Of course, Vuescan adjusts the colors, exposure, contrast, levels, etc when it scans so it's hard to know where the film lets off and the scanner/software takes over. Neither may actually be representative of the originals.
Hi Alan, the above is not strictly correct. Vuescan can be used to produce unadulterated 48bit raw files with no adjustments applied. That's what I do with my negatives. I then let an external program deal with the inversion and removal of the orange mask, which is different based on film maker/stock. The tool I use for this is called Colorperfect (there are other ones though). Colorperfect has inversion profiles for most popular negative films, including Ektar 100, which is what I used above. I would hazard that, provided the inversion profiles closely match the film stocks, a Colorperfect inversion will be a pretty good representation of the original. Of course, any copy of a negative is an interpretation: this includes an imacon scan, a drum scan, a print.
You're right in that Vuescan might apply liberal colour interpretation to the Provia positive. Apart from stretching the histogram all the way through (thus deactivating Vuescan's levels correction) I haven't played with this because as I said I like to keep my post-processing to a minimum - I specifically moved back to film photography after years of digital to spend as little as possible in front of a computer
Another degree of freedom will come of course from the scanner itself - can't exclude that either.
Personally I see a noticeable difference in the two shots above. In general, with my (amateur-level) workflow and equipment, and after doing a fair number of tests, I find Ektar is better in colour, sharpness and dynamic range when scanned on the V550. Given that a brick of Ektar 100 in 120 costs £25 in the UK, versus the £42 for a brick of Provia 100F, for me it's a no brainer, really.
I'm sure Provia is a good film in general though, and might be great for others - I've seen fantastic results from 5x4 Provia drum scans for instance.