first two to deselect them, and then press Delete to get rid of that hidden final one.

You can set Time Remapping by selecting it under the Layer menu or using the shortcut (Ctrl+Alt+T/Cmd+Option+T). This sets and reveals two keyframes at the beginning and end of the layer (not the In and Out points, if they've been edited).

If you've set different In or Out points from the default to the layer being time remapped, then a logical first step is to set Time Remapping keyframes at the In and Out points, then delete the default ones.

Time remapped layers have a potentially infinite duration. When you apply time remapping to a layer whose duration is shorter than that of the composition, the layer automatically becomes the duration of the comp, filling the extra frames by holding on the final time remapped keyframe value.

Still from a Moving Image

Believe it or not, the simplest way to create a still in After Effects from a moving image clip is to set Time Remapping to that clip with just a single keyframe (Figure 2.37). The value of that keyframe is the time at which the desired frame appears in the source clip. It doesn't even matter where on the timeline the one keyframe appears, as long as it is the only one applied.

Figure 2.37. This is what a still frame taken from the first frame of a moving clip looks like; You know it's the first frame of the source not because that's where the keyframe lies (it could be anywhere on the layer, because there's just one), but because the value is set to 00000.

To reverse a clip using Time Remapping, highlight all keyframes by clicking on the Time Remap property name. Next, context-click one of the highlighted keyframes, and select Keyframe Assistant > Time-Reverse Keyframes.

Time remapping can take awhile to get used to in complex situations, but because this book is concerned more with simulating realism than with stylized artistry, I'll use Time Remapping mostly for eases, holds, loops (discussed in Chapter 10), and the occasional time reversal. If you find yourself in deeper, the same principles apply; you just have to track them more closely.