March 7, 2015 posted by

Japanese Color Animation 1907 to 1945 – Part 1



Broadly speaking, Japanese color animation is seen as a post-World War II phenomenon. Only beginning in 1958, with Tōei Dōga’s 東映動画 Hakujaden 白蛇伝 (Legend of the White Snake; US title: Panda and the Magic Serpent), directed by YABUSHITA Taiji 藪下泰司 (1903-1986)(1), could a large number of Japanese admire homegrown color animation films in cinemas (2). So it comes as no surprise that among three recent anime histories (Koyama-Richard[2010]; Tavassi[2012]; Clements[2013]) just one contains an attempt to collect information on pre-war color animation (Clements[2013], 50f.).

This study will discuss Japanese color animation before 1945 systematically, showing that there was more of it than even Clements acknowledges, although, as we will also see, by international standards Japan was a laggard in this field. I will not include the “utsushie” 写し絵and slides for Japanese magic lanterns (“gentō” 幻灯) here, because I do not regard them as animation: they did not rely on the “persistence of vision” principle, which I deem fundamental to any definition of animation. However, I will discuss, among other types of animation for the home market, the color animation filmstrip discovered by MATSUMOTO Natsuki in 2005 and provisionally called Katsudō shashin 活動写真, which very likely is the oldest surviving Japanese animation film to be projected (just not in a cinema), as well as its German counterparts by which it was likely inspired.

The following text falls into three main parts: on natural color systems, on what I will call here artificial coloring methods, and on printed color animation (including Katsudō shashin).

1. Natural color systems

Filming and projecting colors “realistically” had been the aspiration of many people right from the beginning of film itself: the first patent on a “natural color system” was filed by the German Hermann Isensee in Berlin in 1897 (McKernan[2008]). Such systems divide into two classes: additive (“optical”) and subtractive (“chemical”) ones. In the former, as in Isensee’s case, colors are created by mixing light of various wavelengths; in the latter, colors are produced by absorbing certain wavelengths using dyes or pigments (Koshofer[1988], 140, 155). For additive systems filters and special equipment were usually necessary for filming and projecting, whereas subtractive systems resulted in film that could simply be projected with standard equipment.

1.1 European and American developments

lake-gardaThe earliest natural color system to be reasonable successful and widespread was the British two-color additive “Kinemacolor”, which had been developed by George Albert Smith (1864-1959) in 1906, and which was subsequently marketed by the American Charles Urban (1867-1942) (McKernan[2013], chapter 3)(3). Kinemacolor also found its way to Japan in 1913 (4) , after Tōyō Shōkai 東洋商会, soon reorganized under the name Tennenshoku Katsudō Shashin 天然色活動写真 (“Natural Color Films”; in short, Tenkatsu 天活), obtained the East Asian rights (Komatsu[1995], 70-72)(5).

The first drawn animation in natural colors seems to have been John R. Bray’s (1879-1978) The Debut of Thomas Cat in 1920, using the two-color subtractive “Brewster Color” (Cavalier[2011], 75; Nowotny[1983], 127). And one of the most famous natural color systems, the three-color subtractive “Technicolor (Nr. 4)”, was first presented with Walt Disney’s (1901-1966) animation film Flowers and Trees in 1932 (Higgins[2007], 25-26) which soon impressed Japanese audiences, too (Honpō kiwamarinaki[1933]).(7)

flowers-treesOther countries, too, saw the development of natural color systems. In 1933 and 1934 Germans were able to view color animation films by Oskar Fischinger (1900-1967), made with the three-color subtractive Gasparcolor system (Schoemann[2003], 149-151).(9) And from 1935 on various animation films were produced in the Soviet Union using two- and three-color sutractive systems (Mayorov[2012], 247, 250-253).

1.2 Japanese natural color live-action films

It is often claimed that the first Japanese (natural) color feature film had been Karumen kokyō ni kaeru カルメン故郷に帰る (Carmen Comes Home), directed by KINOSHITA Keisuke 木下恵介 (1912-1998) in 1951 (e.g., Sharp[2011], 234). Yet it was just the first one to be made using Japanese color film, in this case by Fuji Film. (The first test of this new film type seems to have been the short, and unreleased, animation Nezumi no kentō ネズミの拳闘 (The boxing-fight of the mice), directed by MORINO Satoshi 森野佐登志 in 1948 (Animēju Henshūbu[1989], 30.))

karyumenAlready in 1937, however, at least three “natural color” films produced by Dai Nippon Tennenshoku Eiga大日本天然色映画 (“Great Japan Natural Color Film”) premiered in Japan: the first one, perhaps qualifying as a feature film, was Tsukigata Hanpeita 月形半平太 (Tsukigata Hanpeita), directed by SHIBA Seika 志波西果 (1900-1937?) (Yomiuri shinbun[1937]). Either the two-color subtractive American Cinecolor system (Sugimoto[1990], 300; see also the next section) or its precursor, the two-color subtractive Multicolor system may have been used for these films (National Film Center[2003], A27; Koshofer[1988], 152).

Of course, Tenkatsu had already made several shorter color films using Kinemacolor, starting in April 1914 with Yoshitsune Senbonzakura 義経千本桜 (Yoshitsune and the Thousand Cherry Trees), directed by YOSHINO Jirō 吉野二郎 (1881-1964) (Komatsu[1995], 73).

1.3 Japanese attempts at natural color animation: Kogane no hana

Although no Japanese natural color animation film reached cinemas before the 1950s, some attempts were made between 1929 and 1937 by well-known animator ŌFUJI Noburō 大藤信郎 (1900-1961) alone or in collaboration with him. Ōfuji, a disciple of KŌUCHI Jun’ichi 幸内純一 (1886-1970), employed various animation techniques in his career (Sano[2013]) and in 1929 produced the chiyogami 千代紙 (colored paper) cut-out animation Kogane no hana こがねの花 (The Golden Flower): not just as a “record-talkie” レコード・トーキー (a film that could be projected in synchronization with a record player), but also as a color film using a “Shinekarā” シネカラー system.(11)

Kogane no hana

Kogane no hana

However, when the film was released in 1930, it was only in monochrome and silent (Yamaguchi/Watanabe[1977], 200; Tsugata[2010]). Nevertheless, it still managed to win the “Best Film Medal” of the Japanese Ministry of Education (Sano[2013], 88).

Yet there remain doubts about the real nature of this film. First of all, no sources are given in the literature for the assertion that it had originally been made in color. Secondly, Ōfuji himself, in his 1934 article on color films, does mention, neither Kogane no hana, nor any “Shinekarā” system. On the contrary, he claims to have given up on chiyogami animation and to have turned to drawn animation because there was no way in Japan to show these films in color, as they deserved (Ōfuji[1934], 66). That Kogane no hana is found in a list appended to Ōfuji’s article under “works from the silent era”, not under “record-talkies”, and that no mention of color is made in the author’s comments on it does not help either (Saitō[1934]). Finally, it remains difficult to ascertain which “Shinekarā” system Ōfuji could have employed: the British Cinecolour system (1925-1930) needed a beam splitter camera; it is very unlikely that Ōfuji had access to such a camera. The French Lumière Cinécolor system using the Kornraster (irregular grain screen) process was introduced only in 1929, the year of Ōfuji’s film, (12) but never fulfilled its promise. And the much better known American Cinecolor system became available only in 1932, much too late for this film (Koshofer[1988], 25, 31, 42).

1.4 Japanese attempts at natural color: Sanba no chō

The problems of additive color systems were well illustrated in Ōfuji’s next, or probably first, attempt, made in collaboration with another highly important animator of the time, MASAOKA Kenzō 正岡憲三 (1898-1988). In 1934 they completed the 16mm short animation Sanba no chō 三羽の蝶 (Three butterflies), apparently with the two-color additive DuPont-Vitacolor system introduced in 1930 for amateurs. This relied on recording and projecting every picture twice in separate colors, similar to the old Kinemacolor system, so the projector needed to be able to run faster than usual – at 32 pictures per second (Koshofer[1988], 23, 145). Yet they lacked such a, foreign-made, device and thus could not project the film properly (Yamaguchi/Watanabe[1977], 213; Sugimoto[1990], 295).

Ōfuji had held high hopes earlier for this color system which he transcribed as “Baitakarā” バイタカラー, because in his opinion it did not need any special projector (he thought 24 frames/ second were enough), but was more advanced than Kinemacolor, and different from the more recent “Uaitakarā” ウ゚アイタカラー system ‒ presumably the American two-color additive Vitacolor system introduced in 1932, which also needed a beam splitter camera (Koshofer[1988], 159), though not in all points as good as Technicolor (Ōfuji[1934], 66f.).

Sugimoto Gorō also covers this film extensively ‒ he does not mention Kogane no hana, and is not aware of any Japanese “Shinekarā” animation, but argues that Masaoka and Ōfuji used a “Uaitakarā” ウ゚アイタカラー system which, by his description, might have been a different Vitacolor system introduced in 1930 (Sugimoto[1990], 290, 295, 300; Koshofer[1988], 42). However, he provides no references; moreover, according to him the film was additionally hand-colored in orange and green, leading to the question whether it should not be listed below with An Expression (2.1).

1.5 Japanese natural color animation: Katsura-hime



Ōfuji’s last experiment with color in the 1930s was in the 16mm film Shikisai manga no dekiru made 色彩漫画の出来る迄 (Making of a color manga[film]) by OGINO Shigeji 荻野茂二 (1899-1991) in 1937. (13) This was a live-action color film showing Ōfuji’s way of preparing a color animation,(14) which included within its five minutes the “experimental” drawn animation Katsurahime かつら姫 (Princess Katsura), lasting two minutes.(15)

It was recorded on (subtractive) Kodachrome film introduced in the US in 1935 for amateurs (Koshofer[1988], 150), so, while Ōfuji seems to have been quite content with the result (Ōfuji[1939]), like the other attempt(s) it was a dead end as far as cinematic release was concerned because the Kodachrome system was not available for 35mm motion picture film. However, Konica’s post-war Konicolor/Sakuracolor system was based on the Kodachrome process (Koshofer[1988], 114), and Ōfuji used this system for his remake of the silhouette animation Kujira くじら (The Whale) in 1952 (Tsugata[2010]).(16)

NEXT WEEK – PART 2: Artificial Color and Printed Color


Many people helped me in the course of this research. Special thanks to Charles Barten, DAIBO Masaki 大傍正規, INABA Chiyō 稲葉千容, INOUE Momoko 井上百子, Darren Nemeth, NISHIMURA Tomohiro西村智弘, OKADA Hidenori 岡田秀則, and most of all to MATSUMOTO Natsuki 松本夏樹. Thanks also to David Tucker for pointing out an error in a reference in the original note put on the web on 17 June 2014.

(1) – Tōei Dōga used Eastman Color, not a Japanese color system (Hu[2010], 91). Two previous, short color animation films had also used Eastman Color: Otogi Pro’s おとぎプロ Fukusuke (Fukusuke) ふくすけ, directed by Yokoyama Ryūichi 横山隆一 (1909-2001) in 1957, and Tōei Kyōiku Eigabu’s 東映教育映画部 Yumemi Dōji (Dreaming Child) 夢見童子, directed by Fukiya Kōji 蕗谷虹児 (1898-1979) in 1958 (Animēju Henshūbu[1989], 37f.).

(2) – Shortly before the premiere of Hakujaden on 22 October 1958, the made-for-TV cut-out color animation Mogura no abanchūru もぐらのアバンチュール (Mole’s adventure; directed by WASHIZUMI Hiroshi 鷲角博) was broadcast by NTV, but seen by most Japanese in black-and-white because color TVs at the time were still rare (http://ja.wikipedia.org/wiki/もぐらのアバンチュール). The 16mm film used the American Ansco Color system (Yamaguchi/Watanabe[1977], 256).

(3) This is the same Charles Urban for whose French company Eclipse Émile Cohl (1857-1938) produced Les exploits de feu follet, which was distributed internationally as The Nipper’s Transformations, and which was shown on 15 April 1912 in Tokyo as Nipparu no henkei ニッパルの変形 (Litten[2013a]).

(4) With only Albert Smith mentioned, but neither Kinemacolor nor Urban, the principles of this process had already been reported in a Japanese newspaper in 1908 (Hanashi no tane[1908]).

(5) Tenkatsu did not use Kinemacolor much after 1914 (Komatsu[1995], 75), but did, inter alia, hire Shimokawa Ōten 下川凹天 (1892-1973), also read Shimokawa Hekoten, and produce the first Japanese animation films for the cinema in 1917 (Litten[2013b]). For a contemporary Japanese note on Kinemacolor see Terasaki[1916], 101.

(6) This is a photograph of the projected film from http://www.brianpritchard.com/Kinemacolor%20Project.htm.

(7) It should be noted that Technicolor No. 4 involved printing on blanc film (Higgins[2007], 24f.).

(8) Screenshot from http://www.disneyshorts.org/shorts.aspx?shortID=166.

(9) Oskar Fischinger’s animation films were also seen in Japan in the 1930s (Nishimura[2008], 134).

(10) Screenshot taken from http://nozawa22.cocolog-nifty.com/nozawa22/2011/11/nozawa22-22.html.

(11) The film is available on the DVD Animēshon no senkusha Ōfuji Noburō ‒ Kokō no tensai, based on a 35mm b/w print (cf. National Film Center[2004], program 2).

(12) Is it co-incidence that one of Ōfuji’s films was shown in France early in 1929 (Sano[2013], 88)?

(13) On Ogino see also below and Makino[2002], 59f. Also cf. Sugimoto[1990], 296f.

(14) This corresponds closely with his introduction to animation production for readers of the Asahi shinbun in 1939 (Ōfuji[1939]).

(15) The film can be found as an extra on the DVD Animēshon no senkusha Ōfuji Noburō ‒ Kokō no tensai.

(16) The film is included in the DVD Animēshon no senkusha Ōfuji Noburō ‒ Kokō no tensai. The lost original of 1927 had been written with the Kanji 鯨. A few other Japanese animation films in the 1950s also used the Konicolor/Sakuracolor system (Sugimoto[1990], 306), but not Fukusuke (see above and Yamaguchi/Watanabe[1977], 251).

Leave a Reply

Your email address will not be published. Required fields are marked *