View Single Post
Old 08-07-2008, 02:03 PM   #16
Ben_Edict
Loyal Member
 
Join Date: Jul 2007
Location: Germany
Gallery Photos: 0
Posts: 517
JIm, as you have read from the many replies there are several options for you:
- get the film scanned in the 1-hour-lab. These scans are usually very low res, certainly nt good enough for 11x14 prints, you want to make
- get the film scanned professionally: this is the route to go, for getting higfh res scans. It is going to be expensive if you scan loads of images
- scan yourself. If you want a 300 dpi print, you need to scan your images with effectively 2800 dpi. That is the realm of the dedicated, good quality film scanner, like a Nikon V or the like (well, not much choice there). Even the best affordable flatbeds like the Epson V700/750 (I have one myself) won't provide more than app. 2300 dpi real resolution (the claimed 6000 dpi are a purely theoretical thing - but this is the same with all other flatbeds, be it Epson, Canon, Mikrotek, HP etc.)
Nevertheless, this still should suffice for a good 11x14 print, because 300 dpoi is not always necessary. 240 dpi is in most cases good enough (subject dependend). If you go for a flatbed, be careful to buy a modell with Silverfast scan software (much cheaper to go with the scanner, than to buy it separately).

Silverfast can improve the scan quality immensely, because it can make multi-exposure scans, which allow to scan much higher densities /better shadows and highlights). Wheras a simple Epson V700/750 has a density range of something like 3.1 and goes up to 3.36. Nikon film scanners get even a bigger boost in desity range.

Up to a certain degree you can have that with Vuescan, too - another independend scanner software package. Both have trial versions, so you can explore, which one you prefer. Silverfast is much more expensive in the end, but gives better control, if you know what you do, wheras Vuescan is probably more user friendly.

The much bigger problem, than pure resolution (or pixels in the final scan file) is colour depth. All 1-hour-lab scans will be 8 bit and thus they won't use the information contained in the image, but lose most of the information.

Pro-scans and your own home-made scans can be (and should be) something between 12-16 bit (these all end up in a 16-bit TIF file, as TIFs don't come in 12 bit or 14 bit flavour). This is considerably more infomration than an 8 bit file. Compare it to making RAWs versus JPGs with a digital camera.

Then you asked whether to use slide or neg films. I think, today it is widely accepted, that negs provide a better base for scanning, as they simply have a far bigger contrast ratio. A negative can contain 12 f-stops of useable information, wheras a slide typically will offer 9 f-stops as the max. That is a huge difference.

When scanning from prints, this is even wortse, as a print does not even transport the
8 or 9 f-stops of the slide film. Its contrast ratio is something like 1:256, versus 1:1000 for the slide and even more for the neg.

If you want to scan prints, this print must already be optimized to contain all the information on the neg/slide. That means it must be made to highest quality standards with contrast reduction (dodging, burning, split exposure etc.) to allow to even apporach, what is reall "in" the neg/slide.


Oops, that's a bit long, now. So I stop here.

Ben
__________________
Pentax user for 25+ years; now K10D, K20D, LX, MX, PZ1p..., 40 lenses, Mamiya 645 system, 4x5 Large Format gear, analogue + digital darkroom
Ben_Edict is offline   Reply With Quote