Daniel Reetz, the founder of the DIY Book Scanner community, has recently started making videos of prototyping and shop tips. If you are tinkering with a book scanner (or any other project) in your home shop, these tips will come in handy. https://www.youtube.com/channel/UCn0gq8 ... g_8K1nfInQ

On comparing one page to another.

DIY Book Scanner Skunk Works. Share your crazy ideas and novel approaches. Home of the "3D structure of a book" thread.
Post Reply
User avatar
daniel_reetz
Posts: 2797
Joined: 03 Jun 2009, 13:56
E-book readers owned: Used to have a PRS-500
Number of books owned: 600
Country: United States
Contact:

On comparing one page to another.

Post by daniel_reetz » 06 Aug 2010, 08:57

It turns out there's a long history of devices that are designed to help scholars compare one page to another. Since this is something we SHOULD be able to do effortlessly in the digital age, it pays to look at their hardware to see how it was done and how it was used.

This article talks a bit about "collators" which were devices that overlaid one page on the other:
http://wraabe.wordpress.com/2008/07/26/ ... ion-draft/

Here's a GIS on McLeod Collator:
http://www.google.com/images?&q=McLeod+ ... 80&bih=688

and on Hinman Collator:
http://www.google.com/images?&q=Hinman+ ... 80&bih=688

spamsickle
Posts: 596
Joined: 06 Jun 2009, 23:57

Re: On comparing one page to another.

Post by spamsickle » 06 Aug 2010, 10:37

To add to the "electronic methods" listed in your first link:

A couple of years ago, I needed to extract a list of numbers from a book which contained lots of information other than just the numbers. The numbers themselves were distinctive (like ISBN-13), but were buried in a lot of other information which I wasn't extracting.

I got my list, but I wanted to verify it as easily and as quickly as I could.

Since it was easy for me, as a human, to find the next number in the book, even with all the "noise," I ran the extracted list through a text-to-voice application while eyeballing the bound source. Worked like a charm.

StevePoling
Posts: 290
Joined: 20 Jun 2009, 12:19
E-book readers owned: SONY PRS-505, Kindle DX
Number of books owned: 9999
Location: Grand Rapids, MI
Contact:

Re: On comparing one page to another.

Post by StevePoling » 06 Aug 2010, 15:11

Years back someone perpetrated a fraud by producing a document with Microsoft Word, claiming it was a memorandum from the Vietnam era. (You can probably guess which fraud I have in mind, but I'd prefer to avoid politics.) The most effective debunking thereof was done with an animated GIF that alternated between the fraud image and an image created with Word.

The eye is pretty good at noticing motion. So, I figure if you can reduce the two pages you want to compare to the same scale/size, an animated GIF flipping back and forth is going to give you the most bang for the buck.

abmartin
Posts: 79
Joined: 15 Sep 2010, 15:33
Number of books owned: 2000
Country: USA
Location: Ohio

Re: On comparing one page to another.

Post by abmartin » 15 Sep 2010, 17:22

Just a few thoughts as someone who has done a bit of bibliographic work. First of all, Hinman collators are a migraine waiting to happen! ;) I've only used one once, since there are rarely two or more copies of an early modern handpress book in the same place (except of course for Hinman's work on the Shakespeare First Folio at the Folger which owns way too many copies!). It was exhausting on the eyes.

I like to bring a copy of my own with me to a library (usually printed to scale from a microfilm or an online source after running it through scantailor, which is a lot easier than doing it with imagemagick!) If I can afford it, I try to print a to-scale set on transparencies. Then I can just put it on top - if there is a blur, there is a discrepancy.

When doing digital collation, I make binarized images and scale them to size. (Scantailor has made that easier) I then choose one to use as the copy that I always compare others to. The others I convert to transparent images. Instead of alternation, I just use two layers in Gimp. A straight overlay doesn't do things perfectly, since pages have shrunk differently over time, so I just move the transparent image around. When there is a blur, it means that there is a difference. It's the fastest way I have managed so far. I figure it takes about two or three minutes to compare two pages, including setup time. I wish I had the rights to share some of these images with you. If there is an interest, I can try to get a letter of permission to show a page from two libraries. (might not happen though)

It is a lot of fun to figure out the editing process of a hand press book. It's necessary too! Books printed by hand will always be unique. Each copy will be a mix of corrected and uncorrected pages. How do we know what a document says if we don't know how the document was made? There is nothing worse than a hardcore textual critic thinking that the use of a semi-colon or a colon in a poetic line is always significant. Printers made mistakes after all. The only way to know if a small detail is important is if someone took the time to change it. Pure analytic bibliography can also be taken too far... (classic example pointing out how ridiculous it can sometimes become)

So, in summary - alternating gifs are a lot more difficult to deal with than just compositing two images together.

Post Reply