Mixed-format timelines have items that are in a format different from the timeline format. Common examples would be SD footage in an HD sequence, or a still image in a sequence.
The RT engine and the Non-RT engine handle mixed formats differently. When an effect is applied to a mixed format timeline, its render method is called one way for RT scrubbing and playback and another way for a movie. In the RT rendering engine, all items are scaled to sequence size before the effect is applied. In Non-RT rendering, the effect is applied to the item at its native resolution.
You can use the renderInfo.scale value to detect when an item is being scaled for RT display. However, this value is not set in the way you might expect.
Consider a 720p item in a DV NTSC sequence. In RT, the following information is sent:
| |
|
|
|
|
| |
|
|
|
|
| |
|
|
|
|
When the sequence is rendered, the following information is sent:
| |
|
|
|
|
| |
|
|
|
|
| |
|
|
|
|
Note that in RT, the HD image is scaled to fit into a DV frame. The width / height aspect is maintained, so the (single-field) image is scaled to 720 x 202; this is to fit inside the 720 x 240 DV field.
In Non-RT rendering, the HD image is at full resolution. The scale value is set to a number greater than 1.0. This tells the effect that the image will be scaled down after render to fit into the sequence.
Last updated: 2007-05-21