I’ll think it through. Though from what I understand, this feature isn’t particularly successful in other apps (i.e., based on feedback I heard from the NLP roll analysis).
There’s an inherent inconsistency in the frames of scanned film, which makes applying a single transform across the roll challenging. For example, even if your exposure is perfect, the density and colour casts will still vary depending on the type of images you create (e.g., high-key, low-key, different light sources, etc.).
Another challenge is the automatic calibration of scanners. At the very least, most hardware will automatically adjust its own exposure, which will change depending on the overall frame density and other factors. This can make every frame appear different, even if the same transformation is applied across the entire thing.
I am, however, looking into ways to save more time scanning this year, so I will definitely keep this idea in mind and research it further to see if there’s anything I can do to add consistency while being transparent about how it all works :)
Dmitri