In previous entries, this website explained how to convert physical lengths of films (how many feet long?) into durations (how many minutes/seconds does the movie run on screen?). Here’s some review, using the added resource of the XML files supplied by the Library of Congress for this MEP project.

XML: eXtensible Markup Language files create common information formats and share the data using standard text.  (See whatis.com, among others.)

When the Library of Congress exports a record for a film in its internal collection management database — MAVIS (Merged AudioVisual Information System) — the resulting XML file creates a lengthy string of coded text. Here’s what the beginning of one such file looks like. Useful terms have been marked in bold for this example.

<mavis database="LOC:mbrp" date="2015-02-28T18:32:41" organisation="Library of Congress">
 <TitleWork descr="Moving Image" xl:href="/TitleWork/key/107086"  
xl:title="Muggsy becomes a hero." 
 <identifier>J144969</identifier> 
<identifierType xl:href="/Code/key/IDENTIFIER_TYPE/TITLE/COP"  
xl:title="Copyright Registration Number">COP</identifierType> 
<AcetateFilmComponent descr="Paper Positive Picture"  
 <identifier>LC2635</identifier>  

With some patience, legible details can be gleaned.

The example of the American Mutoscope and Biograph production entitled The Black Viper (1908) demonstrates how the metadata helps tell the story of the work’s transformation across media. In this case  a 35mm production became a pair of 35mm paper print copies (deposited at LOC July 16, 1908). One of the paper rolls (730 feet long) was copied to 16mm film (292 feet) by Kemp Niver’s Renovare Company in April 1962, and positive film print added to the LOC collection in November 1968. The Library made or acquired a 35mm film copy in 1991. In 2012-13, LOC rehoused the two paper copies in new storage containers and created a digital video (MPEG-4 file) copy. All of this and more is embedded in the XML file for The Black Viper. (Highlighted portions at the bottom of this post.) 

The video streaming from the Mediathread platform has a running time of 8:27.  This includes an opening 10 seconds of black leader and 11 more seconds of “front matter” (LOC title card, copyright information, et al.) before the photographic image begins at 0:22. The movie ends on a freeze frame (Niver’s, not Biograph’s) of the final photographic image at 8:21, followed by another copyright notice and black screen.  So a little more than 8 minutes of the original film image track.

With the XML metadata, we can use a film footage calculator (such as this one at Scene Savers) to determine how closely the DV running time we see matches the known physical lengths. The calculator says that the 292 feet of 16mm print would run 8:06 at 24 frames per second.  The 35mm print at 730 feet would run 8:06 too.  In fact the calculator reveals that 292 feet of 16mm film and 730 feet of 35mm film both consist exactly of 11,680 frames of film. A perfect match. Which is rare. Therefore the video access copy, at just over 8 minutes, well matches the description of the source materials. Which is not to assume it’s a perfect match in terms of what we see on screen. The running time matches if the digital video was copied at 24 frames per second. However, a film of 1908 would’ve been projected at a slower frame rate, probably closer to 18 or 16 fps. And in fact the speed of the streaming version appears faster than natural human movement, so it was likely copied at the modern 24 fps speed used for sound films.

Sometimes the different running times of video derived from the same set of film frames is quite obvious.  For example, Cake Walk (American Mutoscope Co., 1897; copyright 1903) runs 24 seconds when copied at 16 frames per second. This is how it’s displayed on the LOC public site at https://www.loc.gov/item/96520361. On the Mediathread platform, a more recent and higher resolution digital copy of the same Paper Print Collection source material has Cake Walk running only 16 seconds (because copied at 24 fps). 

Here’s a desktop video comparison. 

[archiveorg cake-walk-1897 width=640 height=480 frameborder=0 webkitallowfullscreen=true mozallowfullscreen=true]

Sampling of metadata from the LOC XML file for The Black Viper (1908).

<identifier>H113479</identifier> <“Copyright Registration Number“>
<receivedDate>1962-04-18</receivedDate> <AcetateFilmCarrier descr=“Positive Picture“<actualLength>292</actualLength>
<gauge title=”16mm”>16</gauge> <receivedDate>1968-11-06</receivedDate>
<source xl:href=”/Organisation  =”Renovare Co.“>
<collection>NIVER (KEMP) COLLECTION (LIBRARY OF CONGRESS)</collection> 
<notes>According to paperwork in the Niver acquisition file, this component is an original reversal master positive, however film has not been inspected to confirm.</notes>
<receivedDate>1991-01-01</receivedDate>
IDENTIFIER_TYPE/CARRIER/FORMER PP ROLL <actualLength>730</actualLength>
<modifiedDate>2012-03-27T10:46:08</modifiedDate> <notes>Originally, two copies of Paper Print paper rolls were housed in one box (LC 2691). When rehoused by Conservation Division, copy 1 was housed in Box 0419, copy 2 was housed in a separate box, Box 5251. A. Ainsworth 03/27/2012</notes> 
<labelNote>Copy 2</labelNote> <receivedDate>1908-07-16</receivedDate>
<categoryMaterial xl:href=”/Code/key/CATEGORY/A/P” xl:title=”Preservation Material”>P</categoryMaterial
<format xl:href=”/Code/key/FORMAT//A” xl:title=”Safety Film“>A</format><gauge>35</gauge><copiedDate>2013-02-04</copiedDate>  <copyItem descr=”Video/MPEG4″