Modifying frame boundary detection?

[mabunday]

Just stumbled upon this project and curious if anyone has tried modifying the frame boundary detection algorithm as that seems like a major weak point of the NikonScan software.

I frequently notice that the thumbnail previews are able to find the correct film boundaries when doing full roll scans, but struggle when refocusing for preview or full scans of individual frames. Seems like there's a difference/flaw in how the thumbnails are positioned vs the preview/scans and I'm wondering if anyone else has tried to fix this in the software...

[Archilokos]

This could be especially useful when scanning negatives as positives to convert them in inversion applications nike Negative Lab Pro or ColorPerfect.

For whatever reasons, the frame boundaries detection works pretty well when scanning negatives as negatives, but if you switch to positives after the preview, it resets the detection and starts all over again, messing up the detection. You have to play with the offset tool for a while to set the 36 pictures. A pain.

I am not sure something can be done at this point. The Nikon Scan source code is probably on some backup CDs somewhere in a Nikon office drawer. Or lost.