Abstract: This article focuses on the operation of sound in the split-screen movie. It concentrates, in particular, on instances where the storytelling function of sound is accompanied by the aural exploration of the split screen as a specific spatial form. Different relationships between the soundtrack and multiple frames are demonstrated through examples from The Thomas Crown Affair, The Boston Strangler and Timecode.
Introduction: The Single Frame of Film Sound Analysis
Sound can define sections of the screen, guide our attention to or between them.
– Richard Barsam, Looking at MoviesDiegetic sound can be either onscreen or offscreen, depending on whether its source is within the frame or outside the frame.
– David Bordwell and Kristin Thompson, Film Art.Visual and auditory cues are what define our perception of space and its various aspects… the parameters of the space that convey these cues through sound include frequency range (pitch and timbre), intensity, and reflections (echo and reverberation, or reverb).
– David Sonnenschein, Sound Design.Notice how the car [in the opening sequence of Touch of Evil] will veer out of frame while the camera will go somewhere else; and then it will come back into frame again. That happens three times, and each time, in the new version, the re-entry of the car is preceded by the tune that was on the radio when the ignition was started, which now allows the audience to track the progress of this potentially fatal object – the car with the bomb. – Walter Murch, “Touch of Silence”
The first two quotations cited above are taken from introductory film studies books; the last two are statements by film sound designers. They are all bound by an interest in sound as a means to establish cinematic space, and this places them within a larger literature concerned with the connection between film sound and film space. Whether sound is seen to direct attention towards, or provide a sense of spatial volume for, onscreen elements, or to encourage perception of action that has extended beyond the limits of the frame, all four writers characterise the soundtrack as something that helps to define the physical dimensions of the screen environment.
The quotations from Film Art and Walter Murch, in particular, are exemplary of the tendency to assume that this screen environment has at its centre a defined frame, characterised by the four physical borders of the overall screen (left, right, top and bottom) and by an illusory depth within the image itself. In all four cases, the larger pieces of writing from which these quotations are extracted admit no variation on the assumption that cinematic sound/space relationships are constructed in the context of the single film frame.
The dimensions of the frame may change in terms of aspect ratio, and new practices of sound design may develop in response to this, as John Belton has noted (Belton 1992, 323-331). Advances in soundtrack technology such as Dolby Stereo and Dynamic Digital Sound also allow sound to circulate within and around the frame in new ways. However, the conceptualisation of narrative film as a single-frame phenomenon runs throughout writing on film sound. An indication of the endurance of this tradition is provided by comparing a foundational piece of criticism on film sound with a more contemporary one. In his Theory of the Film, published in 1952, Bela Belazs evokes the single frame as a relatively “tight” space to which the soundtrack refers:
What is not within the film frame cannot be seen by us, even if it is immediately beside the things that are. Light or shadow can be thrown into the picture from outside and the outline of a shadow can betray to the spectator what is outside the frame but still in the same sector of space, although the picture will show only a shadow. In sound things are different. An acoustic environment inevitably encroaches on the close-up shot and what we hear in this case is not a shadow or a beam of light, but the sounds themselves, which can always be heard throughout the whole space of the picture, however small a section of that space is included in the close-up. Sounds cannot be blocked out (1952, 211).
Balazs characterises the image as a clearly bordered space and sound as something that has the capacity to suggest a field beyond that space. In his 2006 article, “Narration in the Cinema of Digital Sound”, Mark Kerins presents a surprisingly similar view of sound/space relationships in film:
In the visual realm the DSC [Digital Sound Cinema] aesthetic relies on quick cutting from one close shot to another; in the aural one it creates complex, quickly moving soundscapes and places the audience in the middle of them (2006, 52).
The concentration on the close shot in both pieces provides a useful point of comparison, but is idiosyncratic (clearly, not all writing on film sound is about the close-up). However, the “reduction” of the image to its most intimate unit in these quotations is useful in indicating the general tendency to conceive of the image as a singular bounded entity and the soundtrack as something that is more expansive, in one sense, but that is also ultimately constrained by its inextricable relationship with the bordered visual space.[1]
Even Michel Chion, a sound theorist who builds his observations on idiosyncratic examples, adopts the critical orthodoxy of understanding film space as a single-frame entity, whilst at the same time alluding to other possibilities:
What is specific to film is that it has just one place for image, as opposed to video installations, slide shows, Sound and Light shows, and other multimedia genres, which can have several. This fact, and no other, accounts for why we speak of the image in the singular … The occasional experiment with multiscreen cinema – Abel Gance’s Napoleon, Michael Wadleigh’s Woodstock, or even Paul Morissey’s Forty Deuce – have not spawned many descendants and as exceptions they prove the rule of the classical frame (1994, 67).
The “exception” of multiscreen cinema is moved to the centre of enquiry in this article, with a specific focus on the operation of sound in the split-screen movie. Whilst the action of the films I discuss does develop within one overarching single screen environment, this becomes the container of frames-within-the-larger-frame, each with its own set of borders to which the soundtrack may refer.
Just as work on film sound has ignored the aural peculiarities of multiple frame cinema, so the limited body of writing on split-screen movies has remained largely silent on the matter of sound. Instead, there has been an understandable concentration on the visual challenge of the split screen to dominant filmic conventions. As Aylish Wood suggests in her book Digital Encounters, the movie screen in the split-screen film becomes self-consciously registered as a technological interface that can host a plethora of visual information from a number of sources, rather than presented as a neutral “window” on a film’s fictional world (Wood 2007, 71-104).
As well as encouraging a general self-consciousness about the screen’s formal qualities, the proliferation of images within the same frame offers unique opportunities for self-conscious comparison and contrast and, potentially, makes the viewer work harder to decipher the relationship between the different visual elements, and to understand this data in narrative terms. The surface complexity of these films’ images encourages critical commentary that focuses on their unique visual characteristics.
When sound in the split-screen movie has been discussed, it is usually ascribed a somewhat monotone function. For example, Urs Hofer, in a recent PhD on split-screen movies, claims:
[a] maxim for many split-screen films is: multiple images – one soundtrack. While the ‘point of view’ is multi-perspectival, the ‘point of hearing’ remains a one-dimensional combination of the different diegetic sound sources, or moves completely to the non-diegetic (2008, 46; my translation).
Two commentaries on Timecode (Mike Figgis, 2000), a split-screen movie discussed in more detail later, replicate Hofer’s distinction between a multi-perspectival image track, on the one hand, and one-dimensional soundtrack, on the other. Both Constantine Verevis and Allan Cameron claim that the sound mixing for Timecode is focused solely upon communicating salient bits of narrative information, through the technique of honing in on the sound of one quadrant at any one time (Verevis 2005, 173; Cameron 2008, 163-164). The soundtrack of the split-screen movie, when discussed at all, is viewed as something that helps to organise the visual material, to impose a dramatic order on images that might otherwise resist narrative explanation.
The concentration on sound as a narrative “clarifier” in the split-screen movie is unnecessarily restricting and, even as a concept in itself, requires further nuance. For example, isolating sound around one of the split-screen frames is only one way of clarifying story material. It is equally possible that the soundtrack may act as a kind of “sonic glue”, suggesting commonalities between all the disparate multiple images. There are also a number of possible uses of sound in the split-screen movie whose function might appear more formal than narrative. For instance, aural material related to each of the multi-frame images may be heard simultaneously, thereby producing a cacophonous soundscape that is of interest for its acoustic properties, but that lacks clarity in storytelling terms. Alternatively, the soundtrack may work according to the same principle of asynchronicity evident in an image track composed of several competing frames.
This article retains an interest in the storytelling potential of the soundtrack in multiple frame cinema, and, as such, does not deal with examples where the soundtrack and images are completely asynchronous (as is the case in many passages of Jean-Luc Godard’s Numero Deux (1975), for example). However, it does explore instances where the narrative function of sound is accompanied by an aural exploration of the split screen as a particular spatial form. In short, this article probes what might constitute a specifically “split-sound” aesthetic in the split-screen movie. Different relationships between the soundtrack and multiple frames will be demonstrated through examples from The Thomas Crown Affair (Norman Jewison, 1968), The Boston Strangler (Richard Fleischer, 1968) and Timecode. The analyses are sequenced to suggest a movement from a split-screen sound aesthetic that works according to “single-frame” principles through to examples that explore the specific spatial dimensions of a multiscreen environment. The interpretations of the extracts have been organized, therefore, to indicate progressive levels of intensity in the aural mapping of multiframe cinema, from a minimally exploratory sound aesthetic to one that is much more extensive in its probing of split-screen space. The conclusion relates the analyses to a model of aural perspective proposed by Theo van Leeuwen (van Leeuwen 1999), using his categories of aural figure, ground and field to make a claim for the uniquely multi-dimensional properties of the split-screen soundtrack.
The Thomas Crown Affair: “Reducing” the Multiscreen Effect Through Sound
The Thomas Crown Affair was amongst a number of Hollywood movies in the late Sixties and early Seventies that made quite extensive use of the split screen (other examples include Grand Prix (John Frankenheimer, 1966), The Boston Strangler (discussed below), Monte Carlo or Bust (Ken Annakin, 1968), Airport (George Seaton, 1969), The Ballad of Cable Hogue (Sam Peckinpah, 1969), Woodstock (Michael Wadleigh, 1970) and The Andromeda Strain (Robert Wise, 1971)). However, a split-screen sequence from early on in the film exhibits the film’s tendency to fill the screen with multiple frames only as a prelude to the reorganisation of the image into a conventional single frame. Even in the moment where multiple frames do potentially vie for the viewer’s attention, the soundtrack ensures that a focus is retained on one clear narrative line, presented as being completely under the control of the title character. The multiscreen sequence also finds a single reference point in the film’s lead, Steve McQueen, with split-screen techniques used to bolster his presentation as star.
Crown is a wealthy financier who is moonlighting as the criminal mastermind behind a bank heist, carried out by unconnected operatives. The sequence under discussion occurs about ten minutes into the film, lasts under a minute, and features Crown phoning four of the operatives from his office, to issue them the command to “go”. By this point, the film has already shown each of the operatives arriving at their assigned phone booths and Crown has already made a number of preliminary calls to two of them, to tell them to keep their positions. The spatial and narrative relationships between the characters have, therefore, already been carefully mapped out (in single frame and brief examples of split-screen) before the film enters into a more sustained passage of multiscreen.
The split-screen sequence is initiated in the middle of a phone conversation between Crown and an operative, the change in image structure signalled by an initial blacking out of the screen, when Crown swivels on his office chair, so that the back of it blocks the camera’s view. During this second of blank screen, Crown is heard saying “go”, whereupon the multiscreen image is revealed, composed of six different frames of varying dimensions, with black space between each frame (Fig. 1).
Fig 1:The Thomas Crown Affair (source: The Thomas Crown Affair DVD, MGM Home Entertainment, 2004)
Initially, all frames but the top right one are filled with an indistinct blur. In the top right frame, the operative smiles at Crown’s order, and says “I’m on my way”. He exits the phone booth, leaving a dramatically “dead” space on the screen, which is, just momentarily, devoid of action overall. However, the narrative impetus is very quickly re-established, as a dissolve within the top right frame reveals Crown in a left side profile shot, telling the next operative to “go”. The top right frame then dissolves into the same blur as the other frames, but, at the same time, the long frame on the left of the screen dissolves into a medium side shot of the second operative leaving his booth on Crown’s command. Through a further dissolve, this frame now plays host to another close profile shot of Crown on the phone, this time viewing his right side. Adhering to a now established pattern, Crown says “go”, “his” frame dissolving into a blur, whilst, simultaneously, the top frame just to the right of this longer frame reveals operative number three leaving his phone booth, the camera this time following him as he gets into his car.
Once the operative is ensconced in his vehicle, and the camera in this frame has ceased its movement, the adjacent frame to the right reveals a close-up of Crown’s eyes, which is quickly magnified so that it cuts across both this frame and the top right one, which also accommodates a view of the phone Crown is clutching to his ear. The long frame to the left is then filled with an image of the back of the chair on which Crown is sitting, whereupon the bottom two frames, previously unused, are filled with the lower half of Crown’s face. The expulsion of the view of operative three from “his” frame, in favour of the remaining piece of Crown’s face (Fig. 2), signals the beginning of the end for the multiscreen passage, as the “jigsaw” is completed, the black spaces between the frames disappear and the screen resolves into a single-frame close-up of Crown (Fig. 3). Commanding the whole screen again, Crown looks nervously excited as he issues his final instruction of “go”, which, the next single-frame shot reveals, triggers operative number four into action.
Fig 2: The Thomas Crown Affair (source: The Thomas Crown Affair DVD, MGM Home Entertainment, 2004)
Fig 3: The Thomas Crown Affair (source: The Thomas Crown Affair DVD, MGM Home Entertainment, 2004)
Visually, the sequence follows a single line of enquiry, despite the potential of the multiscreen frame to follow multiple planes of action. It does this in two ways. Firstly, for the main part of the sequence, dramatic action is confined to one frame only, giving the viewer little option but to concentrate on that section of the screen. Secondly, the passage is punctuated by a choreographed display of Steve McQueen’s star face, which grows more expansive as the sequence progresses. It firstly presents his face in profile in a small frame on the right of the screen. It then stages the pleasurable spectacle of this profile being “thrown” to the other side of the screen, flipped over and magnified. Finally, it offers the novelty of seeing a star face come to dominate the screen piece by piece as the passage ends. In this moment, then, Crown/McQueen are valorised in their roles as chief fictional protagonist and charismatic movie star.
The focusing of the scene around Crown’s narrative directions and McQueen’s charismatic star identity is reinforced through sound. After the first operative puts down his phone, non-diegetic music is heard throughout the passage, featuring driving strings and bursts of brass that lend an urgency to the scene as the heist is set into motion. This reinforces the sense that attention should be focused on the spectacle of the robbery being expertly pieced together by its mastermind.
The music always fades out, however, to allow Crown’s instruction of “go” to be heard without competing soundtrack elements. On a narrative level, the singling out of the word “go” reinforces Crown as the key causal agent within the film’s fictional world (in this world, what he says goes). Its repetition and its insertion between musical beats also gives it a rhythmic quality which adds spectacular value to McQueen’s aural star performance as well as his visual display. Furthermore, the “special” status bestowed upon the command is reinforced by a visual progression in the images with which it is choreographed. The sequence moves tantalisingly from a disembodied “go” at its beginning, played over black just before the split-screen passage begins, through two rhyming multiscreen views of the star issuing the same command, to the moment where the single frame is restored and star body and voice can be seen and heard at full power.
Arguably, the four different treatments of the delivery of the word “go” do involve the multiscreen within a moment of “parametric narration” not normally associated with the Hollywood mainstream. As David Bordwell explains:
Parametric narration establishes a distinctive intrinsic norm, often involving an unusually limited range of stylistic options. It develops this norm in additive fashion. Style thus enters into shifting relations, dominant or subordinate, with the syuzhet [plot]. The spectator is cued to construct a prominent stylistic norm, recognizing style as motivated neither realistically nor compositionally nor transtextually. The viewer must also form assumptions and hypotheses about the stylistic development of the film (1985, pp288-289).
Here, a limited range of visual options (black screen, multiscreen, single frame) are “tested” against the same material: Crown’s delivery of the word “go”. Yet the sense of the sequence offering an exercise in stylistic experimentation is reduced by the strong narrative motivation for the issuing of the command (the operatives need to be set in motion for the central bank heist to take place) and, more importantly, because the end point of the experiment is the reconfiguration of star body and voice in a conventional visual setting. In this context, the multiscreen becomes a lesser partner in a hierarchy that bestows most expressive value to the single frame.
The Boston Strangler: The Screen-Splitting Soundtrack
Unlike The Thomas Crown Affair, which was originally released with a monophonic soundtrack, The Boston Strangler utilised the Westrex 4-track stereo recording system, and the sequence to be discussed uses simple stereo effects to suggest something of the specific dimensions of the multiscreen frame. Contrary to The Thomas Crown Affair, it demonstrates a reciprocal relationship between sound and image that protects the specificity of the split-screen space, using sound to reinforce the difference between its split-screen frames, and making use of split screens to amplify the effect of the soundtrack. It registers an escalating interest in the expressive possibilities of multiscreen as the sequence proceeds, rather than the movement towards single-frame aesthetics demonstrated by the example from The Thomas Crown Affair.
The film, based on a true story, follows the police investigation into the murders of serial killer Anthony deSalvo, convicted of killing thirteen women between 1962 and 1964. It deploys split screens for a number of purposes: to depict multiple details of the killings and the crime scene whilst withholding the identity of the killer (played by Tony Curtis, whose face is only revealed an hour into the film); to provide an idea of the scale of the public reaction to the killings and the extensive nature of the investigation into it; to offer a sense, through the cutting up of the female body by the split screen, of the sadistic voyeuristic impulses that lie behind both the killings and the activities of other male suspects; and to create suspense for the viewer, by splitting the screen between the activities of the killer on the one hand and the simultaneous actions of his unwitting prey on the other.
This latter function is evident, with a twist, in the sequence I wish to discuss. Early in the film, two elderly women discover the body of their fellow lodger Myra Edwards, who has been killed by deSalvo in her bed. The sequence moves into multiscreen mode self-consciously, as a single-frame pan, following a postman turning off the pavement towards a house entrance, cuts itself in half by retracting from the left side of the screen (so that, momentarily, the left side is filled only by black space). There follows a high angled shot from the top of the stairs inside the house, still only occupying the right side of the screen. This shows the postman entering the hall and putting down the mail, which is picked up and carried upstairs by a “Mr Harris”. Once Harris has reached the top of the stairs, a third shot shows him dropping the mail off at three doors on the same level before disappearing up a further flight of stairs. In the same shot, two women come out of their rooms to collect their mail, the first from front of frame left, the second from back centre, and indulge in neighbourly chatter. The conversation turns towards the whereabouts of the third occupant on their level, Myra Edwards, who has not appeared to collect her mail. When Harris, who passes by them once more on the way down to the hall again, fails to shed any light on the situation, the woman at the back of the frame buzzes on Myra’s door, to no response. Discovering her door to be unlocked, the two women venture inside Myra’s room, which is where this shot ends.
The frame occupying the left side of the screen is activated towards the start of the shot just described. As the first woman comes out of her room, the light lifts in the left-hand frame slightly, to reveal, albeit still very gloomily, a pair of sprawled legs, motionless on a bed (Fig. 4). This mysterious and chilling image remains throughout the women’s conversation, with light (literally) only thrown on its specifics, in terms of graphic detail and location, when the two women open the door and see Myra’s body.
Fig 4: The Boston Strangler (source: The Boston Strangler DVD, 20th Century Fox Home Entertainment, 2007)
This moment of revelation occurs in both frames simultaneously (Fig. 5). The right frame shows the two women in medium close-up, without the body at the front of the frame, whilst the left shows them entering the room from the position established from its start, so the lower half of Myra’s body remains in view. The scene ends with one of the women gasping and the other screaming twice for Mr Harris.
Fig 5: The Boston Strangler (source: The Boston Strangler DVD, 20th Century Fox Home Entertainment, 2007)
Sound in this sequence moves from banal everyday conversation to a collection of screams, and the multiscreen helps to both add tension to the dialogue in the opening part of the sequence and to reinforce the transition to overt horror at the end. The impact of the screams is intensified by being choreographed with the doubling of visual information that occurs when both frames represent the same event. Whereas the final command of “go” in The Thomas Crown Affair is “amplified” by the re-emergence of the single frame (and the “unfragmented” star face), The Boston Strangler achieves a similar sense of dramatic amplification whilst still retaining a distinction between simultaneously presented frames.
When the door to Myra’s room is opened, the horror that had lurked silently in the left frame of this sequence infiltrates the right frame, which, up until this point, had represented the undisturbed unfolding of the everyday. The distinction between the sinister (in the left hand frame, appropriately) and the banal (in the right) is maintained until the end of the sequence by the use of stereo effects. From the very first shot of the postman, the sound of someone practising the piano is audible in the right speaker. When the action settles onto the lengthy shot of the women talking in the hallway, with dialogue centred, the piano continues in the right channel, so that we assume that the music is coming from a room upstairs to the right.
When one of the women rings Myra’s buzzer, which is positioned on the left of the corridor, the sound is registered in the left stereo channel. This signals a focusing of interest away from the everyday and on to the horrific events that have been silently chronicled through the emergence of the second frame to the left (once noticed, the silence of the left frame provides a disquieting undertone for the conversation in itself). The piano playing ceases in time for the second press on the buzzer, thereby signalling the suspension of normal activity absolutely. This prefigures the visual “invasion” of the right frame by the left in the shot that shows the women opening the door, the creak of which occupies the left channel, thereby underlining the movement from one atmosphere to another.
The effect of the sequence, then, is based, in part, on a soundtrack that recognises the two different frames on display, distinguishing between their atmospheres and supporting a movement from one mode – the women’s unwitting continuation of everyday activity – to another, the revelation of a horrific act, conducted within their midst. The “joining” of the two spaces does not result in the re-emergence of a single frame, but rather is registered through a multiscreen-specific doubling of the women’s horrified reaction.
Timecode: Testing the Limits of the Multiscreen Space
Timecode incorporates the sound/space relationships evident in the examples from The Thomas Crown Affair and The Boston Strangler. At different points it uses sound to bring unity to its disparate images, like The Thomas Crown Affair, and to assert the independence of the multiple frames, like The Boston Strangler. It also involves sound in a more formal exploration of the dimensions of the split screen, with the overall effect that the film acts as a compendium of the relationships that might exist between sound and space in multiframe cinema, at least when these elements are working harmoniously.
Timecode divides the screen into four quadrants for its duration, with the action in each quarter filmed in one take and simultaneously with the other segments. It takes as its central location the offices of an LA film production company, which provides the setting for a number of personal conflicts. The sequence under discussion occurs two-thirds into the film and features the following characters: in the top-left quadrant sits Lauren, the partner of aspiring actress Rose, who occupies the bottom right quadrant, and who is being led to an audition room by film director Lester Moore. Lauren has just discovered that Rose is having an affair with Alex, the head of the production company, who can be seen in the bottom left quadrant, in the midst of a nervous breakdown. This quadrant is shared by his colleague Rene, who is trying to console him. Meanwhile, in the top right quadrant, Alex’s newly estranged wife, Emma, accompanied by her actress friend Cherine, bumps into filmmaker Ana, her musician boyfriend and her agent, as they head to the production offices to make a pitch for a new type of digital split-screen movie to Alex and his fellow executives. As this synopsis demonstrates, the film is full of “story”, and sound does work to organise this narrative – to straighten the film out, so to speak. However, the soundtrack also exhibits a split-screen specific multi-dimensionality that works alongside its more conventional storytelling duties.
Within a very short timeframe, the sequence moves between four different phases, each of which demonstrates a different place for sound in the split-screen movie. The first phase sees sound operating as a binding agent for the four quadrants, thus unifying on-screen action in the same way as the music did in the example from The Thomas Crown Affair. This unifying effect is achieved by the non-diegetic musical score, which had been playing unobtrusively before this scene, rising to the fore, signalled by the introduction of electronically treated drums, strings soaring more dramatically than previously, and a wailing female vocal. For the first twenty seconds of this sequence, diegetic sound is either silenced altogether, in the bottom quadrants, or all but inaudible, in the top quadrants.
This has the effect of offering the viewer a moment of narrative contemplation. This sense is exacerbated by the musical score rising in tandem with the renewed movement of Rose, bottom right, who is in the process of relocating to the audition room. This “functional” movement, between two different dramatic spaces, is used as a moment of pause to allow the viewer to consider the state of play between characters at this point. In particular, the viewer may notice that Lauren and Rose now occupy non-adjacent quadrants, as do Alex and his wife Emma (by non-adjacent I mean the quadrants do not share a full border with each other, although, obviously, all the frames meet at a point in the middle). This suggests visually the estrangement that each couple now faces. The commonalities between the situations are emphasised through a musical soundtrack that does not “belong” to a particular quadrant and through the subduing of diegetic sound that would emphasise the independent life of each quarter of the screen.
So, at the start of the extract, the music brings together disparate elements of the drama in a manner that is not particularly split-screen specific: music works in this way to bring together characters and storylines in the montage sequence of non-split-screen movies, for example. However, twenty seconds into the extract, competing elements re-emerge on the soundtrack, thereby changing its relationship to image and narrative. In this second phase, whilst the music does continue to be heard, this is overlaid by a succession of sound bites from three of the four quadrants: bottom left, Rene offers to call Alex a cab, reassuring him that she is there for him, whilst Alex apologises miserably; this is swiftly superseded by sound from the bottom right quadrant, with Rose’s response to another aspiring actress, “yes honey I’m the bitch”, attaining a privileged position on the soundtrack. This is followed immediately by the filmmaker Ana, in the top right quarter, asking her boyfriend for his cigarette: “give me a drag, Joey”.
Two things are being achieved by the soundtrack here. Firstly, sound navigates the viewer’s attention from one story situation to another. This differs from the binding function of the music at the start of the extract, in that it characterises each quadrant as an entity in its own right, with its own soundtrack. However, it still provides points of connection between the quadrants by representing them as environments containing stories that are entwined, as the momentary intermingling of sounds from the two bottom quadrants makes explicit, bringing together as they do the two characters who are having an affair with each other.
As well as continuing to organise the material narratively, the sound at this point works more formally to highlight the sonic specificities of the split-screen image. It does this by choosing to relay between the sounds in a recognisable visual pattern: anti-clockwise from bottom left, through bottom right to top right. This functions as a partial sonic circling around the dimensions of the split screen. There is also an acceleration in the sonic relay that registers an excitement in being able to tour around this particular screen environment. The relatively lengthy passage of dialogue between Rene and Alex slides into the shorter exchange between the annoyed auditionee and Rose, which gives way, this time without any bridging effect, to the even shorter utterance of Ana. The dialogue moves more quickly between frames as it proceeds, and by so doing, draws attention to the fact that the image offers a proliferation of frames into which the sound may migrate. There is a formal play involved in the movement of sound here, which is only enabled by the peculiar characteristics of the split screen. At the same time, this formal play does not replace the organisational qualities of sound in terms of narrative: sound is still giving us insights into the film’s characters, by characterising, through dialogue, Alex’s hopelessness, Rose’s ambition and Ana’s jealousy.
The third phase of the extract, lasting another twenty seconds, sees diegetic sound being formed into a limited collage effect. After Ana asks her boyfriend for a cigarette, the chatter around them in this quadrant can still be heard, even as there is an aural refocusing towards the bottom left quarter, where Rene tells Alex she will order him a cab (and he continues to wallow). The sounds from both these quadrants then momentarily compete with each other, to the extent that precise dialogue cannot be deciphered. Whilst the bottom right quadrant is mainly silenced for this phase, there is one brief moment when this soundtrack is added to the mix, as the director says something to Rose on the way to the audition room – the sound from this quarter is then phased out again momentarily.
This relative lack of discrimination between sounds is brief but audible, and hints at a soundtrack that is ready to abandon its organisational role in terms of narrative. If the relay of voices in the second phase of the extract focuses on the movement between split-screen spaces, the collage effect gives the split screen a specific sense of overall solidness. However, the very brief phasing in and out of sound from the bottom right quadrant reminds the viewer that this solidness is constituted from separate frames, each with a soundtrack that can be manipulated separately as well as existing as part of a wider whole. Again, sound is functioning to characterise the split screen as a specific type of space.
The collage effect might be seen to threaten narrative intelligibility and the sudden return to more streamlined sound in the fourth and final phase of the sequence can be viewed, in part, as a return to the dramatic situation. There is an ironic visual allusion to the need to redirect the viewer to the story, as the establishing of the film director as the aural focus of interest comes just as he reaches the on-screen microphone in the audition room. It is as if the director has contrived to find his way to a single source of sonic amplification in order to save the film from being engulfed by an incomprehensible proliferation of voices from its disparate corners.
The sense of sounds being singled out, rather than allowed to intermingle, is also underlined by the very abrupt termination of the music, and the appearance of a strong, lone off-screen voice, whose owner, a visiting masseur, is only revealed subsequently through a camera movement. But viewing this moment only as a return to narrative order through sound misses out on some of its nuances. For one thing, the fact that the voice comes from off-screen means that the expansiveness of the split screen is still self-consciously being referred to: not only is this a film which features four frames for the price of one, each of these frames contains off-screen space too! Of course, nearly all films use sound to indicate off-screen space, but this moment is unusual in terms of Timecode’s visual aesthetic: hitherto the camera movement has been planned precisely to capture characters’ actions in “real time”. In this context, it is surprising that the camera “gets caught out” by the masseur’s vocal intervention. The existence of multiple frames is registered precisely because, just for a moment, they are shown to be fallible in their failure to survey a significant element of the film’s world.
So, a return to narrative intelligibility is accompanied by a further probing of the split-screen’s formal capacities. The moment also registers the dimensional qualities of sound in a split-screen environment in another way. Although the masseur’s is the only voice that is heard at the end of the extract, it is not the only sound. The abrupt end of the music is also followed immediately by the clink of a glass hitting a bin in the bottom left quadrant. The cessation of the music, together with the abandonment of the collage effect also draws attention to sound from the top left quadrant, which has been characterised by silence up to this point. With other sounds subdued, we can now hear clearly Lauren’s tapping of a bottle cap on the window frame of the car door, a detail into which the camera has just zoomed. At the same time as the film seems to be making things simple again by focusing on one voice in one quadrant, it is also stretching our attention to the furthest corners of its split-screen environment: from a voice off-screen bottom right, to a tap on the far side of the quadrant top left. Once again, direction towards significant narrative events is accompanied by direction towards the specific physical properties of the split-screen environment – where sounds from very different spaces are available to be isolated, relayed or intermingled in an extremely flexible manner.
Conclusion
What might account for the different intensities with which each sequence explores the dimensions of the split screen through sound? One potential enabling factor, kept to the margins of the argument thus far, is the technological means the filmmakers have at their disposal to construct their soundtracks. Timecode makes use of Dolby 5.1 and Sony Digital Dynamic Sound and this certainly enables a polyphony of sounds to be mixed with a precision and clarity far superior to the monophonic The Thomas Crown Affair or to the primitive stereo effects of The Boston Strangler. In addition, these new sound technologies allow for a spatial movement of sound around the screen that was simply impossible in earlier film sound practice.
However, it would be a mistake to cast the development of sound technology as all-determining to the shape of split-screen/sound relationships. The key differences between the sequences’ use of sound would still stand, even if different technologies had been involved in the soundtrack’s production. Imagine, for example, the extract from The Thomas Crown Affair benefiting from recourse to digital surround sound. This would allow each of Crown’s instructions of “go” to be heard in spatial correspondence to the image of Crown’s mouth which, as previously described, moves between different split-screen frames. It would also allow his voice to be heard across all the available audio channels when Crown’s face is made whole at the end of the split-screen sequence. These effects would contribute to, rather than undermine, the sequence’s organisation around its protagonist and star, its casting of the split screen as a novel detour on the road to the reaffirmation of the single frame as the default big-screen deliverer of charismatic star performances and narrative information.
Similarly, the basic division between the two sides of the screen in The Boston Strangler could be maintained in mono or digital surround. Digital sound technology would have allowed the filmmakers to combine a richer palette of sounds and make wider use of the stereo spectrum, but, as with The Thomas Crown Affair, this would have amplified the original sound/space relationships, rather than changing them. Without stereo effects, the sonic distinction between left and right frames would be inaudible, in a literal sense, but would still be suggested to the film audience through sound’s implied relationship with concurrent visual and narrative information. The left side of the frame is characterised by near darkness and a disturbing lack of bodily movement; the right side by everyday conversation and activity. In these circumstances, it would be counter-intuitive to associate the unvisualised sound – the piano playing – with anything other than the right side of the frame.
Finally, the complex sound design of Timecode would not have been possible to achieve in exactly the same way without digital technology, but the different ways of organising sound identified in my analysis – “centralised” (through the use of non-diegetic music), relayed, formed into a collage and “flung” across the split-screen environment – would be attainable using any recording system that allowed multi-tracking. The relay of voices from bottom left quadrant to top right, for example, is achieved as much by raising and lowering the volume from one quadrant to another as it is by the spatial movement of sound around the speakers. Even without the physical movement of sound, the film audience can appreciate the relay effect due to the synchronisation of particular sounds with their logical visual counterparts in a specific quadrant. The crucial point of distinction between the sequences, then, is not the sound technology itself, but rather the way that technology is used to construct a soundtrack’s implied relationship with visual and narrative material.
Theo van Leeuwen’s Speech, Music, Sound provides a useful model for understanding the parameters of the implied relationship between sound and the split-screen environment. Its subject is sound as a widespread social practice and, as such, it avoids the narrow focus on sound as a single-frame phenomenon that dominates film sound studies. In the chapter on sound perspective, van Leeuwen identifies the positions of figure, ground and field as crucial to the creation of aural perspective. He states that “the significance of these positions can be glossed as follows”:
Figure
If a sound or group of sounds is positioned as Figure, it is thereby treated as the most important sound, the sound which the listener must identify with, and/or react to and/or act upon.Ground
If a sound or group of sounds is positioned as Ground, it is thereby treated as still part of the listener’s social world, but only in a minor and less involved way. We are to treat it as we treat the familiar faces we see every day and the familiar places we move through every day, in other words, as a context we take for granted and only notice when it is not there any longer.Field
If a sound or group of sounds is positioned as Field, it is thereby treated as existing, not in the listener’s social, but in his or her physical world. We are to treat it as we would treat the people that crowd the streets through which we walk, or the trees that populate the forest past which we drive (van Leeuwen 1999, 23).
A split-screen environment provides a very specific context in which sounds are configured in the arrangements of figure, ground and field. Each individual frame offers a distinct visual and narrative setting within which aural perspective may be created, with the “raw material” from which figure/ground/field relationships are construed often differing significantly from frame to frame. In The Thomas Crown Affair, the “denial” of split-screen specificity is achieved, in part, by a refusal to explore aural perspective within the sequence: the sound is almost exclusively “figure”, either in the shape of the music or Crown’s voice. In The Boston Strangler, there is a contrast between a frame whose sound world is composed of figure (the neighbour’s chatter) and ground (the piano playing) and a frame whose soundtrack is – uncannily – all field (the silence of Myra’s room). The two worlds are brought together when “figure” – the women’s gasps and screams – dominates both frames and each is robbed of its wider sonic context (the “ground” disappears to the right when the piano playing stops; the “field” of silence disappears to the left with the women’s cries). In Timecode, what constitutes figure, ground and field is in a constant state of flux as “environmental” sounds, such as the clink of a glass being thrown into the bin or a bottle top being tapped on a car door, are brought to the fore, and voices – normally showcased as figure in film – are made to mingle as “ground”. Far from existing in the “one-dimensional” form assumed by Hofer in his comments on split-screen sound, the application of van Leeuwen’s terms allows the soundtrack to be understood as a provider of aural perspective, which has the potential to adopt as many multiple shifting forms as the multiframe image itself.
References
Balazs, Bela. 1952. Theory of the Film: Character and Growth of a New Art. London: Dennis Dobson.
Barsam, Richard. 2007. Looking at Movies: An Introduction to Film, 2nd ed. New York: W.W. Norton and Company.
Belton, John. 1992. Technology and Aesthetics of Film Sound. In Film Theory and Criticism: Introductory Readings, eds. Gerald Mast, Marshall Cohen and Leo Baudry, 4th ed., 323-331. New York: Oxford University Press.
Bordwell, David. 1985. Narration in the Fiction Film. Madison, WI: University of Wisconsin Press.
Bordwell, David and Kristin Thompson. 2004. Film Art: An Introduction, 7th ed. London: MacGraw-Hill.
Cameron, Allan. 2008. Modular Narratives in Contemporary Cinema. Basingstoke: Palgrave Macmillan.
Chion, Michel. 1994. Audio-Vision: Sound on Screen. New York: Columbia University Press.
Hofer, Urs. 2008. Split Screens: Prolegomena zu einer Analyse des geteilten Filmbildes. PhD diss., University of Zurich. http://www.lulu.com/content/913777.
Kerins, Mark. 2006. Narration in the Cinema of Digital Sound. The Velvet Light Trap, 58 (1): 41-54.
van Leeuwen, Theo. 1999. Speech, Music, Sound. Basingstoke: Macmillan Press.
Murch, Walter. 2003. Touch of Silence. In Soundscape: The School of Sound Lectures 1998-2001, eds. Larry Sider, Diane Freeman and Jerry Sider, 83-102. London: Wallflower Press.
Sonnenschein, David. 2001. Sound Design: The Expressive Power of Music, Voice, and Sound Effects in Cinema. Michigan: McNaughton and Gunn.
Verevis, Constantine. 2005. Time Code and the Screen. In New Punk Cinema, ed. Nicholas Rombes, 168-179. Edinburgh: Edinburgh University Press.
Weis, Elisabeth and John Belton, eds. 1985. Film Sound: Theory and Practice. New York: Columbia University Press.
Wood, Aylish. 2007. Digital Encounters. London: Routledge.
Notes
[1] The pervasiveness of the ‘single-frame assumption’ in film sound writing is usefully illustrated by following the numerous references under ‘sound and space’ in the index for the seminal film sound collection, Weis and Belton (1985).
Author Bio
Ian Garwood is a lecturer in Film and Television Studies at University of Glasgow. He has written a number of articles on the film soundtrack, including the book-length “The Pop Song in Film”, included in Close-Up: 01, eds. John Gibbs and Douglas Pye (London: Wallflower Press, 2005).
Contact Email: