[plug] scanners

Craig Ringer craig at postnewspapers.com.au
Tue Oct 5 13:12:04 WST 2004


Ben Jensz wrote:
> The other choice you've got is that you can use Vuescan 
> (http://www.hamrick.com/vuescan/) if your scanner isn't supported under 
> Linux by sane.  Vuescan has versions for Windows, OS/X and Linux.  There 
> is a list of scanners supported (and unsupported) by Vuescan here:  
> http://www.hamrick.com/vuescan/vuescan.htm#supported
> 
> Also, as a side note..  flatbed scanners with film scanning capabilities 
> aren't all that fantastic really.

I would go so far as "remarkably poor" for most of them. Cheap scanners 
often claim 1200dpi or more, but they can't always actually resolve that 
level of detail - they can capture it, but it's often blurry so you 
aren't getting a "real" result much better than 300 or 600 dpi. Much 
like with digital cameras, a high resolution CCD is far from the only 
thing required for a good image. Interlacing exacerbates that issue 
further. Detail issues that may be acceptable for scanning 15x10 prints 
will quickly become a serious problem for scanning 35mm slides, so be 
careful.

When it comes to colour, be aware that the key measure is not bit depth 
but dynamic range. Dynamic range is essentially the difference between 
the darkest and lightest colours the scanner can "see" in the same 
image. Bigger is better. It ranges from 0 to 4 - any dynamic range 
quotes greater than 4 (and some under, unfortunately) are using a dodgy 
marketing trick that calculates a "theoretical maximum" dynamic range 
rather than measuring the scanner's actual performance. We all know what 
meaning "theoretical maximum" has in computing - 802.11G or USB2 anybody?

Getting the real measured dynamic range out of a scanner vendor can be 
like wringing blood from a rock. If you can manage it, though, it's very 
good information to have.

If you buy a flatbed scanner, especially one with transparency scanning, 
test it at the shop if at all possible. If you have a scanner 
calibration target, definitely bring it along and use it, as well as a 
couple of photos and/or slides, to compare. Check for colour fidelity, 
dynamic range, and blurring at high detail levels.

I'm also going to warn you away from the MicroTek 8700. It's a very 
inferior scanner, especially for the price.

If speed and detailed control isn't an issue, I've seen some incredible 
results from even very cheap consumer scanners. Test carefully and 
you'll probably come away with a winner.

 > If you're going to be doing a lot of
> scanning of (35mm) negative and/or colour reversal film on a regular 
> basis, you might want to possibly look at getting a dedicated film 
> scanner.  You can pick up low-end ones for around $500-600 new.

Not that they're all that super either, at least the cheap ones. The 
POST had an older Nikon model that gave even worse results than a 
flatbed. On the other hand, I have seen some spectacular results from 
even cheaper newer models, so YMMV.

--
Craig Ringer




More information about the plug mailing list