Convert VSI to GBR

Technical note

How VSI and GBR differ in practice

VSI is made of pixels and GBR stores geometry, so the conversion traces the image into real lines, arcs and polylines. The result is editable in CAD: entities can be selected, snapped to and measured, rather than a bitmap wrapped inside a GBR container, which is what a viewer or a lightweight converter usually produces. It is still derived geometry, so expect contours rather than a structured drawing. Dimensions, blocks and meaningful layer names were never in the VSI to begin with. A line in a VSI image is several pixels wide, so tracing it can mean different things and the mode is yours to choose. Centreline returns a single line running down the middle of each stroke, which is what a scanned drawing needs: one entity per line, snappable and dimensionable, rather than a long thin rectangle around it. Outline follows both edges and returns closed shapes, which suits clipart, logos and anything that will be filled or cut. A precision setting controls how closely the traced paths follow the pixel edges. One thing to plan for: vectorised output from this path is black and white, so colour in the VSI does not survive as colour in the GBR.