Just to be clear(as I'm still learning), when you say I would need to vector trace the hairstyle myself, you mean I would need to trace out a vector image of the hairstyle?
For a static SWF hairstyle:
yes, that would be correct. Your SVG file will typically contain a few reference or temporary layers, and then three real vector layers, corresponding to the three hair layers normally used by SDT. These layers are equivalent to the three active quadrants of a PNG hairstyle image. If your design includes a hat then we can tack on an extra layer in the SVG.
Here's one layer of the vectors for the
Princess Rosalina Static Hairstyle (specifically - the layer which is drawn ON TOP OF her ear).
----------------
For a dynamic SWF hairstyle:
you need to trace out each strand of hair. Including the hidden bits. You need to think about
how different parts of the hair pass over and under other parts, and then draw each strand on a separate layer of your SVG. Which means that you need to "invent" some parts, because
they simply aren't visible in your reference image. If you're lucky then you might be able to trace them from other images. If not, then you'll simply need to imagine them.
You can be a bit lazy with the details, because those hidden bits will tend to
remain hidden during SDT gameplay.
Here's that same set of vectors for the
Princess Rosalina Dynamic Hairstyle. It now consists of three separate layers.