Ruivo, Céline. "le Livre De Fabrication De La Compagnie Générale Des Phonographes Cinématographes Et Appareils De Précision" - The Pathé Frères Laboratory Logbook Of Joinville-le-pont (1906–8)

Le Livre de fabrication de la compagnie générale des phonographes cinématographes et appareils de précision: The Pathé Frères Laboratory Logbook of Joinville-Le-Pont (1906–8) Author(s): Céline Ruivo Source: The Moving Image: The Journal of the Association of Moving Image Archivists , Vol. 15, No. 1 (Spring 2015), pp. 85-92 Published by: University of Minnesota Press Stable URL: https://www.jstor.org/stable/10.5749/movingimage.15.1.0085 JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at https://about.jstor.org/terms

University of Minnesota Press is collaborating with JSTOR to digitize, preserve and extend access to The Moving Image: The Journal of the Association of Moving Image Archivists

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

UMforumforum Le Livre de fabrication de la compagnie générale des phonographes cinématographes et appareils de précision

The Pathé Frères Laboratory Logbook of Joinville-Le-Pont (1906–8) C é line R uivo In 1905, the Pathé Frères Company opened a second laboratory in Joinville-Le-Pont to expand the activities of its Vincennes operation, where the tinting and stenciling of film prints were part of the daily work.1 Thierry Lefebvre’s study of the Joinville factory shows that 265 prints were manufactured every day.2 Until 1910, only the Vincennes laboratory was processing camera negatives. Le Livre de fabrication de la compagnie générale des phonographes cinématographes et appareils de précision is a five-hundred-page unpublished logbook written by the Joinville engineers in the research and development department between June 16, 1906, and September 3, 1908.3 The department’s main researcher was Clément Lair, a former student of the French scientist Pierre Curie who would much later become president of Kodak-Pathé in 1939, following the earlier merger of the two companies in 1927.4 The logbook has been preserved by the Kodak France Archives, which created the Cercle des Conservateurs de l’Image Latente (CECIL) to manage Kodak’s French archival holdings. The logbook can be con-

sulted at the Fondation Jérôme Seydoux-Pathé (http://fondation-jeromeseydoux-pathe.com). Le Livre de fabrication combines a series of reports written by Pathé’s engineers, some of which were addressed directly to Charles Pathé, codirector of the company. Along with detailed accounts of the chemical analysis of color, the logbook describes a variety of other technical developments, including a new machine for print processing, the creation of film titles, comparison tests between different types of film stock (Blair, Eastman, Lumière, etc.), experiments in stereoscopy and with nonflammable stocks, and notes on increasing the properties of toning baths. The logbook lists each day’s experiments with meticulous entries on their results and the overall progress of the lab’s various technical innovations. Each of the daily entries has subtopics, which allow the reader to easily analyze the results, especially concerning research on color. Although the logbook touches upon many aspects of research and development at the Joinville factory, we focus here primarily on the techniques of coloring during these early years, as well as on the diversity of the tools and methods used, including the hand painting of the positives with brushes, stenciling, tinting, and toning baths. One of the most interesting things that we learn from the logbook is that Pathé was already experimenting with an additive coloring system as early as 1907. On the basis of these various entries, we demonstrate how Le Livre de fabrication is important for the history of color, particularly as Pathé was already established at the time as one of the leaders of color cinema. From the logbook, it is clear that between 1906 and 1908, Pathé was innovating its methods and styles of coloring, looking for more

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

f o r u m 86

natural and subtle coloring techniques for its films. At that time the stakes were high both economically and technologically. Color film sold much better abroad, and Charles Pathé wanted to beat the competition in this field by having the most advanced technology possible. New competitors, such as Éclair in 1907, made for an increasingly competitive marketplace. At the same time, Gaumont’s Chronophone system, which brought effective sync sound to the screens, helped to push Pathé’s business model toward increased specialization, resulting in its growing expertise in color, as reflected in the logbook. We should also remember that Charles Pathé was actively seeking out inventors and technicians to support the company’s continued effort at patenting new mechanical devices that would reduce costs and the timeconsuming and uneven technical labor required in the laboratory.5 For instance, in 1907, he ordered a stencil machine with a cutting stylus attached to a pantograph from inventor Jean Méry to help cut holes in stencil reels. The project for this patent is described in 1907 in the logbook, including a drawing most likely by Méry, but without attribution to him, along with a second device with a mechanical brush used for applying the dyes to the positives (Figure 1). Another interesting point about Le Livre de fabrication is that it contains a few references to the American Pathé factory in Bound Brook, New Jersey. The factory opened in 1907 in the former World’s Dispensary Medical Association building, using French workers recruited from other Pathé factories, and it developed the company’s prints for the American market.6 Although the New Jersey factory is not known to have been involved in the research and development of coloring methods, the logbook reveals that it did send color formulas back to the Pathé laboratory in Joinville. How the Pathé factory in Bound Brook came into possession of these formulas, though, remains unknown.

THE COLORS OF THE TONING PROCESS Before examining the logbook in detail, it is useful to understand the practice of the toning process at the turn of the last century. Initially, it could be described as an almost artisanal practice, with prints varying dramatically, depending on the weekly mixture of chemicals

and the grading processes selected by the individual technicians. According to Joshua Yumibe, it was Henry V. Hopwood in his book Living Pictures (1899) who made a reference to some of the earliest experiments of moving picture toning, but the actual first toned moving picture still remains impossible to identify. At Pathé, a dedicated department focused on quality control and innovation and industrialization in the field of toning was not founded until the opening of Joinville in 1905, though the company was toning prints in a more artisanal fashion before this. The toning process is distinguished from tinting in that it involves the coloration of the dark shadows of images, that is to say, the unexposed parts. Toning was used in silent cinema to color the black-and-white positive print. The process can be found much earlier in nineteenth-century still photography, such as in photographs created by John Herschel in 1840 using the “cyanotype” process, where a Prussian blue replaced the dark shadows of the photograph. The silver salts are first bleached and converted into silver ferrocyanide. A second bath is necessary to replace the salts with different types of metals. Typically, copper ferrocyanide is used for the sepia tone. Iron ferrocyanide can usually create a blue tone, or Prussian blue. Uranium was used for the red tone, which J. C. Burnett experimented with for the first time in 1859.7 Vanadium could re-create a yellow-green tone: the color was discovered by Rodolfo Namias in 1901, even though this element had already been tested before by the Lumière Brothers for its reactivity to light.8 All these chemical components were dangerous for the workers of the Pathé laboratory because of the high level of toxicity found in the compounds, though the writers of the logbook never refer to these hazards.

THE INDUSTRIALIZATION OF THE TONING PROCESS Charles Pathé started the industrial refinement of the toning process in 1906 with the opening of the Joinville facility. The main goal was to find the best manufacturing methods to increase production by improving the workflow and technical standardization of practices. We know that Charles Pathé would later reveal in

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

f o r u m 87

Figure 1. Drawing of the stencil machine (pochoirs), May 20, 1908. Excerpt from Livre de fabrication de la compagnie générale des phonographes cinématographes et appareils de précision, livre de fabrication no. 1, 467. Courtesy of Association CECIL and Fondation Jérôme Seydoux-Pathé.

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

f o r u m 88

his autobiography his deep attachment to Frederick Winslow Taylor’s scientific management theories, but it is not yet known if between 1906 and 1908 he was already aware of Taylor’s work. However, the logbook does contain related directives on how to improve the efficiency of the workers, as coloring productivity had to attain a high standard of quality to surpass the competition. This is also why the engineers carried out so many quality control tests, and it also provides a sense of the industrial leadership in color that Pathé was building over the years in the early twentieth century.

THE PRINTING AND PROCESSING OF TONED POSITIVES Until now, the absence of primary technical descriptions has made it difficult to understand the interaction of the processing and the printing of positives that were to be toned, but Pathé’s logbook sheds useful light on this early technical history. In the logbook, as early as 1906, a Pathé engineer named Dubois explained that print processing needed to be slower to increase the contrast for toning.9 Although the standardization of the processing was quite problematic in Joinville, the negatives processed by Vincennes also created a few problems that Dubois mentioned. A few titles, such as Tour du monde, Steeple Chase, Odyssée d’un paysan à la ville (1905), and Pantalon, had poor camera exposure during filming, which necessitated a fifteen-minute increase in the processing, according to Dubois. Dubois also frequently complained about the disorder of the processing, calling for increased regulation of the workers in charge of the development, and specifically criticized their practice of adding more developer and water into the baths whenever they wanted, which made each tray very different from the others with regard to chemical composition.10 Laboratory printing methods in 1906 were approximate, as the grading was done by eye and by adjusting the distance of the bulb in the printer from the film being printed to modify the contrast, a much cruder method than later practices of modifying bulb intensity through voltage. To increase the sharpness of the image, potassium bromide was added to the processing bath, a common practice in still

photography. With increased contrast in the toning process, the brightness and density of the color in the print emerge, bringing a more colorful image to life. The logbook offers film restorers important insight into print density for toned materials, especially in instances where positive reference prints no longer exist, as is the case for the Cinémathèque Française, which holds the largest collection of original Pathé negatives in the world, allowing it to reproduce new prints representative of the original effort. The writers of the logbook regularly explain that toning was time consuming, given the extensive chemical manipulation involved, but it was still cost effective because of the metals employed. The trays used for the baths had to be switched from lead to wood to avoid reactions of the chemical components, a problem encountered mostly with vanadium (for yellow-green tones). Moreover, this simple change in trays made the toning baths much more efficient in terms of longevity. The logbook documents, through almost daily entries, Pathé’s dedication to understanding and refining the toning process by recording the engineers’ adjustment of colors, including specific measurements of the chemical components, as well as the laboratory’s understanding of the longevity of processing chemicals when mass-producing multiple prints. The following topic headings correspond to titles from logbook entries and illustrate developments in quality control and the refinement of Pathé’s various color processes.

The Sepia Tone As early as 1906, the lab researchers wanted to reinforce the sepia density of the positives. They experimented with different formulas, including one with “urane” (uranium) to make the sepia stronger, as the strength of the uranium element was problematic and had a tendency to wash out all the grays and middle tones. It was only in July 1908 that Pathé’s engineers received a new formula from the American Pathé factory in Bound Brook that had been used at least during that year: potassium ferricyanide (800 g), potassium bromide (300 g), muriatic (hydrochloric) acid (200 g), and water (120 l). The tones could change according to the amount of muriatic acid added to the formula.11

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

f o r u m 89

The Brown Tone The brown tone was different from sepia, and Pathé marketed it separately, as can be seen in the frames available in the company’s tinting guide for producers, Le Film vierge Pathé (1926).12 In 1906, one engineer described the process in the following terms: “You first need to convert the silver into an halogenous derivative.”13 The conversion was done with potassium bichromate, which had a brownish property because it came initially from a redorange crystal; this was then transformed into sulfide with sodium sulfide.

The Green Tone The instability of the vanadium bath is described throughout the logbook as one of the most problematic issues concerning color rendering. This information is quite important because we have noticed that green toning did not survive as well as other tones, and it is quite rare to find in film archives. It could be possible that the production of the tone was lower precisely because of the short-term bath longevity and the cost of the vanadium. Even when the team tried to replace vanadium with cobalt, an element usually used for coloring glass an emerald hue, nothing could produce a strong green as well as vanadium.

The Blue Tone Pathé modified its blue tone in October 1907: “We wanted to change the iron oxalate and replace it with iron chloride, which is much cheaper.”14 Aesthetically, the result was quite satisfactory, as the images gained more transparency while the blue looked resplendent. The cost was reduced by one-third—five francs and thirty cents per tray instead of eighteen francs—and the bath could take sixty film processing roller carriers before the chemicals weakened.15

THE COMBINATION OF TINTING AND TONING In July 1906, the Pathé team experimented with its first tinting and toning combination on the title Les Rapides de la rivière Ozu (André Legrand, Pathé no. 1493, 1906). A sepia toning was combined with blue tinting, and the result was quite successful. The engineer

Dubois noted that it was difficult to achieve perfect regularity on such combined colors.16

Multiple Color Toning An excerpt of the logbook shows that highly experimental tests on a multiple toning process were carried out in 1908 using a double-sided color print. Clément Lair describes a positive with two emulsions, and consequently the colors were applied on both sides of the positive. Lair refers to an existing photographic process, the carbon process, to create this new type of positive.17 The carbon process implies the reactivity of chrome to light, which was discovered by Louis Nicolas Vauquelin in 1798.18 Based on Vauquelin’s research, Alphonse Poitevin discovered the actual carbon process in 1855. He applied a mix of potassium bromide and pigments coming from coal powder onto a piece of paper. After drying the emulsion, Poitevin exposed the paper to sunlight in contact with a negative, and the image was then revealed in a hot water bath. Later, John Pouncy discovered this process separately in 1863. It was in 1864 that Joseph Swan presented the various possibilities of carbon printing to the Photographic Society of Edinburgh.19 Carbon printing was combined with the photographic toning process at the end of the nineteenth century. Regarding Clément Lair’s research in Joinville, he describes in the logbook a new concept of combining toning using the gum bichromate in the gelatin. He wanted to make a double emulsion positive with silver bromide on one side, which was sensitive to light, and a gelatin with pigment on the other side using the carbon process (Figure 2).20 In 1908, Lair wanted to patent this process and asked for the help of the Pathé factory in England to create the double emulsion positive, which would, according to him, help reduce expenses because they could use less metal for the toning process. Unfortunately, Pathé abandoned his groundbreaking research.21

Three-Color Process Multiple references to a system of three-color still photography appear in the logbook in June 1906. Several tests were made with a nine by twelve centimeter optic chamber with three glass plates that were placed in front of three lenses with separate color filters. The

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

f o r u m 90

projections of engraved photography were also done with carbon arc, and the glass plates were sensitized before the projection, referred to in the logbook as the “pinachrom” technique, used in an early experiment in June 1906.22 The process was useful in increasing the sensitivity of the red glass plate and used leuco dyes without ammonia, some of which were colorless and could be transformed into colorful dyes by ultraviolet radiation. Commercially introduced by the German company Meister, Lucius, and Brüning in 1904,23 the process was abandoned because it did not produce stable results in light due to oxidization. The test described in the notebook with pinachrom introduced irregularities into the color rendition process and was discarded. Dubois started his research later in 1908, on the pinatype color process, while Lair was doing research on the carbon process. The pinatype process was quite different from the carbon process because it used dyes that were absorbed by an emulsion that had reliefs on it. This color process was discovered by the French photographer Charles Cros, who patented it under the name Hydrotype. Another French inventor, Léon Didier, improved the process and sold it to Meister, Lucius, and Brüning under the name Pinatype in 1903.24 The components of certain dyes were more appropriate for the Pinatype system. Finally, they were simply called the pinatypes or the pina dyes and were produced by the I. G. Farben factory, a German

Figure 2. Drawing of a twoemulsion positive, January 10, 1908, by Clément Lair. Excerpt from Livre de fabrication de la compagnie générale des phonographes cinématographes et appareils de précision, livre de fabrication no. 1, 385. Courtesy of Association CECIL and Fondation Jérôme Seydoux-Pathé.

industrial conglomerate that included BASF and Agfa. A pina manual was also published by Agfa. Unfortunately, the research on the pinatypes was not successful at Pathé Frères. In April 1908, Dubois experimented with a combination of stencils and pinatype coloration on one print, which, according to him, was beautiful for a still projection.25 Meanwhile, Dubois pursued research on a three-color additive system in 1908. It included three negatives exposed through three color filters: red, yellow, and blue (or violet). When the negatives were processed, the complementary colors appeared on the positives. Dubois wondered about how he could make a movie camera that could separate the three colors, and Dubois thought about a rotating shuttle that could be placed in front of the lens. The three colors did not work effectively with this system, so Dubois tried it with a two-color system, which was easier to control. Given the timing, it is clear that the British invention of Kinemacolor, a two-color additive

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

f o r u m 91

process that was distributed by Charles Urban in France as early as 1908, influenced Dubois’s research on natural colors, even though Dubois never mentioned Kinemacolor in his writing.26 Also the three-color additive system Dubois described resembles the Chronochrome process, produced by Gaumont in 1911. Dubois was unsuccessful in the experiments he conducted and recognized at the end of the logbook that finally it was much easier to color the positives with stencils—especially since by then they had the new stencil machines innovated from Méry’s designs. Another solution proposed was to produce a system combining stenciling and the three-color process.27 However, the logbook does not list any hybrid films produced that combined these two methods, even though a few test prints were most likely made, given Dubois’s proposals. The Cinémathèque Française has three experimental color cameras manufactured by Pathé (undated, though seemingly from this period), and each had three lenses attached vertically to the cameras, but none of these prototypes was put into production (Plate 10). Even though various publications, such as E. J. Wall’s The History of Three-Color Photography (1925), demonstrated that three-color additive technology did not start with Gaumont’s Chronochrome, nor did the concept of three separated negatives begin with Technicolor. Wall’s account shows how important the research on natural color processes had become for Pathé, the company already recognized for its stenciling, tinting, and toning techniques. Moreover, Wall reports a patent presented by Pathé for new a prism block in 1913,28 two years after the first projection of a Chronochrome film, which shows that Pathé wanted to enter the three-color market as well. Did World War I stop that introduction? It is difficult to know, but it is a distinct possibility. The logbook also raises the question of intellectual property, because many ideas were not original and were in varying stages of development based on photographic patents already in existence. In addition, Clément Lair’s research on a cemented two-color positive can be connected with later patents, such as for the Kodachrome I (1916) and the Technicolor process II MultipleAdditive. One of the problems that film archives

encounter today is finding and identifying the surviving elements and prototypes that were involved in this quest for natural colors in films. The identification of older color technologies needs more than ever to be documented to understand the interconnections among those technologies: cameras and optical devices, dyes, workflows, and projection devices. Through Pathé’s documentation, both of its successes and failures, we gain more than just an in-depth look at one of the greatest research laboratories of its time; the logbook also gives the reader an acute sense of the correlation between scientific research and economic development at the Pathé Frères Company.

Céline Ruivo is a doctoral student at the University of Paris III Sorbonne Nouvelle. She has been the curator of the film collection at the Cinémathèque Française since 2011. Notes

The author would like to thank Stéphanie Salmon from the Fondation Jérôme Seydoux-Pathé and Laurent Mannoni from the Cinémathèque Française for their contributions. 1. In 1906, Pathé expanded its laboratory in Vincennes, located on rue du Polygone, with an additional building dedicated to stenciling located on rue des Vignerons. 2. Thierry Lefebvre, “L’Usine de Joinville, établissement dangereux,” in La Firme Pathé Frères 1896–1914, ed. Michel Marie and Laurent Le Forestier (Paris: AFRHC, 2004), 52. 3. Le Livre de fabrication de la compagnie générale des phonographes cinématographes et appareils de précision, livre de fabrication no. 1 (Paris: Pathé Frères, 1906–8). 4. Michel Remond, Histoire d’une Aventure: Kodak Pathé Vincennes 1896–1927–1986 (Vincennes, France: Kodak-Pathé, 1986), 151. 5. “Modifications à la machine à colorier, à velours,” in Le Livre de fabrication (1908), 470. 6. Stéphanie Salmon, Pathé à la Conquête du Cinéma 1896–1929 (Paris: Tallandier, 2014), 156–57. 7. Joseph Friedman, History of Color Photography (Boston: American, 1945), 304–9. 8. Auguste Lumière and Louis Lumière, “Sur

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms

f o r u m 92

les propriétés photographiques des sels de vanadium,” in Paris-Photographe, revue mensuelle illustrée, 4ème année (March 30, 1894), 121–22. 9. Reytinas, “Tirage et développement des films projetés en couleurs,” in Le Livre de fabrication (1906),1–2. 10. Dubois, “Développement,” in Le Livre de fabrication ( June 23, 1906), 8. 11. Roussel, “Formule de virage sépia,” in Le Livre de fabrication (1908), 500. 12. Le Film vierge Pathé: manuel de développement et de tirage (Paris: Les Établissements Pathé-Cinéma, 1926). 13. Jaupitre, “Essais de virage en brun de la pellicule positive,” in Le Livre de fabrication (1906), 22. 14. Zelger, “Nouveau bain de virage bleu,” in Le Livre de fabrication (1907), 315. 15. Ibid. 16. Dubois, “Virage,” in Le Livre de fabrication (1906), 52. 17. Clément Lair, “Nouveau procédé pour virage en toutes couleurs,” in Le Livre de fabrication (1908), 383–85. 18. Edward J. Wall, Carbon Printing (London: Hazel, Watson, and Viney, 1904), 5–104. 19. Ibid., 17. 20. Clément Lair, “Nouveau procédé pour virage en toutes couleurs.” 21. Ibid. 22. Ducour, “Photographie Trichrome,” in Le Livre de fabrication (June 1906), 12–13. 23. Edward J. Wall, The History of Three-Color Photography (Boston: American Photographic, 1925), 422. 24. Friedman, History of Color Photography, 462. 25. Dubois, “Essais de pinatypie,” in Le Livre de fabrication (April 1908), 451. 26. Benoît Turquety, “Le Naturel et le mécanique: le Kinémacolor à la conquête de Paris, ou Charles Urban vs. Charles Pathé,” in 1895, Revue d’histoire du cinéma, no. 71 (Winter 2013): 81–105. 27. Dubois, “Sur l’obtention d’une vision en couleurs au Cinématographe,” in Le Livre de fabrication (1908), 408–10. 28. Wall, History of Three-Color Photography, 195.

Hand-Painted Abstractions

Experimental Color in the Creation and Restoration of Ballet mécanique R ossella C atanese , G uy E dmonds , and B regt L ameris Given the canonical status in avant-garde media history of Fernand Léger and Dudley Murphy’s Ballet mécanique (1924), its preservation history is perhaps surprisingly complex. As a consequence, it has been the object of many debates on the practice and ethics of restoration and reconstruction.1 In 2011, EYE Film Institute Netherlands, in collaboration with the Haghefilm Foundation, carried out the most recent restoration work of the film and employed coloring techniques as used in the 1920s. Although the work conducted did not result in a new projection copy, the project remains a potential next step in the evolving restoration history of Ballet mécanique. Still, one wonders how to position this restoration work that did not result in an exhibition print. Can we still define this as a restoration, or are we confronted with something different and new? We will describe the history of the film and its colors in the various prints, versions, and restorations to contextualize and historicize the work. Finally, we wish to situate the project within an archival approach called experimental media archaeology that, though being outside of standard restoration practice, will allow a better understanding of the achievements of the restoration. Ballet mécanique, a seminal work of the 1920s avant-garde, explores the machine age using an experimental approach grounded in montage, camera movement, unorthodox camera angles, and animation. Some of the archival prints of the film also contain colored inserts, mainly of animated circles and triangles in different sizes, but also of other images evoking simple shapes (such as a hat, a line of bottles, or a pair of shoes) in blue, green, red, and yellow. Fernand Léger and Dudley Murphy made the film, in collaboration with Man Ray.2 According to William Moritz, the latter made all

This content downloaded from 189.54.144.168 on Thu, 28 Mar 2019 04:26:49 UTC All use subject to https://about.jstor.org/terms