Rendering
FCE has made dramatic improvements in real-time effects performance, but when you finally run out of processing power, you must render. This section outlines FCE's playback and rendering quality settings and rendering protocols and explains how to navigate the Sequence menu's vast forest of Render commands. For more information, see Chapter 34 of Apple's Final Cut Express Help PDF.
Rendering protocols
Here are some general guidelines to help you understand Final Cut Express's rendering protocols. What needs to be rendered In general, you must render any transitions, effects, and composited layers that exceed the real-time capacity of your FCE system. Also, before your sequence can be exported or printed to video, you have to render any source media with frame rate, frame size, or video or audio compression settings that differ from those settings in your sequence. Audio tracks with transitions or filters, as well as multiple audio tracks over your real-time playback limit, have to be rendered before playback. Clips whose speed has been changed must also be rendered before playback. What doesn't need to be rendered Final Cut Express sequences that include real-time-supported transitions, effects, and composited layers can be played back in real time without rendering, as long as the sequence size and frame rate match those of the original source material. Multiple audio tracks can be played back in real time without rendering. (See Table 18.1, "Final Cut Express Audio Track Costs," later in this chapter for more information.)
Table 18.1. Final Cut Express Audio Track Costs ITEM | TYPE | TRACK COST |
|---|
Each mono track | Track | 1 | Each mono track with transitions applied | Transition | 2 | Pair of stereo tracks | Track | 2 | Sub-sequence containing track | Track | 4 (one stereo) | Each track referencing a sub-sequence | Track | Number of audio tracks in sub-sequence | Compressor/limiter | Filter | 6 | Expander/gate | Filter | 6 | Vocal de-esser | Filter | 6 | Vocal de-popper | Filter | 6 | Reverb/echo | Filter | 6 | All other filters | Filter | 3 |
Rendering order Video frames in sequences are rendered in the following order: Effects applied within individual tracks are processed first. Speed changes, filters, motion, motion blur, opacity, and transitions are processed next, in that order. After processing of all the individual tracks is complete, the processed tracks are composited, starting with the top layer of video (the highest-numbered track), which is composited onto the track below it.
It is possible to change the order of rendering by using nested sequences. For more information, see "Working with Multiple Sequences" in Chapter 4.
Rendering indicators
As you build your sequence, you'll encounter the following indicators: When you try to play video material that requires rendering in the Viewer or the Canvas, a blue background with the word "Unrendered" appears (Figure 18.6 ), indicating that the video can't play in real time.
Figure 18.6. The "Unrendered" screen appears when you try to play back video that cannot be played back in real time without rendering.
When you try to play audio material that requires rendering in the Viewer or the Canvas, you'll hear a steady beeping sound, indicating that the audio can't play in real time. In the Timeline, the rendering indicator above the ruler (Figure 18.7 ) consists of two thin bars. These bars indicate which sections of the sequence currently require rendering to play back smoothly. The upper bar indicates the render status of video tracks; the lower bar shows the audio track status.
Figure 18.7. The rendering status bars are located above the Timeline ruler. The upper bar indicates the render status of video tracks. The lower bar shows the audio track status. Pause the pointer over the rendering indicator bars, and a tooltip displays render status details.
A sequence audio clip requiring sample-rate conversion displays a render status indicator overlay right on the clip in the Timeline (Figure 18.8 ).
Figure 18.8. An audio clip requiring sample-rate conversion displays an item-level render status indicator overlay right on the clip in the Timeline.
Non-real-time FCE systems use the following status codes in the rendering indicators: Red (Needs Render): The material needs to be rendered. Gray: No rendering is required. Blue-gray: The material has already been rendered.
Real-time effects previewing adds six (!) more color codes to the render status indicator: Dark green (Full): The material can be played back and output at full quality in real time. No rendering is required. Green (Preview): No rendering is required for real-time playback in the Canvas or the Viewer. For playback on an external device, green indicates a lower-quality output of scaling and motion effects. Yellow (Proxy): Real-time playback is an approximation of the final effect. To get the true final effect, you still need to render the material. You'll see the final effect when playback is stopped or during scrubbing. Dark yellow: The material has been rendered at a quality lower than that currently specified on your sequence's Render Control tab. FCE preserves these render files even if you modify the sequence's render quality settings later. Orange (Unlimited): Indicates effects enabled by selecting Unlimited RT in the Timeline's RT pop-up menu. These effects exceed your computer's "official" real-time playback capabilities and may drop frames during playback, but you can get a rough idea of how a complex effect is shaping up.
FCE Protocol: Audio Rendering Options
Here's the rundown on the audio rendering options available in the Render submenus: For Playback: Only the portions of your sequence audio tracks that require rendering to play back in real-time are processed. Item Level: This option renders individual audio clips that require sample-rate conversion to match the sequence sample rate, and clips with applied audio filters. Item-level processing generates item-level render cache files. Audio levels and multitrack playback are still processed in real time. Mixdown Audio: Your audio tracks, along with all applied filters, levels, and transitions, are processed into a set of render files, a single file for each audio output. This command is available only in the Render Only submenu.
|
|