The Media Ecology Project shared spreadsheet for the Domitor / Library of Congress pilot project contains 20 fields (columns). Enter metadata for these, using the guidelines below.

If you are in the Film Historiography class, you will receive an e-mail invitation to access a spreadsheet named  MEP_LOC_titles_metadata_FHH2017_SHARED.xlsx

title: Use the title printed on the opening shot of the movie; if none appears, use the title in the copyright catalog. If they differ, note this in the field “alternateTitle.”

date: Year of release; or year of production if release date is not known.

releaseDate: Enter the day and month, or closest documented period (e.g., 25 Feb 1909; Dec 1908; Mar-Jul 1911; ca. 1901. (Default to the DD-MM-YYYY format created by Microsoft Excel.)

copyrightDate: 31-08-1899 (Default to the DD-MM-YYYY format created by Microsoft Excel.)

director: If a director is credited. In the early years of cinema, films did not always have a “director” per se. A camera operator shooting alone will sometimes get a director’s credit in reference books created years later. Before the director-driven studio system developed in the 1910s and 20s, it was not uncommon to have two directors credited on a film. Although the AFI Catalog lists the director of Three American Beauties (1906) as Edwin S. Porter, many other authoritative sources credited Wallace McCutcheon as co-director with Porter. 

producer: Leave this field empty. (For this period of film history the role of the producer was not well defined in the industry. We also did not want this field to be a confusing mix of  (a) names of people who oversaw a movie production and (b) corporate names of production companies.

writer: as credited in primary or reliable secondary sources. If a motion picture is adapted from a literary source, for example, that author’s name will go here, after a film scenarist or scriptwriter (the term used in Kemp Niver’s catalog Early Motion Pictures).

cinematography: Niver uses the term “cameraman.” I would have used “cinematographer” to be consistent with director, producer, writer, actor categories. But “cinematography” is what previous MEP contributors have been using. You may list camera operators other than the head of the camera unit, but that’s optional. Most titles will have one person listed, while a few will have two. 

actors: List any identified performers or people on screen (or on the soundtrack). I would’ve used the term “cast,” since nonfiction recordings are not generally of actors in the conventional sense.

music: if credited. Thus far, none of the silent films supplied by LOC for this project include musical scores or any other sound elements. For other films appearing in MEP classes, you might also find music that is not in official credits but which can be identified. For example, The MIRC note for Fox Movietone News Story 5-679, Massed bands — outtakes; John Philip Sousa conducts selected members of 23 bands in playing “Stars and Stripes.” Edwin Franko Goldman leads band in playing “Young America.”

editor: if credited or known.

synopsis: Paste in any pre-existing synopses, including the Niver “summary.” Other sources might appear in producer-distributor catalogs or trade magazines. Put each found source in quotation marks, followed by a citation of the source in parentheses. List more than one if you locate them. If no existing synopsis is found, create your own based on your viewing of the movie. We are no prescribing specific terms or methods. Do summarize. Don’t describe every shot unless the work is very short.

length: The total running time of the movie in hours, minutes, and seconds. Use the formula that spreadsheets require: 0:08:44. Since the running time will include video added before and/or after the movie itself, you might mention this in the notes section. (“This video file is linked to the LOC webspace for Waterfall in the Catskills (1897) www.loc.gov/item/00694329 . This embedded MP4 is 60 seconds in duration, but the images from Waterfall in the Catskills run only 25 seconds.”)

source: Use only for literary or other sources for a movie’s story or subject. This does not refer to archival sources for the films, nor to citations of printed sources of information about the film.

productionCompany:  Occasionally there might be more than one company to enter.

physicalLength: The length of the film strip. Very often, a film will be issued in more than one length. Where known, list all you can confirm. In this field, you can identify the source of the data. Where needed, clarify what the measurement refers to. The physical length of a 35mm film will usually equal the length of the paper print made from that film. But copyright deposits varied widely; some producers deposited only excerpts or stills. Others only a few frames. Note also that 16mm and 35mm copies of the same film will have different measurements (35mm always longer). List whatever you can verify. If the source says “85 ft.” enter that, even if you do not know the film gauge. If a source lists meters instead of feet, use that. No need to convert the units.

When seeking to understand the relationship between physical length and running time, you can use an online tool, such as this Film Footage Calculator from Scene Savers. (Click here.) It allows you to convert feet of film into minutes and seconds, factoring in both the film gauge (35mm, 16mm, et al.) and speed that the film would be run (through a projector or a camera). A 35mm film that is 85 feet long (running at 16fps) has a duration of 1 minute, 25 seconds.  Other tools do similar work, such as Kodak’s Film Calculator (“Speed, feet, frames, and time”) and Panavision New Zealand (“Film Footage, Frame and Running Time Calculator.” You can also determine if the number of feet you see listed in a catalog is the length of a 16mm print or 35mm print.

subjectHeadingLOC: For items in the LOC catalog, a trained moving image cataloger has assigned “controlled vocabulary” for conventional library descriptions. Copy that information if it exists. Other sources might assign a film with a genre or category, etc. (novelty, scenic, topical, etc.) using non-controlled vocabulary; but only use LOC Subject Headings (LCSH) in this MEP spreadsheet. Where such non-LOC information is useful, enter it in the notes field (e.g., Moving Picture World listed this film under the categories “Educational” and “Travel”). If you want to know about the official Subject Headings developed by LOC, there’s a lot of information here.

citations: Enter citations for the publications in which you found the metadata. Cite as you would in a footnote. Use Chicago Manual of Style. Many titles will have Niver’s catalog listed as well as the Howard Lamarr Walls copyright catalog. For these two citations, you can list the short form: Niver, Early Motion Pictures; Walls, Motion Pictures, 1894-1912. 

alternateTitle: List any published variations on the text in your title field. Even if variations are minor, list them. If translated titles are found, list them. Edison Kinetoscopic Record of a Sneeze, Jan. 7, 1894 (the copyright title) has the alternate titles Edison Kinetoscopic Record of a Sneeze; Fred Ott’s Sneeze; The Sneeze; and simply Sneeze. Niver and the LOC catalog often identify alternate titles. You will likely encounter other variations in your research. Often sources will omit The, A, or An when listing a title. Or they will simply get a word wrong. Or misspell something. Include even those. Copyright title Circular Panorama of Horse Shoe Falls Winter (Niagara Falls), alternate title Circular Panorama of Horseshoe Falls Winter (Niagara Falls).

notes: Any other facts you find significant. The notes you create for the spreadsheet field can be of any length. There is also a Notes field in the Mediathread platform (click on EDIT THIS ITEM or the pencil icon when viewing a video). You can paste the same notes into both the spreadsheet and the Mediathread entry.

---

Sometimes, as you know, a source (even library catalogs and well-researched books) will get facts wrong. If you find discrepancies or errors, list whatever is the most reliable information. In the notes field, describe where errors or discrepancies are known. Examples (hypothetical): “Musser credits Porter and McCutcheon as co-directors, but all other sources omit McCutcheon.” “The NYPL catalog mistakenly lists Frank Wilson as director; all other sources credit Gaston Méliès.” “The newsreel catalog describes one of the teams in the 1930 World Series of baseball as the Oakland Athletics. It was the Philadelphia Athletics.”  These need not be in the initial notes you enter into the spreadsheet, but save this information in your off-line work. It will likely be useful in a later report or “composition” you write.

A few addenda to the data and metadata about Three American Beauties.

 The Kino-MoMA Edison DVD says that this film was “mastered at 14 frame per second,” while the earlier Treasures from American Film Archives DVD indicates the transfer was done at 16 frames per second. Both sources used the MoMA color print. When these two digital copies run on a web platform the speeds yield running times of roughly 45 seconds (16 fps) and 50 seconds (14 fps).

Here’s a comparison.

The video on the left is from the NFPF website; on the right is a copy of the the film hacked from the Kino DVD and posted on YouTube. (To download the file 3ABx2x.mov, go to the Internet Archive.)

---

Color: hand-applied dye, stenciling, and tinting.

The sources we reviewed ascribed three different processes to this film. The only copy that has been studied or digitized is the Museum of Modern Art’s 35mm nitrate print.

The original Edison descriptions said that Three American Beauties was “handcolored.” The Timeline of Historical Film Colors identifies “Hand coloring” as one category: “Coloring of individual frames by the use of very fine brushes. The process was previously applied to lantern slides. Any water-based translucent dye was suited . . . most often acid dyes.” The Timeline gallery offers 4 photographs taken from the MoMA print. “In contrast to stencilled films, hand-colored ones often have soft outlines and the application of color varies from frame to frame.”

When inspecting the print before photographing frames for her Timeline, Barbara Flueckiger reports (in a February 14 Facebook response to the previous blog post) that MoMA film archivist James Layton thought that Three American Beauties appear to have both hand-coloring and stencil color.

The Timeline continues: “Stencil colored films can be identified by the sharp outlines that define the colored areas. Color hues were most often soft pastels. The stencil colored images have a painterly quality.” Also: “Stencil coloring required the manual cutting, frame by frame, of the area which was to be tinted onto another identical print, one for each color.”

Adding colors to black-and-white film prints in the first two decades of cinema was invariably done by teams of women doing the tedious work. In France, the Pathé company employed up to 200 women colorists to work on stencil color prints. In the United States, the hand coloring of kinetoscope films was “contracted out to the wives of Edison employees” as early as 1894-95 (Musser, Before the Nickelodeon, 21).

Finally, the head title card created for Three American Beauties on the Kino-MoMA DVD calls the film “Hand Tinted.” However, the term tinting for film laboratories refers to putting positive prints through dye baths to add a single color.

As copies and “instantiations” multiply, note one other alteration to what you see above in the YouTube stream. Once uploaded, YouTube (speaking as “we”) asked me (the uploader) if they should “enhance” the “video’s lighting.” I allowed it, just to see what it would produce.

 

As Paolo Cherchi Usai’s useful dictum puts it “Every print of a film is a unique object. . . .”^† 

Paraphrasing, we might say there are no films, only prints of films. With a broader media archaeology approach we must expand that to say copies (digital, video, film, paper) of a motion-picture work are seldom if ever exact clones. In this case, an automated YouTube technology created a new and unique digital file, an “enhanced” “copy” of a QuickTime movie I created on February 15, 2017, and first saved as a file named “3ABx2x480.mov” with these properties: dimensions of 1284 × 460 pixels; with an H.264 codec; color profile HD (1-1-1); duration 50 seconds. Meanwhile, in a parallel universe, the Internet Archive generated three variations on the MOV file, there named ThreeAmerBeautiesx2 : MP4 and OGV video files and a torrent file. 

The metadata details here are not inherently significant; what matters is understanding that any time you study a “film” you are analyzing but one iteration of a work. Media archaeology is the practice of digging for other iterations and assessing their relationships. The humanities methodology of “close reading” of a work is arguably more relevant than ever, since meaningful differences appear across objects. In the case of the various items called Three American Beauties, we have seen differences in color, speed, length, duration, added sound, titles, and more — all without ever seeing or touching an actual film object.

---

† Paolo Cherchi Usai, Silent Cinema: An Introduction (London: BFI, 2000; 2nd ed., 2010), chapter 3, “The Ethics of Film Preservation.” Previously published as Burning Passions: An Introduction to the Study of Silent Cinema, tr. Emma Sansone Rittle (BFI, 1994), and Una passione infiammabile: guida allo studio del cinema muto (Turin: UTET, 1991). 

 

This blog gives instructions and tips on doing film history research, particularly the work our course does on the Media Ecology Project’s pilot initiatives with (1) early cinema from the Library of Congress (Paper Print Collection and more) and other online sources, and (2) the University of South Carolina Moving Image Research Collections (MIRC), including its newsreel and newsfilm holdings, Chinese Film Collection, and others. Here’s the first in a series of posts. — Dan Streible

---

Looking for silent-era films? A first checkpoint is Treasures from the Film Archives, a database maintained by the International Federation of Film Archives (FIAF). It contains information about more than 60,000 silent-era works held in more than 100 film archives. The database contains both filmographic data (about the works) and metadata about the physical holdings of specific copies of those works. Below, is a case study of researching the Edison/Porter film Three American Beauties (1906). (Or is it 1907? Read on. . . . )

NYU Libraries offers two on-line portals to Treasures from the Film Archives, although finding them is not as obvious as it could be. Searching “Treasures from the Film Archives” in BobCat (Bobst Catalog) first lists a book by the same title: Ronald S. Magliozzi, Treasures from the Film Archives: A Catalog of Short Silent Fiction Films Held by FIAF Archives (Scarecrow, 1988). Find that in the Bobst Reference room, call number PN1995.75 .M335 1988).

The second entry is the database you want. (Confusingly, it’s labeled as a “Journal” and given the title FIAF international filmarchive database (Online : SilverPlatter). When you click to access it, ProQuest leads you to a page called FIAF International Index to Film Periodicals Database. Despite the misleading name, searching this database yields citations for film and television periodicals but also for film holdings in archives (i.e., the Treasures metadatans-serif;”>Another ProQuest database, “Screen Studies Collection,” searches the same sources as above, but also searches the invaluable reference we mentioned in class: The AFI Catalog. In this database, the records display an icon for FIAF items and a red icon for AFI entriesns-serif;”>Here’s a search for the Edison/Porter film we watched called Three American Beauties .

Another idiosyncratic element of this database is that the entries for motion pictures (as opposed to books, journal articles, etc.) are categorized as “Artistic & Aesthetic Works” (an odd term, no?) and marked with an icon of a painter’s palette. You can filter the search to see only the records for film holdings. Helpful.

From the FIAF Treasures database, we glean this:

THREE AMERICAN BEAUTIES. 1906.
Produced and distributed by Edison Mfg. Co. (US)
Director: Porter, Edwin S.

Five archives hold film copies:
• Museum of Modern Art ­ Department of Film (New York) [USM]
• Cinémathèque Québecoise (Montreal) [CAQ]
• BFI National Archive (London) [GBB]
• Library of Congress (Washington) [USW]
• Academy Film Archive (Los Angeles) [USF]

Holdings that can be accessed (i.e., studied on site or loaned) are: a 35mm print (Québec), a 16mm print (LOC), and one of a format not indicated (BFI).  The Library of Congress has a 16mm acetate dupe negative and a 16mm acetate positive (in other words a print to watch on a flatbed viewer or put on a 16mm film projector). Acetate refers to the “safety film” stock used after the highly flammable nitrate cellulose film stock ceased being manufactured circa 1951. We also know from other sources, including the booklet for the DVD we watched of Three American Beauties, that the MoMA copy is a nitrate original 35mm print, hand-colored. That’s one preserved and digitized. Thus far I’ve found no copies of 3AB on DVD or online that are not derived from the MoMA print.

Always check the Library of Congress public catalog. It will not show you the full and deeper interior database of motion picture holdings (called MAVIS), but it might confirm that LOC holds a copy of the film. Remember that many films in the collection are not in the public catalog. However, it will often give you rich metadata about the physical items in the Library’s collection. Searching “three american beauties” at catalog.loc.gov does show a record for the Edson film. However, we learn that Three American Beauties is NOT part of the Paper Print Collection. Instead it’s a 16mm print from the George Kleine Collection, preserved in 16mm. (George Kleine was a major distributor of films in the early era of the U.S. motion picture industry.)

Always check the catalog created by our hero from The Film of Her, Howard Lamarr Walls, Motion Pictures, 1894-1912: Identified from the Records of the United States Copyright Office (Library of Congress, 1953). It lists more than 8,500 “works” corresponding to more than 6,000 titles registered in the Copyright Office as photographs — but which Walls verified were motion pictures. His list, therefore, is roughly twice as long as Niver’s.

Here we note a discrepancy: two copyrighted works using the same title, but the later (from October 1907) has four scenes instead of three.

What about the two AFI Catalog records for Three American Beauties? Entry one contains most of the information from FIAF, but brings additional details. 

Produced April 1906, and Publication date (meaning when it was first sold or exhibited) June 1906.
Physical description: b&w; 60 or 65 feet. (The length of the 35mm film prints sold by the Edison company.)
Copyright Information: Thomas A. Edison, 1 May 1906, H76443-H76445 (the LOC copyright registration number, as also found in the Howard Walls catalog and Niver book, Early Motion Pictures).

Then, very helpfully, the AFI Catalog lists the primary sources in which its researchers found about Three American Beauties:  LCMP (the Walls catalog), p. 61; New York Clipper (a show business trade magazine from the early twentieth century), 16 June 1906, p. 470; The George Kleine Collection, (another reference book from LOC, on PDF in our shared folder), p. 135, and Magliozzi’s book Treasures from the Film Archives, p. 250. Going first to this reference database saves you time in searching for the primary resources of 1906. Now you can go straight to the Clipper item (an Edison ad, as it turns out).

The other AFI Catalog record is for the title Three American Beauties [No. 2]. This is described as “Silent: b&w: Handcolored, 85 feet.”  Published on November 23, 1907; deposited for copyright October 7 for Thomas A. Edison, with new registration numbers H100637-­H100640. So it’s a remake of the first film. This film was mentioned in the New York Clipper, November 23, 1907, and in Moving Picture World (the most important trade journal for the American film industry, starting that year), November 23 and again January 18, 1908. Such documentation is especially important because No. 2  is not known to survive in any form.

Here’s the MPW ad from Edison, Nov. 23, 1907.

An important reminder about Lantern: The best method of searching and retrieving original motion picture trade press is from the powerful Lantern search platform built as part of the Media History Digital Library, which you have now read about in The Archlight Guidebook. Exploring search terms at lantern.mediahist.org allows you to read even very small items and ads. You can track how a film or a subject was covered month by months in some cases. Searching “three american beauties” and “edison” returns (as of today) 54 citations. One is the 1960 BFI catalog (see below). The other 53 are ads and release information in Moving Picture World and the New York Clipper — 53 instead of the AFI catalog’s 4. The majority are from the Clipper, which is significant because a previous generation of early cinema history relied heavily on access the Moving Picture World  on microfilm.

Switching to The George Kleine Collection of Early Motion Pictures in the Library of Congress: A Catalog (1980) we get new details.

Note that the cataloguer has added the subject terms novelty and cartoon (is this a cartoon?). Then in the classification system at the end of the record we see each of the three shots labeled “Drama” (are these dramas?). We also get a specific day of production and precise location — Porter shooting March 17, 1906, in the Edison studio in New York. The summary above was written by the cataloguers (Rita Horwitz and Harriet Harrison with  Wendy White) using film analysis terms (“extreme close-up”) rather than quoting from a 1906-07 description. But then that curious and surprising sentence: “A field of stars, supposedly from the flag, is animated to form the words ‘Good night.'”  

Looking at the video copy we have from MoMA’s print, I see the field of stars (in a fourth shot after the three shots of “beauties”), but I see no animation of the stars to spell out “Good night.” (Do you?)  Does the LOC print in the Kleine Collection contain extra footage? a longer shot of stars that transforms into text? Or is this description taken from some written source rather than any surviving film print? We’ve seen other Edison-Porter films of 1906 that use animated letters for novel effect (the How Jones series, seen in the MEP Mediathread collection).

The BFI’s National Film Archive Catalog (1960) adds another clue, but not a clear answer. A telegraphic style of description. The date assigned is 1907, but the title does not include No. 2, which is inconsistent with the other data we’ve accumulated. The physical length of the BFI’s film print — 78 feet — might suggest it is part of the longer film, No. 2, said to be 85 feet long. If the first version of Three American Beauties from 1906 was only 60 to 65 feet long, it seems more likely this British archival copy is No. 2.  Presuming that the metadata is accurate. Which it might not be. The note about the animated “Good night” suggests that this part might have only been part of No. 2.  We’d have to see the BFI print to know. (The BFI catalog is also in our MEP folder.) 

[* Update; August 9, 2023. Confirmation from BFI National Archive that it holds a 1907 print, which indeed shows different footage. It will doubtless make this copy available in the near future.]

So even for a very short film from early cinema, we were able to discern some of its characteristics and history just by using available reference books and databases. Going to primary sources would tell us more.

And of course we can compare that with what we see when watching it.

For fun, here’s one of the YouTube copies, posted for some reason with a Russian translation of the title. Published on the “History VA” channel, July 16, 2015. Like other copies online, this file was ripped from the DVD set Edison: The Invention of the Movies (Kino International, 2005). It has a piano score (created for the DVD by Philip Carli) that differs from the piano music Martin Marks recorded for the earlier DVD Treasures from American Film Archives (National Film Preservation Foundation, 2000).

The DVD notes from both packages are given away online by the producers. Charles Musser’s notes for the Kino set add a co-director for the film, Wallace McCutcheon with Edwin Porter. “Often hand-tinted, this short film was typically used by exhibitors to conclude their programs. It elaborated on a popular practice among exhibitors of the 1890s. They ended their programs with a film of the American flag waving in the breeze,” Musser adds.

Scott Simmon’s notes for the NFPF edition of Three American Beauties add still more. The notes from these Treasures DVD series are an excellent model for doing deep research and translating them into the most relevant context for contemporary lay viewers. Film scholarship for a public readership.

Transfer Note: Copied at 16 frames per second from a 35mm print preserved by the Museum of Modern Art. New Music: Martin Marks. Running Time: 40 seconds.

Three American Beauties is colored by stencil, a more common process in France than in America. Color was painted on release prints using a separate cut-out stencil for each color. The brief, patriotic film was shown in nickelodeon theaters at the end of a program, and proved so popular that Edison’s original negative wore out and had to be shot again, complete with in-camera dissolves. The Museum of Modern Art preservation copy reproduced here is probably from the first version, shot by Edwin S. Porter and Wallace McCutcheon in New York in March 1906. Porter remade the film in September 1907. —Scott Simmon

Note that this DVD edition says the film was transferred at 16 frames per second.

For most every case of Edwin Porter films or the Edison company, the go-to scholarly source is Charles Musser’s work, especially his book Before the Nickelodeon. But this interesting little film Three American Beauties is not mentioned at all in his long and scrupulously written book (based on his NYU Cinema Studies dissertation). Perhaps the film was not so accessible at the time he was researching. The scholarly tome that does feature the film is this one, edited by Andre Gaudreault, Gunning’s co-inventor of the term “cinema of attractions.”  But other than identifying the film’s title on the back cover photo credit, the book does not mention Three American Beauties at all.

The film is mentioned briefly in Musser’s other key work for our study, The Emergence of Cinema: The American Screen to 1907, first published in 1991. He describes the Lubin company of Philadelphia and its practice of quickly remaking popular subjects made by competing companies. Lubin’s film Good Night (November 1906, 65 feet) is said to be similar to Edison’s. This is the first indication that the animated stars spelling out “Good night” were likely present in the original 1906 version as well as the 1907 Edison remake. Alas, Lubin’s Good Night is a lost film too. For now any way. It might well be sitting in an archive unidentified or awaiting rediscovery in an orphaned private collection.

But wait! Lantern shows us several ads for Good Night that Lubin placed in the New York Clipper, including this one from 1907. The film might be lost, but S. Lubin published two frames from it (lacking the color, of course).

Lubin’s Good Night  is designed rather differently  from Edison’s.

A  final question:  Who is the woman in the Edison film? (For that matter, did the same actress appear in the version Porter re-filmed a year a half later?) The Edison catalog only refers to “a bust of beautiful American girl.” It seems to be a fact one could find. But so far, none of the reference sources or archival clues tell us who she is.

---

 

 

---

Next up: a list of key reference books and other databases for researching early cinema.