Scans from Prints – settings

Scans from Prints – settings

Robin Gillanders and a few other friends have asked what my work-flow is for scanning prints that I’ve been posting here.

Here goes:

Epson Expression 10000 XL flatbed scanner, reflective (I also have a transparency head for negatives)

Epson Scan software (I have tried other drivers, but this is a clear interface and does the job) with the following settings:

Screen Shot 2015-09-18 at 10.56.35 AM

click on the image to see more details

Scans are large (about 250MB for an 11×14 sheet), but I want this for archival purposes. Files are saved as TIFF, imported into Lightroom. I do not process these further, other than cropping as described below. For posts to this web site, I export as jpeg, sRGB, maximum size of 1280×1280 pixels at 72 ppi. I export three versions of each scan:

  • the full scan, showing Robin Myers’ wonderful color target (this is very important for verifying color balance and output for any future uses of these files);
  • the cropped scan, showing just the image as I would display it matted;
  • a detail view of about 15-25% of the image area.

The scan resolution, at 48 bit and 600 dpi is high enough to render paper-fiber details without pixelation.

I hope this helps!

2 responses to “Scans from Prints – settings”

  1. Hi Urs – you are welcome! And there are no secrets 🙂 I know you know that. From the very earliest days of my work with platinum-palladium printing, I resolved to share openly whatever I did in the dark(room). Mike Ware and I have gained so much from this approach, and most of all, have had wonderful people like you choose to become our friends! So, thank you!!
    All that said, I wanted to set out some ‘standard’ approaches to sharing the physical details of the prints we all make via this virtual and very variable digital space. I’ve seen many people post scans of their prints, but with little indication of how the scans were, if at all, post-processed, and of ways of even approximately rendering the color according to some standards. I hope more of us will take the above approach (or suggest ways of improving on it without loosing sight of the objective) when sharing work on the internet. Fondly…