regarding my previous mail for this topic, I have in the meantime identified my misconception. actually, `computeContour3d' returns the vertices just fine in the correct coordinate frame. the misconception was caused basically by assuming that the `level' argument was a fractional threshold relative to the maximum of the array. so I believed that the rendered cube actually is the "outer surface" of the defined object in the example provided in the manpage. I know understandt it's an absolute level and `example(computeContour3d)' consequently displays some "interior" isocontour. this explains all my apparent errors. I believe the manpage would benefit from a slight clarification that `level' actually is an absolute, not a relative/fractional threshold. apologies for the noise. j. ps: it of course would still be nice, if the surface area (or a vector containing the individual triangle areas) were returned to the caller as well ...
how to derive true surface area from `computeContour3d' (misc3d package) -- follow up
1 message · j. van den hoff