Camera scanning is a viable option, but it has limitations. As Peter explained above, many people already have the needed gear. Peter built a very cool semi-automatic system that is described in the long "DSLR scanning" thread. I built a cheap manual solution and described it here: https://www.largeformatphotography.i...ample-approach I'm scanning 4x5 negatives at 2,666 ppi, but the same concept would work just fine for 8x10 (albeit with a lot more pictures). You could shoot at 2,000 ppi and still have a very detailed file.
Aligning everything is actually simple. I use the same trick as darkroom workers use to align their enlargers (a mirror). In my setup I'm making 12 frames for a 4x5 negative, and getting those 2,666 ppi using an APS-C Fuji camera with no anti-aliasing filter. The final resolution is more than adequate for my current needs, and the time it takes is very reasonable. From fluid mounting the film to stitching the 12 images together takes maybe 5 minutes.
But... the major limitation of camera scanning approaches is that some pictures don't stitch because of the contents, e.g., a landscape where the top part of the picture is a clean sky will not stitch automatically. I've tested problem pictures using Lightroom, Photoshop, PTGui, Microsoft ICE... no luck. For difficult pictures that won't stitch you need to be able to use a system that stitches them based on known physical positions rather than contents (and that means using a system's like Peter's that allows for that kind of precision). This is a situation where an actual scanner wins.
--------------
Something else to keep in mind that I'm struggling with now is the impact of the sensing technology on the look of the resulting "scan". I'd like to standardize on one developer so I've been making careful, controlled comparisons of HC-110 and Xtol for HP5+ negatives. According to Kodak, Xtol at full strength is supposed to provide better shadow detail, grain and sharpness than HC-110 at Dilution B (1:31). From what I can see, any differences on the negative are lost when camera scanning. In other words, I find it very difficult to see a difference between the Xtol and HC-110 negatives; they both look equally good. I need to do more testing, but I'm becoming convinced that the characteristics of the digital camera sensor overwhelm the subtle characteristics of the negative. It's not simply about resolution either. A digital sensor isn't film so we shouldn't expect a camera scan (or a scanner scan, which is still a digital image) to look like film. I'm going to be sending out some negatives for drum scanning soon and I'm curious to see whether or not (and how much) more faithful the drum scan is to the characteristics of the negative.
Of course there's a whole other variable, which is the printing side. If you're scanning with the intention of printing, e.g., on an inkjet printer, then the subtle differences you might be able to see on the negative with a loupe are most likely not going to be visible in prints. First, there's what the scanning process does. Second, the printer, the driver, the ink and the paper have a major impact on what the print looks like. I use an Epson 3880 and print with a monochrome inkset via Quadtone RIP. Test prints at 16x20 size of my Xtol versus HC-110 negatives are indistinguishable.
To be clear, I'm not really seeing a "problem" here! I'm pleased with the results and getting the prints I want. In that context, the details and quirks are not particularly important to me.
Reply With Quote
Originally Posted by Randy Moe
Bookmarks