Sparrow said:
Two years ago I came to the conclusion that to preserve it’s RF nature in use, what I needed was a camera with a manual interface (focus and exposure), a full frame 135 sensor to keep the FOV and to allow for 12x18” prints at 400dpi some 30-35mp of file size.
I still hold that view. Am I being unreasonable?
That depends on what you consider "reasonable." If you mean "practical," well -- SLR cameras sell in much higher volumes than RF cameras, so manufacturers tend to concentrate their R&D budgets in the SLR sector. As a result, innovations tend to occur first in the DSLR sector.
So, how many DSLR cameras are available right now that incorporate all the features you want? Um, exactly two, they're both from the same manufacturer (Canon) and they're both at the top of their respective price ranges. (Even these don't have the 35-mp pixel count you've convinced yourself is necessary, so let's leave that for the second part of this answer.)
Back to the two Canons: Although they have good reputations, and photographers seem to aspire to them, that hasn't been enough to induce any other manufacturer to take the same route. And although online forums suggest that many photographers want them, the fact is that the majority wind up buying something else, either another Canon model or another manufacturer's camera.
RF cameras are always more expensive than their SLR equivalents because of the complexity of an optical rangefinder and the greater precision required to assemble them, plus the fact that since they sell in smaller numbers, the manufacturer can't expect to "make it up on volume." So, my guess is that the type of camera you describe would cost something just under $10,000 (the cost of a top-of-the-line Canon EOS plus a "rangefinder premium" to cover the cost of the RF subsystem, R&D costs, and lower volume.)
Is that a
reasonable expectation? Well, Canon has good business judgment and plenty of development funds, and they're in business to make money; I suspect that if they felt such a camera would sell, they would be making it already.
Am I misunderstanding the maths? Can 10 or 12mp really be as good as a 40mp scan from a negative?
You're not so much misunderstanding the maths as trying to compare apples and oranges. One of the key reasons that negatives need very high scan resolutions is to avoid the effects of
grain aliasing, the interaction between film-grain edges and scanner-pixel edges that causes a scan to look much coarser-grained than a wet print. (This probably accounts for most of what your kids are seeing when they distinguish between the higher- and lower-resolution scans.)
A direct digital image doesn't have this problem, and also doesn't incur further "generational losses" from the scanner's film holder, cover glass, mirror, lens, and other analog elements that are involved in the transfer of image data from the negative to the digital file.
In other words, all other things being equal, a digital file made directly from the image cast by the camera lens is much more faithful to the original scene than a file that has had to go through the long chain of camera lens > film > developing process > scanning process. An image scanned from a negative requires much higher scan resolution to compensate for this loss in fidelity.
On the output side, the extra pixels of a film scan don't always translate into better
printed results because there's no common printing process that reproduces your digital file pixel-for-pixel. Inkjet printers form their image by making up overlapping patterns of ink droplets that blur the pixel boundaries; dye-sublimation printers use a gaseous diffusion process that does the same; and so does printing-press reproduction, in which the image is formed of overlapping screen patterns called rosettes.
The effect of all this -- plus the fact that you're evaluating the output with a wet, jiggly, imprecise little blob of jelly called the human eyeball -- is that there's not much demonstrable benefit in sending your printing device more than 300
output pixels per inch, and most people are quite satisfied with the overall appearance of prints made with 200 output pixels per inch, or even less for larger prints that are seen from long distances. (I can tell you from experience that a 6-megapixel digicam image makes a perfectly acceptable billboard.)
So, if you're routinely making 20 x 30 prints which people are going to examine at close reading distance -- in-store display posters, for example -- then yes, you are going to need a lot more than 10 megapixels, which is why commercial photographers are willing to spend $30,000-plus for those 39-megapixel digital backs.
But the vast majority of us seldom make prints that large -- and for smaller prints, a camera with a lower pixel count produces just as good a final image, because any "extra" pixels are simply thrown away by the printing process.
So, a very common compromise is to buy and use a digital camera that you can afford and that generates enough image data to meet
most of your needs. For the occasional huge, highly-detailed print, it's more cost-effective to keep a medium- or large-format film camera around than to buy a digital back that has capabilities you'll hardly ever use (and is much bulkier and less convenient to use than a more compact digital camera.)
Even the most determined film zealots don't use 8 x 10 sheet film for
everything -- that's why there have always been so many camera sizes and film formats.
It's worth reminding ourselves occasionally that
image quality isn't the same as
picture quality, and that a picture you got with your small, fast-operating camera is always going to be better than the one you missed while setting up your big, cumbersome large-format camera. The same goes for digital as well as film.
