CUTs 1.03: How Filmmakers Influenced Exhibition of Silent Films
Production Accounted for Exhibitor Versions
Filming and Projection Dynamics
The notion that projectionists destabilized film texts by not accurately reproducing a picture’s original taking speed relies on two erroneous assumptions. The first is that the cameraman always shot a picture at a stable and consistent rate. The second is that equivalence between pro-jection and taking speeds should ensure textual stability. Both likely stem from the popular mis-conception that all silent films were filmed at a rate of 16 fps.
Historian Kevin Brownlow has demonstrated that this is attributable to the proliferation of film projectors in the latter half of the 20th century equipped with a “silent” speed option of about 16-18 fps and a “sound” speed option of 24 fps. In fact, taking speed was variable and could differ from film to film in the pre-transi-tion period.1 The reason for this is the cameraman, like the projectionist, was a live performer. The rate at which he cranked the camera could depend on various factors, including “the quality of the lighting on the scene, the film’s sensitivity, the kind of action the camera had to record.”2
As variation was inherent to the live performance of the cameraman then, even the most accurate reproduction by the projectionist would still retain instability of speed.
A clear illustration of this can be seen in the fact that many silent pictures use multiple taking speeds. In Nosferatu (1922, dir. Murnau), for example, the taking rate of some scenes is 18 fps, while that of others is 20 fps. Salt references this inconsistency in his argument that silent films today “can only be properly shown with a projector with continuously variable speed…”3 In such a case, constantly varying the cranking rate of the projector in accordance with changes in the taking speed becomes neces-sary to ensure accurate reproduction of onscreen action and movement.
Meanwhile, consistent projection at either rate means certain segments play at nor-mal speed, while others appear in slow or fast motion, resulting in a variant distinct from the intended one. Nosferatu thus demonstrates that, in contrast to the dominant rhetoric, a standard and consistent projection rate wouldn’t necessarily ensure textual stability, while an unstable projection rate wouldn’t always result in an alternate version of the picture.
Expressive Use of Motion
One could view the inherent instability of hand cranking as evidence that the presence of different taking speeds in the same picture was never intentional. However, this runs counter to the evidence that filmmakers expressively utilized motion to generate or heighten the effects of comedic, dramatic, and action sequences.4 Stylized and unnatural movement has been seen as a filmmaking convention of silent slapstick comedies in particular.
Kracauer describes fast and slow motion effects as “technical tricks” silent comedies use to transform reality into fantasy by “exaggerating natural movements.”5 Similarly, Rob King states that synch sound led to “a regularization of tempo that mitigated against many of the undercranking effects on which silent comedy had depended.”6 Both authors clearly position silent comedy’s fantastic speed as deliberately created by the producers, with neither referencing the agency or role of exhibitors.
Partially addressing this omission is Walter Kerr’s book The Silent Clowns, which argues against the notion that silent films in general should move at a “natural” rhythm, as a degree of accelerated motion “was used in all silent film and was accepted as normal for the form.”7
Filmmakers like Keaton and Chaplin apparently realized that accelerated motion could benefit their pictures in eliciting audience laughter, inspiring them to intentionally shoot their films below the expected projection rate. Key to this was the use of musical cue sheets, which would feature instructions for projectionists regarding what speed to utilize for a given scene or sequence.8
Slap-stick comedies thus exemplify how “silent film chose, by control of the camera and through instructions to projectionists, to move at an unreal, stylized, in effect fantasized rate.”9 In this sense, exhibitors could actually fulfill filmmaker intention.
Studies of musical cue sheets by film preservationist James Card and historian Kevin Brownlow strongly support Kerr’s conclusions about accelerated motion being normalized. Both authors argue that there was no stable projection speed, with a given title’s correct rate often exceeding 16 fps.10 Card cites multiple cue sheets by Triangle Company for various 1916 pictures, including Stranded, The Captive God, and The Half-Breed, to illustrate this.11 Unfortunately, he does not provide a comparison with the taking speeds of the titles he references.
Brownlow, by contrast, does provide taking speeds, which he has estimated by running multiple pictures from 1918 through 1928 through a variable speed telecine machine called a Polygon. He then provides a chart comparing “probable” taking speeds with the recommended projection speeds in each title’s respective cue sheet.12
Figure 1.1 compiles some of Brownlow’s data in a short table to show the difference between taking and projection speeds. As this data illustrates, producers regularly intended and encouraged faster-than-natural motion during exhibition. In this sense, they supported and enabled variation, rather than accurate reproduction.
Crucially, neither Card nor Brownlow provide any evidence that projectionists actually followed the cue sheet instructions they reference as providing the actual theatrical projection speeds. Indeed, given the lack of standardization, producers had no way of guaranteeing a projectionist would abide by the instructions.
So, there could be both correspondences and differences between the speed(s) at which a silent film was shot, at which the producers wanted to show it, and at which the projectionists actually screened it. Examining the numerous ways that a pro-ducer could influence the projection speed of a given picture reveals that filmmakers could have just as much, if not more, authorial control over a specific exhibition version of a silent film than managers, sound performers, etc.
Live performances then did not make exhibitors the sole causes of textual instability, as the text was already fluid at the production and distribution stages. Certainly, instances of accelerated or slow motion could result from managers attempting to gain more screenings per day. But they could also stem from the cameraman varying his shooting speed, the director desiring to produce a comic effect, the cue sheet instructions, etc. The fact that producers and distributors sometimes supported an exhibitor alteration of a text proves they did not necessarily oppose textual variation.
Further research, in fact, evinces that they regularly developed multiple silent film variants for overseas markets, meaning variation occurred prior to the stage of exhibition. Rather than a singular pre-exhibition version then, there were actually multiple base or master prints, from which managers, musicians, projectionists, etc. could derive a unique variant for a given screening.
This article is the fourth entry in an ongoing series called Cinema’s Unstable Texts (CUTs). If you like it, then please, by all means, share it!
Want to learn more? Then be sure to click below and check out the other entries!
Kevin Brownlow, "SILENT FILMS WHAT WAS THE RIGHT SPEED?" Sight and Sound 49, no. 3 (Summer, 1980): 164.
Paolo Cherchi Usai, Silent Cinema. An Introduction (London: British Film Institute, 2000), 9.
Barry Salt, Film Style and Technology: History and Analysis (London: Starword, 2009), 175.
Leo Enticknap, Moving Image Technology: From Zoetrope to Digital (London: Wallflower, 2005), 50.
Siegfried Kracauer, Theory of Film: The Redemption of Physical Reality (New York: Oxford University Press, 1960), 88.
Rob King, “The American Slapstick Short and the Coming of Sound,” in Companion to Film Comedy, eds. Andrew Horton and Joanna E. Rapf (Somerset: Wiley, 2012), 68.
Walter Kerr, The Silent Clowns (New York: Knopf, 1975), 35. Original emphasis.
Ibid., 36.
Ibid., 37.
One should note that both Card and Brownlow presume that “proper” projection speed is always equal to the taking speed of a given silent film, whereas Kerr questions the notion of what “proper” or “correct” projection speed actually is. Thus, they both support the notion that silent film motion should be natural, which Kerr staunchly opposes. Ironically, their study of cue sheets corroborates Kerr’s argument.
James Card, Seductive Cinema: The Art of Silent Film (New York: Knopf, 1994), 54-55.
Brownlow, Sight and Sound, 165-166.