Liberating the Arlberg “Viennese Manuscript”

A couple of months ago I posted about extracting Vigil Raber’s sixteenth-century Wappenbuch der Arlberg-Bruderschaft from the “click to pan and zoom” web interface in which it is hosted, but as it turns out, this is not the only armorial manuscript created by this brotherhood.

For a bit of context, the Brotherhood of St. Christopher was established in the fourteenth century to shelter and assist travelers who were crossing the Alps using the Arlberg pass between Italy and Austria. For hundreds of years, they recorded the identities of armigerous travelers (and donors to the brotherhood) by painting the arms in a series of manuscript guest books.

In addition to Vigil Raber’s manuscript (circa 1548), several other copies of these books have survived to the present, although it appears that others were lost over the centuries.

Among the survivors is the so-called Viennese Manuscript, which dates to the fourteenth century. The online version uses the same hosting service as the Vigil Raber manuscript, so I was able to use the same technique to extract the images.

You’ll need to install the ImageMagick toolkit, and grab my little exa_extract.pl Perl script, and run the below command:

perl exa_extract.pl http://bilderserver.at/wappenbuecher/WienerHandschriftEXAv2_52z3/img 324

You can keep these as separate high-resolution JPEG files, but I find it convenient to combine them into a PDF for ease of browsing.

As with the other armorials I’ve processed in recent months, I am making available a lower-quality version of this PDF file (compressed down to just 49 MB rather than the full 844 MB size) for use by those who are not able to run the above script themselves.

Liberating the Livro do Armeiro-Mor

The Livro do Armeiro-Mor (Book of Great Armigers) was painted in Portugal 1506-09 by Jean Du Cros.

As with a number of other period armorials, it has been photographically digitized at high resolution, but the only publicly-available source for these scans is a “click to pan and zoom” web interface operated by the Portuguese Archives Network, which is hard to flip through rapidly, and can’t be accessed when offline at an event.

Gunnvôr silfrahárr has a PDF version posted which has a very helpful name index, but the images have been compressed, so the quality is significantly lower than the original scans. Some of the pages are available at the Wikimedia Commons, but the resolution isn’t as high, and the collection is incomplete.

So, I downloaded each of the full scanned images and combined them into PDF file. There wasn’t any technical wizardry involved — just clicking the download link 277 times.

As with the previous rounds, I’m posting a compressed PDF here (13 MB) and will share the full size PDF or JPEGs (210 MB) by request.

Liberating the Wappenbuch der Arlberg-Bruderschaft

Having had some success with the Gelre armorial, I thought I’d take a stab at extracting another renaissance-era armorial that is only available through a “click to pan and zoom” web interface: the Wappenbuch der Arlberg-Bruderschaft, painted by Vigil (sometimes spelled Virgil) Raber around 1550 in Tyrol, on the border between northern Italy and western Austria.

To run the below Perl script you will need the ImageMagick tools, which provide the montage command that stitches image tiles together. Mac users can install ImageMagick using the following commands:

# Install Homebrew
ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"

# Install Dependencies
brew install imagemagick

This quick-and-dirty Perl script works through the 462 pages of the book, fetching each of the 64 image tiles needed to assemble that page.

#!/usr/bin/perl

my $base_url = 'http://bilderserver.at/wappenbuecher/VirgilRaberEXAv2_52z2/img';

foreach ( 1 .. 462 ) {
    my $page = sprintf( "%04d", $) );
    foreach ( 0 .. 63 ) {
        my $tile = sprintf( "%02d", $_ );
        qx{ curl -o page_$page-tile_$tile.jpeg $base_url/$page/__00000$tile.jpg };
    }
    qx{ montage page_$page-tile_*.jpeg -geometry +0+0 -tile 8x8 page_$page.jpeg };
    qx{ rm page_$page-tile_*.jpeg };
}

The result is a folder full of JPEGs, each 7.7 megapixels covering two pages, averaging 2.3 MB each, which I then assembled into a single 1.1 GB PDF file.

As with the Gelre armorial, I am making a lower-quality version of this file available, which has been compressed to a mere 21 MB, which is somewhat fuzzy but sufficient to scroll through to locate interesting items to view in greater detail via the online viewer.

If you’d like a more-detailed version in either PDF or JPEG format, and are not able to run the above script yourself, contact me for a copy.

Liberating The Gelre Armorial

The Gelre Armorial is a medieval manuscript including over 1700 coats of arms that was painted around 1395 near Geldern, presumably by Claes Heinenzoon, herald to the Duke of Guelders.

It is kept in the Royal Library of Belgium, or KBR, and until very recently it was not available in full online, although images of several pages or noteworthy arms had been posted, such as the earliest known color depiction of the flag of Denmark.

Very recently, KBR published high-resolution images of the complete book, but they were only available through an interactive point-and-click “zoom to view” web interface, and could not be downloaded in a simple PDF format.

These image viewer tools are fairly common in the online library world, and when I looked around I discovered that several people had developed open-source tools that automated the process of downloading the raw image tiles and stitching them together.

With a bit of fiddling, I was able to use one of these tools to capture all 260 high-resolution scans of the Gelre Armorial.

The steps I used are shown below. The first two sections are specifically for Mac users. (Folks on Linux probably already have wget and imagemagick installed, or can use their own package manager tools to obtain them.)

# Install Homebrew
ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"

# Install Dependencies
brew install openssl
brew install wget
brew install imagemagick

# Install Dezoom
wget -O dezoom.sh "https://raw.githubusercontent.com/lovasoa/dezoom.sh/master/dezoom.sh"
chmod +x dezoom.sh

# Download Gelre
perl -e '$t = qq{./dezoom.sh -o gelre-NNNN.jpeg -X 3 -Y 5 -p "--referer=https://viewerd.kbr.be/gallery.php?map=A/1/5/8/9/7/3/5/0000-00-00_00/" "https://viewerd.kbr.be/display/A/1/5/8/9/7/3/5/0000-00-00_00/zoomtiles/BE-KBR00_A-1589735_0000-00-00_00_NNNN/2-%X-%Y.jpg" && sleep 5\n}; foreach $n ( 1 .. 260 ) { $p = sprintf( "%04d", $n ); $s = $t; $s =~ s/NNNN/$p/g; print $s }' | bash

When that process finished, I had a directory full of 260 JPEGs, averaging about 2 MB each. I opened them all in Preview and used the “Print > Save as PDF” feature to create a single massive 494 MB PDF file.

Then I used the “Export > PDF > Reduce File Size” option to knock that down to a low-resolution PDF file that’s only 9 MB in size. The smaller PDF is pretty fuzzy, and while it does have enough detail to allow arms to be identified, one would still want to switch over to the higher-quality file to see any fine details.

You can download a copy of that 9 MB PDF file of the Gelre Armorial for your own reference, or follow the above procedure to capture your own set of high-quality scans.

While I appreciate the work KBR has done to preserve this book, and to capture and share these images, making people click around their site to view individual pages seems unnecessarily restrictive, and as the book is six hundred years old, it can not possibly be restricted by copyright law, so I don’t think I am infringing on anyone’s rights by making this document available to the heraldic community.

A few other notes for folks perusing this armorial:

  • The yellow used for Or has faded almost to white, while the silver used for argent has tarnished and in many cases appears black. In most places the vert has faded a lot, but in a few it’s gotten darker and looks almost black. The gules and azure generally remain vibrant.
  • When trying to interpret a confusing image, this page-by-page index of names and blazons is very helpful.