NCC News: News

Japanese Studies Spotlight: The Japanese Paper Film Project: Preserving the History of 1930s Kami Firumu

The NCC is collaborating with institutions and scholars to release a monthly series on our blog entitled Japanese Studies Spotlight. These features showcase exciting online collections available to researchers and students in Japanese Studies, introducing the archive or project, describing their contents, and demonstrating how they can be usefully engaged in research or in the classroom. If you are interested in submitting something to the series, please contact Paula R. Curtis, NCC’s Digital Media Manager, at digitalmediamanager@nccjapan.org.

Eric Faden, Professor of Film/Media Studies, Bucknell University

In 2017, while conducting research on Japanese cinema and teaching at Kyoto’s Doshisha University, I set out to work on Japanese magic lanterns (called utsushi-e or nishiki kage-e), an early type of moving image projection. However, in my first week in Japan, I met with Yoneo Ota, the Director of the おもちゃ映画ミュージアム (Toy Film Museum) in Kyoto, and my focus suddenly changed.

When discussing nishiki kage-e, Ota-sensei said [and I am paraphrasing]: “Nishiki kage-e is sort of interesting but what you should really be working on is this…” He then showed me a very old, small cardboard box and inside was a truly unusual object–a wound up film strip that was not celluloid film. Instead, it was a film printed on paper. Called kami firumu 紙フィルム (lit. “paper film”), it had framelines and perforations like a celluloid film but the paper was opaque rather than transparent.

Figure 1. Paper films at Kyoto's Toy Film Museum

Ota-sensei then showed me a paper film projector and explained how it worked. Excited by this unusual format, I asked Ota-sensei if we could watch some of the films. He replied: “No . . . it is not possible,” and then explained how fragile the films were and the poor condition of the original projectors. Ota-sensei then slyly wondered aloud: “if only there was a way these films could be digitized.” And thus began The Japanese Paper Film Project.
 

Kami Firumu in Context

Producing films on paper may seem retrograde given celluloid film’s clear success as a moving image medium by the 1930s. However, thinking about paper and moving image culture in a larger context helps make sense of kami firumu. Japan has a long history of using paper and images as a temporal/spatial narrative medium. For instance, the earliest surviving examples of emakimono, illustrated narrative handscrolls made of paper, can be found beginning in the Heian period (794–1185) and expanded during the Kamakura period (1185–1333). Importing and modifying a practice from China and Korea, Japan used these paper scrolls as a medium for telling stories with images from earlier sections connecting to the images that followed as the scroll unfurled, not unlike a reel of film.

In the nineteenth century (during the Meiji Restoration), we can think about another paper-based moving picture, kamishibai. In the 1930s, gaitō kamishibai 街頭紙芝居 (streetcorner kamishibai) featured performers who traveled by bicycle from neighborhood to neighborhood, setting up their performance in public spaces. Perched on the bicycle’s rear shelf was a large wooden box with a proscenium (a structure, often with a stage curtain or wall, that forms or frames the stage). The proscenium housed a stack of ten to fifteen paper or cardboard pictures, which the performer then used to narrate a story, deftly removing each of the pictures to reveal the next slide underneath.

Scholar Sharalyn Orbaugh notes some kamishibai artists studied film theory and incorporated cinematic techniques into their paper theatre so that they produced “a variety of visual and narrative effects through such cinema-inspired techniques as continuity . . . ‘panning shots’ or ‘close-ups.’” [1] Indeed, kamishibai’s visual style was so popular that when television premiered in post-war Japan, it was initially called denki kamishibai 電気紙芝居 (electric kamishibai). We should remember that many early pre-cinematic devices used around the world like the Thaumatrope, Kineograph, Phenakistoscope, and flip books (all popular toys in Japan) were paper-based devices that created moving pictures. Given Japan’s rich image culture, film and paper have enjoyed a long history together.

A Very Brief History of Kami Firumu

Several companies produced kami firumu for the home market between 1932–1938. [2] The three main companies were REFCY (Tokyo), Katei Toki (Osaka), and the recently discovered Tsukiboshi (Tokyo). [3] The films were live action (copied from celluloid) or original animated films. Films could be in black and white or color and had running times between one to four minutes, although there are several cases of multi-reel films where the story continued into the next reel. Many films even had synchronized soundtracks via 78rpm records that paired with the films. Each company manufactured and sold their own proprietary hand-cranked projector. The projectors were episcopic projectors that reflected light off of the opaque paper and into a lens.

Figure 2. Katei Toki's Tokkyū Chūshingura （特急忠臣蔵） Because the projector's lens "reversed" the image, the paper film strip is printed as a mirror image. Print courtesy of the Toy Film Museum.

REFCY was the first company to market and curator Chiyo Inaba details the work of printing engineer Tsujimoto Hidegoro and the various patents he filed in 1932–1933 for paper films. [4] Importantly, Hidegoro’s background was not in the film industry but the printing industry, which shaped paper film’s form. For example, scholar Fumiaki Itakura notes that REFCY film lengths (15, 30, and 45 meters respectively) resulted from the 787mm x 545mm printing sheets which, when cut in 27mm film strips, yielded 15 meters. [5] Perforations were punched into each individual strip and then they were glued together every 750mm to create a roll of film.

Katei Toki and Tsukiboshi films and projectors appeared shortly after REFCY. Katei Toki’s system was remarkably similar to REFCY’s, with a nearly identical frame size, although their projector was incompatible with REFCY films. The most obvious difference was Katei Toki’s perforation system which, perhaps concerned about the paper’s delicate nature, used two horizontal perforations rather than REFCY’s single perforation. In addition, Katei Toki and Tsukiboshi embraced sound by using 78rpm records roughly synched to their films.

Kami Firumu Forms and Genres

Kami firumu took many different forms and spanned a variety of genres. Live action films were duplicates (known as “dupes”) of other studios’ content [6] and might fall into several categories: documentary, ethnographic, newsreels, propaganda (the Second Sino-Japanese War looms large in many films), as well as highly condensed versions of domestic and international theatrical films. In some cases, the live action kami firumu version contain fragments of lost films. For example, the 35mm print of Tomiyasu Ikeda's 1930 Genroku Kaikyo Daichūshingura (元禄快挙大忠臣蔵) featuring early Japanese star Denjirō Ōkōchi is believed lost. However, the private Machiko Kusahara Collection includes a 2-reel kami firumu version of the film.

Anime proves the most dominant kami firumu form but, like live action kami firumu, there is a surprising variety. Each manufacturer had a distinctive style. Tsukiboshi films used extremely rich, saturated colors while REFCY and Katei Toki echo the 1930s Fleischer Studios’ animation style, such as the large eyes and bold lines of characters like Betty Boop. Films incorporated sophisticated editing and animation techniques. For example, REFCY’s Muteki Ōhei kappa odori (無敵凹平かつぱ踊) contains seven scenes with thirty-four shots plus eleven intertitles and opening/closing credit text. Editing transitions include straight cuts, iris transitions, and wipes. Shots utilized narrative continuity editing protocols of screen direction, close-ups (vignetted), match on action, and eyeline matches between shots. Finally, the film also mixes panning and tracking shots with the static frames.

In terms of content, films focused on popular historical figures or events that have taken on a quasi-mythical air like the revenge of the 47 rōnin. The Katei Toki version, called Tokkyu Chūshingura (特急忠臣蔵), is in two parts and includes a synchronized soundtrack with music, sound effects, and benshi narration. Set in the Edo period, the film features period appropriate costumes, rituals, and architecture (although one character does ride a motorcycle!). Many animated kami firumu celebrate this era with characters like Sarutobi Sasuke (a superhero ninja), Tsukigata Hanpeita, and the fictional rōnin Tange Sazen.

Beyond historical figures (real and fictional), several animated kami firumu draw on Japanese mythologies, fairy tales, and Shinto/Buddhist mythical creatures, such as kappa (Shinto god water sprite), foxes, and, of course, shape-shifting tanuki (raccoon dogs). In addition, popular manga and anime characters from the 1920s and ‘30s find their way into kami firumu, like manga artist Tagawa Suihō’s Norakuro (an orphaned dog who serves in the military) and Sutakora Sacchan. Finally, other kami firumu anime concerned the Japanese military and their involvement in China and ranged from films for children (Military Parade with animated marches plus anthropomorphic tanks and trains) to Tsukiboshi’s Japanese People are Here, which recounted a 1933 historical incident in China and appealed to a more adult audience.

The Challenge of Preserving Kami Firumu

The Japanese Paper Film Project, based at Bucknell University, aims to preserve and promote research on these important materials of film history. An interdisciplinary and international collaboration between Bucknell faculty and students as well as The Toy Film Museum in Kyoto, the Itabashi Science Museum in Tokyo, and many other partners, our collaborators worked at length to devise the most effective way to handle the film. Kami firumu’s quasi-handcrafted nature presents significant preservation challenges. Given the films’ unusual format and the conditions outlined earlier, the films’ production process caused significant inconsistencies in the film strip that make using a traditional film scanner difficult. Moreover, because many paper films were held by Japanese collectors or archives unwilling to loan them out (some museums classify their paper film collection as national treasures), we needed a solution where the scanner could be lightweight and portable so it could travel to Japan.

Figure 3. Japanese Paper Films will disintegrate without proper care or preservation, highlighting the urgent need to scan and preserve these films.

My Bucknell colleagues and their students in Mechanical Engineering and Computer Science collaborated to develop a scanner specific to paper films that could automate the process as much as possible. We approached the problem by building a rig, nicknamed Kyōrinrin, after a Japanese yōkai (spirit) that protects and preserves lost, forgotten, or unstudied Japanese scrolls, that could transport the film via light tension (instead of by sprockets/perforation) with relatively few touch points. [7]

Given paper film’s delicacy, we wanted a rig that could transport the film strip but minimize any wear/tear or damage to the original materials. We then record the film strip in motion with a 6K digital camera shooting at a very high shutter speed and frame rate (1/480th and 60 frames per second). This approach allowed us to quickly scan the raw materials, with each film reel taking about ten minutes.

Figure 4. Our paper film scanner, Kyōrinrin.

Once captured, we then have a “stream” of images of the paper film progressively moving through the rig’s gate. At this point, we use custom software to sort through the ocean of frames to select and align individual progressive frames. Those selected frames are then stitched together to reconstitute the original paper film’s individual frames. Each film takes approximately fifteen minutes to process. This stage provides us with a roughly stabilized, playable film. Finally, we use post production software like DaVinci Resolve and Fusion to smooth the stabilization, adjust playback speed, synchronize sound, and do any final color correction. While certainly not “labor free,” this process is far more automated, produces excellent results, and given the scanner’s transportability, a convenient package for working with Japanese museums, film archives, and collectors.

The Japanese Paper Film Project has come a long way from when I first saw kami firumu in 2017. So far, we have processed and preserved about 70 films of the 200 scanned. As we process the films, we are also cataloging each print while researching the content so it can be contextualized for audiences. Ideally, we hope to publish the catalog plus the films themselves at some future point so the films can be enjoyed, researched more, and taught in the classroom. They do, after all, offer a fascinating snapshot of 1930s Japan. In the meantime, we are curating select screenings of the preserved films. Last April, we premiered a short program of paper films at The Museum of the Moving Image as part of the Orphan Film Symposium. Next month, we will present a longer program at Brooklyn’s Light Industry. For short clips, please visit our Twitter page.

Acknowledgements

The Japanese Paper Film Project could not exist without collaborations from many institutions, colleagues, and students. At Bucknell, Engineering Professor Nate Siegel and his student Alina Arko designed Kyōrinrin. In Computer Science, Professor Joshua Stough and his students Yuhan Chen and Jackson Rubiano created our frame selection software. Film/Media Studies students Harry Winter and John Ogunwomoju worked on stabilization. My research collaborator in East Asian Studies, Professor Elizabeth Armstrong, guided this project’s research and served as chief translator and interpreter. Finally, we are indebted to The Toy Film Museum (Kyoto), The Itabashi Science Museum (Tokyo), The National Film Archive of Japan (Tokyo), The Kobe Planet Film Archive (Kobe), the Minpaku National Museum of Ethnology, and collectors Natsuki Matsumoto, Tsuyoshi Yamabata, Machiko Kusahara, and Kogan Ashiya for allowing us to scan their collections.

WORKS CITED

Fukushima, Kanako. The Mixture of Prewar Visual Culture: Magic lanterns, Toy Films, and Miniature Films. Kyoto: Shibunkaku, 2022.

Inaba, Chiyo. “Paper Film and the Hand-cranked Projector: Animation and movies of 70 years ago,” Nihon hitokata gengu gakkai-shi, 21, 2010: 91 - 99.

Itakura, Fumiaki. “History and Preservation of Paper Film in Japan,” The 63rd Congress of the International Federation of Film Archives in Tokyo, 2007, Tokyo: National Museum of Modern Art, 2012: 83.

Orbaugh, Sharalyn. "Kamishibai and the Art of the Interval." Mechademia, vol. 7, 2012: 78-100.

[1] Sharalyn Orbaugh, "Kamishibai and the Art of the Interval." Mechademia, vol. 7, 2012: 86 - 87.↩

[2] While there is not room to discuss here, there were actually paper films before kami firumu. In the U.S. films were printed on paper frame-by-frame for copyright deposit until about 1912. Importantly, unlike kami firumu, these films were never sold to the public nor intended for projection.↩

[3] Scholar Kanako Fukushima writes about Tsukiboshi in her excellent book, Unorganized Visual Cultures before World War II : Magic Lantern, Toy Film, Cine Film. Kyoto: Shibunkaku, 2022. Researcher Tsuyoshi Yamabata (Itabashi Science Museum) has located a handful of Tsukiboshi paper film prints.↩

[4] Chiyo Inaba, “Paper Film and the Hand-cranked Projector: Animation and movies of 70 years ago,” Nihon hitokata gengu gakkai-shi, 21, 2010.↩

[5] Fumiaki Itakura, “History and Preservation of Paper Film in Japan,” The 63rd Congress of the International Federation of Film Archives in Tokyo, 2007, Tokyo: National Museum of Modern Art, 2012: 83.↩

[6] Not surprisingly, given potential copyright concerns, many live action paper films omit the celluloid film’s original title and studio provenance. Thus, tracing the paper films to their original source material involves significant sleuthing.↩

[7] Kyōrinrin appears as a kind of dragon made of strips of paper and his purpose and appearance seemed an appropriate mascot for our project.↩