InterfaceTransformOptions

Whether to make the image as large as the transformed one, which then fills it (a rotated image grows to fit its corners, a skewed one gets wider). Without it, the image keeps its size and the parts outside are lost. Frames only grow up and left as far as their offset allows. Default: true

How the pixels between the source's are sampled: 'nearest-neighbor' (hard pixels, no smoothing, for pixel art), 'bilinear' (smooth), 'bicubic' (sharper, slower, can overshoot a little at hard edges). Default: 'bilinear'

The point the transform is applied around, in pixels: a rotation turns around it, a scale grows from it. Default: the middle of the image. It only decides where the result lies on a canvas that stays as large as the image (expand: false)