Repository navigation
transform scale rotate
Home > Transform > Scale & Rotate
< Previous: Rotate | Next: Mirror >
Scale & Rotate (tool)Scale and rotate at once so a reference point lands on the handle.
Menu: Transform > Scale & Rotate | In depth: Transform tools
One gesture that scales and turns at the same time, so that a point you nominate lands exactly where you want it.
Pick the center, pick a reference point on the selection, then drag the handle. The selection is scaled by how much farther the handle is from the center than the reference was, and turned by the angle between them. The reference point comes to rest on the handle.
Because "put that corner there" is one intention, and doing it as a scale and then a rotate is two guesses that have to agree.
The classic case is fitting a motif into a space it was not drawn for: its corner has to reach a particular crossing, and it has to stay the shape it is. Scale it to the right size and the corner points the wrong way; rotate it to the right angle and the corner is short. This tool solves both at once by definition, because scale and angle are read from the same pair of positions.
What it performs is a similarity: every angle inside the selection is preserved and every length changes by the same factor. Squares stay square, tangents stay tangent, and a pattern stays the pattern it was.
Like Rotate and Scale, it moves hand-placed free points and lets everything derived follow. Select Roots first if a selection seems immovable; it will be made of computed points.
Each click takes a point, a crossing, or a position along a curve, which is what makes this the tool for "land that corner on that crossing". Both the reference and the handle can be things the drawing already contains, so the fit is exact rather than eyeballed.
Empty space is allowed and drops a loose handle. A missed click does not cancel; Escape does.
You can also type a factor and an angle if you know them, which turns the same tool into an exact similarity rather than a dragged one.
The reference point wants to be something you can see clearly and would notice landing in the wrong place. A corner, an endpoint, a crossing. Pick something vague and the transform will be correct and still look wrong, because the part you were watching is not the part you told it to land.