Colour Film Scanning Revisited - Part 3

EDIT: Part one of this series can be found here.  Part two can be found here.

Yo, what's up film-togs...

Sorry, where did that come from?

At long last, I present Part 3 of this tutorial.  I had originally intended this part to focus on the development of film-specific targets, that could be applied quickly (and with a single click) to *any* image taken on that stock.  This has proved more difficult than I imagined, most (I believe) because of the different response of film in different light.

Also, over the last few months I've found that my original method (with modifications below) to be remarkably resilient and easy to implement across all colour and B&W negative film types.  Slide film (which forms the bulk of my film shooting, especially in 35mm) will be addressed in Part 4 (basically, I'm working on a strategy to develop ICC profiles for specific film types, along the lines of what you'd do for a proper scanner).

So, in this instalment, I'm going to give the final (as far as these things are ever final...) method for my DSLR scanning protocol, and discuss how the method can be used for traditional scanning as well (hint: it's a bit different).  So here we go.

The following video summarizes the process, which is very slightly modified from the method shown in Part 2.


NOTE: The action used in this video can be downloaded from here.  This download includes a section action for B&W film, which is essentially the same as the colour action, except, well, there's no colour correction layer.  Also, in the case of B&W film, there is no grey eyedropper.  Just use the white eyedropper instead.

If you are scanning with a traditional film scanner, the only difference is that you need to scan as a gamma-encoded TIFF (as opposed to a Raw linear TIFF), with your scanning software set to scan as "slide".  Then, you should be able to follow this method and get comparable results.


DSLR Scanning Is Really, Really Good...


9 image stitch from a Nikon D800 + Tokina 100 f/2.8 Macro, producing a 75MP final.  

Your flatbed can't do this.

This is a stitch of a 4x5 Portra 160 negative (stitch done entirely in Lightroom CC) of 9 images (~20% overlap) from a Nikon D800 with a Tokina 100 f/2.8 Macro.  Converted using the method published here (incidentally, Part 3 is still in the works).  The final cropped image is 75MP.  If you've ever wondered what a large format negative scanned at 75MP looks like at 100%, this is it:

...and I could do better.

75MP is not taxing this negative at all.  I could easily push this to >200MP, if I really needed detail for large prints.

Apart from dust removal (which I'd do manually anyway), there is literally no downside to DSLR scanning.