What actually changes.
JPG saves space by combining image information. Small blocks and bright halos can appear around sharp black and white edges. They are easy to miss in a photograph but each transition may become part of a traced contour.
PNG stores hard edges without lossy compression and can include an alpha channel. That is ideal for logos, lettering and isolated artwork. The real source still matters. A blurred PNG does not become precise because of its file extension.
My quick decision
- Use PNG when transparency or a hard graphic edge is available
- Use JPG when it is the only photograph and the resolution is sufficient
- Avoid saving repeatedly as JPG because every pass adds artefacts
- Inspect Original and Result closely with either format
Step by step.
- Load the original PNG when it is available.
- For JPG, set Black level first and watch the halos around edges.
- Increase Contrast only until the intended shape separates clearly.
- Check Fine mode to see whether the contour becomes calmer.
- Use Bézier simplification carefully and inspect small corners.
- Download SVG once counters and the outside contour look right.
A better file format cannot replace a good capture. Perspective, motion blur and reflections are already part of the image and need to be solved before or during separation.
Questions before export.
Does a JPG become a genuine SVG?
Yes. Vector Boy does not wrap the JPG inside an SVG. It traces the detected form as closed paths.
Is PNG always sharper?
Not automatically. A small or upscaled PNG may contain less real detail than a good JPG.
What happens to a transparent background?
The alpha channel separates visible artwork from transparency. No white rectangle is added as a vector object.