Footage and Composition Settings

Why devote a section to footage and composition settings if After Effects does its best to automate them and stay out of your way? Because by trying to make your life easier, After Effects is inadvertently making it more difficult: Shielding you from the details contributes to user ignorance of them. And you need to understand these details, so you can get them right from the start. A misinterpreted alpha channel, nonsquare pixel footage interpreted as square, undetected scan lines, and similar issues will leave you fighting an uphill battle if you don't make the proper settings before you ever start putting together a composition.

Although the automatic settings in After Effect are great most of the time, it pays to understand how they work for those times you need to do some manual tweaking. Take a closer look.

Interpreting Footage

This book generally eschews the practice of taking you menu by menu through After Effects. Sometimes, however, the UI perfectly encapsulates a set of production challenges. The Interpret Footage dialog, then, is like a section-by-section checklist of all that can go wrong when you import footage: misinterpreted Alpha, incorrect Frame Rate, misadjusted Fields and Pulldown settings, incorrect Pixel Aspect Ratio. To bring up the Interpret Footage dialog, select some footage in the Project window and press Ctrl+F/Cmd+F or context-click and select Interpret Footage > Main.

Alpha

Alpha settings are more complicated than you might think when it comes to compositing; most After Effects users have no idea just how complicated, in fact. Figure 1.19 shows the common symptoms of a misinterpreted alpha channel.

Figure 1.19. It's easy to distinguish a good (right) from a bad alpha channel (left). Note the grey area in the cockpit canopy and the black line around the misinterpreted version (left). In this case, the image had a premultiplied alpha, and the black data in the edges is the result of misinterpreting it as straight.

For the time being, I offer a couple pointers and reminders:

• If you're unclear about which type of alpha channel you're working with, click Guess in the Interpretation dialog that shows up when you import footage with alpha. This will typically get the setting right.

• Under Preferences > Import is a setting for how you want After Effects to handle footage that comes in with an alpha channel. Beware of setting this to anything besides Ask User until you are certain you know what you are doing with alpha channels and that circumstances aren't likely to change without you noticing.

For a more in-depth discussion of alpha channels and how they operate in the After Effects pipeline, see Chapter 3, "Selections: The Key to Compositing."

Frame Rate

The clearest symptom of an incorrectly set frame rate is footage that does not play smoothly in a composition with the target frame rate. Interpreting frame rate properly is an issue only when you import image sequences, which, of course, is the most common way that you will import moving footage in a visual effects setting. Image sequences are preferred to formats such as QuickTime (.mov) or Windows Media (.avi) in a production setting because

• If something goes wrong with a rendered image sequence, only the missing frames need to be replaced. With a movie file, the whole file typically has to be discarded and replaced, potentially costing hours of render time.

• Movie file formats are not as universally recognized and standardized across various platforms and programs as still image formats. QuickTime is the most robust and flexible of the moving image formats, yet many Windows-based programs do not even natively support it.

Therefore, when importing still image sequences to use as moving footage, remember

• Just because you've set your Project Settings to the proper frame rate (say, 24 fps for a feature film project), your image sequences may still import at 30 fps by default. You can change this default setting under Preferences > Import.

• You can assign whatever frame rate you like to a footage sequence if you determine the default setting is not correct.

• Just because an imported moving image file (such as QuickTime) has its own frame rate does not mean you cannot override this rate as needed by checking Assume This Frame Rate and entering a value. Be certain, however, that this is really what you want to do; typically, it is only a correction for outputs that were set incorrectly in another application.

Keep in mind that you can highlight any clip in your Project window and see its current frame rate, along with other default settings, displayed at the top of the window (Figure 1.20).

Figure 1.20. To find useful information about any file in the Project window, highlight the file and look at the top of the window. You'll find frame rate, duration, color depth, pixel aspect, and how many times the clip is used in the project. To see and select specific comps in which it is used, click the little arrow at the end of the file name.

Fields, Pulldown, PAR

As creators of digital movie clips, we all look forward to the day when interlaced footage is a thing of the past. In the meantime, you must guard against symptoms such as those depicted in Figure 1.21. The best way to avoid field artifacts in moving footage is by making sure your Field setting matches that of your incoming footage; Separate Fields unweaves the two fields embedded in a single frame to be two separate frames, allowing you to transform them without creating a complete mess.

Figure 1.21. The foreground pickup truck spells trouble if you're planning on doing much more than a simple color correction; fields were not removed for this clip. If you see a problem like this, check your footage interpret settings immediately.

For DV footage, this process is automated and After Effects knows what to do by default. For other formats, you need to know not only whether your footage is coming in with fields but also whether the upper or lower field is first. Furthermore, with film footage you may have to deal with pulldown.

Digital Source Formats

After Effects is capable of importing and exporting a wide array of footage formats, yet only a small subset of these recur typically in visual effects production. Here are some of the most popular raster image formats and their advantages (for a summary, see Table 1.1):

• TIFF (.tif): In many ways the most flexible of formats, with the possible exception of PNG, TIFF is longstanding enough to be supported by most applications. It can be compressed effectively (and losslessly) by enabling LZW compression, which looks for recurring instances and patterns in the image and encodes them. Although 16-bit support is not native to After Effects, it is available via a free third-party plug-in (for more on this see Chapter 11).

• Targa (.tga): This format has the dual advantages of being universal to most computer graphics applications and offering lossless RLE (run length encoding) compression, which looks for sets of identical pixels that can be losslessly grouped together (such as a solid color in the background or alpha channel).

• PNG (.png): In many ways the most misunderstood of formats, PNG natively supports images of 8 or 16 bits per channel in Photoshop, and it often does the most effective job of losslessly keeping files sizes small. Because PNG has a reputation as a Web-oriented format, users sometimes mistakenly assume that it adds lossy compression, as is the case with JPEG files.

• Cineon (.cin): A common format for transferring digitized film images, this format works in nonlinear, 10-bit color and is discussed in greater detail in Chapter 11.

• Photoshop (.psd): Although a universally supported format with many built-in extras, such as support for individual layer transparency, these files can be huge, as Photoshop offers no lossless compression options.

• Open EXR (.exr): This is an advanced format for use solely with high dynamic range imaging (HDRI), source that has been taken in bit depths higher than linear 16 bit. Moreover, it is not supported natively by After Effects but requires a special plug-in. If this all sounds like gobbledygook but you still want to learn more, have a look at Chapter 11.

Table 1.1. Raster Image Formats and Their Advantages

FORMAT	BIT DEPTH	LOSSLESS COMPRESSION	ALPHA CHANNEL
TIFF	8 or 16 bit	Y	Y
TGA	8 bit	Y	Y
PNG	8 or 16 bit	Y	Y
CIN	10 bit
PSD	8 or 16 bit		Y (multiple via layers)
EXR (non-native)	16 bit, 32 bit (floating point)	Y

So which format should you use? I would give TIFF the edge for 8-bit images, but use PNG for 16-bit images, because the file sizes are losslessly so much smaller. Cineon is the only choice for 10-bit log files, and there's nothing particularly wrong with Targa (except perhaps slow run length encoding) or Photoshop (except the huge file sizes).

Be sure to note that 8 bit means 8 bits per channel, for a total 24 bits in an RGB image or 32 bits in an RGBA image. Targa includes a 16-bit option that actually means 16 bits total, or 5 per channel plus 1. Avoid this.

Adobe Formats

After Effects supports some special features for dealing with files created by other Adobe applications. For effects work, the most handy one is support for multilayer Photoshop images, including layer names, transfer modes, and transparency settings.

To take advantage of this feature, import the Photoshop source as a composition, which leaves all of the properties of each individual layer editable in After Effects. The alternative is to import only a single layer or to flatten all layers and import the entire file as one image.

The ability to import a Photoshop file as a composition means you can set up a shot as a Photoshop still and import it with everything already in place. This is particularly useful with matte paintings that include separate elements, such as multiple planes of depth.

If you don't need the elements ready to go as layers of a composition, however, I would forego this option because it makes editing and updating the source trickier. Each layer becomes a separate element that must be updated separately, and it's wasted effort if you're never editing the associated composition.

Composition Settings

There are two relatively sure-fire methods for ensuring that your composition settings are exactly as they should be:

• Use a prebuilt project template that includes compositions whose settings are already correct.

• Create new compositions by dragging a clip to the Create a New Composition icon and using only clips whose settings match the target output size and frame rate (Figure 1.22).

  Figure 1.22. Dragging a source background clip to the highlighted icon at the bottom of the Project window creates a new composition with the clip's duration, pixel dimensions, pixel aspect, and frame rate. It's a reasonably foolproof way to set up a new composition if you are working with a master background clip.

  

  
  

The crucial settings to get correct in the Composition Settings window are the pixel dimensions, Pixel Aspect Ratio, Frame Rate, and Duration. (Actually, even Duration is negotiable so long as it is not too short.) If you're working with a footage format that isn't accurately described in any of the Preset options and you're going to be using this format again and again, then by all means create your own Preset setting by clicking on the small icon adjacent to the Preset pull-down menu, the icon that looks like a little floppy disk (Figure 1.23).

Figure 1.23. Feel free to create your own preset if none of the listed ones match one that you will be using continually for your project.

And what of that other tab, labeled Advanced in the Composition Settings dialog? It pertains to specific options for dealing with time and space (see Chapter 4) and for working with motion blur and 3D (see Chapter 9, "Virtual Cinematography").