From RMIT Visual Effects
Jump to: navigation, search

The Reformat node simply reformats to a target format any source that is fed into it. The top parameters ('type' and 'output format') control the size of the target format. The drop-down list of output formats contains a list of all common formats that Nuke recognizes. At the bottom of this list are automatically included any non-standard formats used within the script. If you wish to reformat to a value not on this list, then the 'type' value may be changed to 'box', which allows any dimensions to be specified.

The top parameters of the Reformat node.

The bottom parameters control how the reformat is committed. This is needed when the target format is not the same proportion and the source format. In such a case you will have to decide whether to preserve the width of the source, or its height. The 'resize type' parameter controls this behavior. The other key parameter is black outside, which controls what happens to any 'void' space within the image.

The bottom parameters of the Reformat node.

In the third image below, this void space is filled with stretched pixels. In the forth image, black outside fills this with black. Remember: the black outside behavior extends to the alpha.

From left:the original image, reformat with height preserved (clipped on the left and right), reformat with width preserved (with streaky pixels filling the 'void' and reformat with width preserved and black outside.

This node can also be used to 'flip' or 'flop' the image. This is actually rather useful for checking the 'visual togetherness' of a composite (painters traditionally do a similar thing when they check their painting in the mirror).

It can be processor intensive to reformat a read, especially an image sequence. You should consider rendering out the reformat material and reading it back in.

The Reformat should not be used to fix problems caused by setting the format wrong in Project Settings.