Introduction

The following is an efficient way to prepare scanned book or manuscript pages for use in digitization projects. In fact, it’s kind of a shortcut to have a fast workflow which nonetheless allows for visual control of quality issues.

RAW Input

The  input files should always – repeat: always – be RAW files, at least if the digitizations are done by commercial cameras using CMOS sensors. In case of dedicated scanners TIFFs will be the only available output.

In either case, aim for the best available gamut setting (like AdobeRGB color space) and a fitting file format, like 12bit or 14bit for NEFs and CR2s – or whatever your RAW formats are – and 16bit for TIFFs generated from scanners. In case of the latter, LZW compression does wonders without inflicting losses.

A conversion into the DNG format is not useful and in fact to be avoided because it does not preserve the original output stream from the camera sensor or the scanner but instead only its conversion and interpretation. Different DNGs generated from the same RAW file – say: in 2008, in 2012 and 2014 – are in fact different.

Preserving Recto/Verso

Front and back sides of scanned pages should be saved in separate directories, aptly named recto and verso. Some operations like renaming and cropping will be much faster this way.

Your folder will thus have the following structure:

root
  └signature
      └recto
      └verso

Obviously, this is not necessary if you always scan both pages of a book in a single operation.

Image Operations

For the following operations I will be using Aftershot Pro, originally developed by Bibble but now (unfortunately) in Corel’s hands. FLOSS software like Darktable is obviously preferrable and I might change my workflow accordingly.

We don’t need no separate import operation, as Aftershot allows direct access to the recto and verso folders. This comes handy and is – at least in our case, where keywords and metadata are not much important – much better than, say, Lightroom’s or Capture One’s behaviour.

Right now, I apply the following operations. Some of these are important only if your are using camera devices like the »Wolfenbütteler Buchspiegel« which use a mirror. Here is a link to the relevant Aftershot Presets files stored in .AfterShotPro/Presets/.

• Mirroring
• Recto
• Verso
• White Balance
• EV & Perfectly Clear
• Wavelet Sharpen
• Copyright Information

Fortunately, the possibilities of a RAW developing engine like Aftershot allows for intensive fine tuning and instant control.

Cropping

Nearly always there will be the need to crop the images because they do not show the 2:3 ratio of FF cameras. Aftershot allows to do this once, copy the crop settings and and to apply them to the rest of the images.

Renaming Files

As the files are for now in sequential order only and both recto and verso need to be brought together, we will have to rename them.

Here is the schema used in Aftershot:  [opath-2]-[3rseq]-[opath-1]

Export

Exporting the files is only for specified workflows, e.g. the generation of PDF’s via convert -compress jpeg -quality 75 -limit memory 22000 *.jpg [filename].pdf. These PDFs may be enhanced with indices as described here.

Please note that exports are not in any way suitable to serve as archival formats: only the RAW files (maybe accompanied by the XMP sidecar files generated by Aftershot) represent the original captured data.