Making "The Inferno Unseen"

How one of the most tantalizing unfinished film projects became a live showcase of the studio tests for Henri-Georges Clouzot’s "Inferno".
Marketa Uhlirova


The Inferno Unseen

Henri-Georges Clouzot’s Inferno (1964) is one of the most tantalizing unfinished projects in cinema history. If completed, it would have told a story of extreme jealousy and obsession. The plot is simple—a hotel owner, Marcel (Serge Reggiani), begins to suffer from nightmarish visions in which his young wife Odette (Romy Schneider) appears in various lascivious poses and sometimes erotically interacts with another man. Marcel gradually descends into madness and may, in the end, be driven to kill his wife. Generously backed by Columbia (via the French production company Orsay Films), Clouzot shot in black and white as well as in color, employing three separate film crews, no less than 12 cameras, and a large number of technicians and film craftsmen, including some of the most established industry names of the time. For six months, three cameramen—Claude Renoir, Armand Thirard and Andréas Winding—shot seemingly endless studio tests and, in the final three weeks, also some scenes on location. Many of these were shot repetitively, in subtle variations across numerous takes, resulting in the 12 hours of footage that survive today and that contain no dialogue or sound. Clouzot’s extravagant process reveals an unusual sense of freedom from budgetary constraints that would normally dictate strict planning, discipline and structure. Fueled by an ambition to elevate cinema to the status of avant-garde art (a concern that is again clearly manifest in Clouzot’s later film La prisonnière, 1968), the making of Inferno turned into a prolonged search for novel expressions in filmmaking—a search that was as ambitious and intense as it was often quite hellish and confusing for those involved. The film was never finished; it was shelved when Clouzot suffered from a heart attack (and though this could on the face of it be read as a cause-and-effect situation, it has also been suggested that both were deeply interrelated and perhaps even inevitable in equal measure).

The footage forms several distinct groups: an extensive set of wardrobe tests for the lead actors Schneider and Reggiani, and the key supporting actors Dany Carrel and Jean-Claude Bercq, which are largely in black and white, and range from full-body shots to close-ups. There are also several prolonged fabric and surface tests with grey-scale and color charts juxtaposed against various backgrounds, almost completely static. But ultimately, it is the kinetic and optical tests—Marcel’s delirious hallucinations and the effects they have on his psyche—that make up the bulk of the studio-shot material. They are also the most ground-breaking aspect of this project.

These ‘hallucinations’ display a remarkable range of visual experimentation. There are abstract explorations of changing shapes, colors and patterns, or dynamic phenomena (bubbling liquids, smoke), informed by contemporary artistic movements such as Kinetic art and Op Art, while also belonging to an older tradition of interwar avant-garde cinema. There are also numerous eye-popping visions which treat the human face as raw photographic material to be manipulated—sometimes to the point of complete abstraction—through distortions, splicing, double exposures, mirroring, or kaleidoscopic multiplications. But what is singular in Clouzot’s experiments is the extensive application of fluid light and color projections onto the human face and body and costume, as well as their various psychedelic distortions. Here the filmic and the sartorial unite to form striking visual effects, at once beautiful and haunting, that become a somewhat improbable missing link between earlier (as well as contemporary) formal experiments with light and film by the likes of John and James Whitney and Jordan Belson on the one hand, and Andy Warhol’s later Exploding Plastic Inevitable projections on the other.


The Inferno Unseen

My own interest in this film began in 2008 when I was spending time at Lobster Films researching early dance, trick and féerie films for a forthcoming Fashion in Film Festival season. At the time, Lobster’s Serge Bromberg was digitizing the discovered footage of Inferno, some of which would be used in an experimental documentary he was planning to produce (the film, Henri-Georges Clouzot's Inferno, would be released a year later, in 2009 and would win numerous prestigious awards). Over coffee Bromberg told me a highly entertaining story of how he got to work with this material that many before him had unsuccessfully attempted to get access to. Like the others before him, Bromberg was persistently nagging Henri-Georges Clouzot’s widow Inès to let him see the footage and, as with the others, the answer was always a resolute no. However, after much seemingly futile effort, he did eventually manage to ‘seal the deal’ thanks to something that can only be called an act of divine intervention: he and Mrs. Clouzot got stuck in the lift together while he was being gracefully escorted out of her Parisian apartment. In the two hours spent in the lift Mrs. Clouzot finally gave in.

I am forever indebted to Marianne Lère, one of the producers of Bromberg’s documentary, who was then working at Lobster and who showed me a handful of the Inferno rushes, just as they had come in, in their raw and unprocessed form. She sat me down in front of a large computer screen with the words, “if you are interested in fashion, then you really should see this!” And I almost fell off my chair: the footage was utterly transfixing. Serendipitously, it seemed to me strongly aligned with my research at the time, which looked into early cinema’s interpretations of Loïe Fuller’s dance spectacles of the 1890s and 1900s, which used costume, light and colors as their primary media. I distinctly remember one of the prolonged takes of Romy Schneider facing the camera, engulfed by darkness, with her head wrapped up in cellophane like a human bouquet, a big ribbon fastened around her neck. She was being strangled by a pair of hands emerging from behind her. As her head was swung violently from side to side, a rainbow of projected colored lights danced around the shiny surface. I remember the enigmatic bride (also Schneider) slowly lifting her veil, sharpening her beautiful features hitherto softened by a layer of muslin, with colored lights and shadows again circling the face and torso. And there were other striking scenes. These moments stayed with me and have continued to reverberate, even after I got my proper ‘fix’ in the form of the 2009 documentary and the extras that accompanied its DVD release.

In 2015 I began to collaborate with the cinema historian Tom Gunning on a festival season, which explored the relations between cinema, time and fashion. The program looked at time through four distinct categories—past, present, future and dream. While Gunning mostly seemed to gravitate towards the past and the dream, I was intrigued by the prospect of investigating the present and future. As I was thinking about fashion and cinema in relation to the present, and also a sense of physical presence (for example, of a person or garment in an image), Inferno surfaced again. I had in fact first toyed with the idea of programming Andy Warhol’s Screen Tests from the 1960s, which uphold cinema as an art of duration while at the same time foregrounding the sartorial and the accidental. In many ways, Clouzot’s cinematic tests invite a comparison with Warhol’s, and it is pertinent that they were also created at exactly the same time. Although Clouzot’s intentions and working methods were very different (if only in not being conceived as stand-alone pieces), and his subjects were instructed differently, he too was clearly fascinated by the potential of a screen test—an otherwise rudimentary studio device used to determine an actor’s photogenic appeal—as a study of ephemeral and often involuntary gestures, with a unique, almost unsettling temporality. In focusing on moments of holding a pose, rather than acting, he too was exposing the person behind the actor. Though here of course the actor and their dress were treated as a screen on which to project colored lights, shadows and patterns.

Gunning and I wanted a live element in our festival program, so I decided to go back to Bromberg to see if he would allow me to revisit Inferno for the purpose of staging a live show. He agreed, which I saw—and still see—as a magnanimous gesture. I then met with Kiri Inglis at MUBI who was interested in getting MUBI to stage more unique live shows as an alternative to online streaming. Together we began to define the project in more concrete terms. We initially wanted a rather abstract showcase of Clouzot’s rushes, focusing, within the remit of our festival, on the shots led by clothes, textiles and make up. I proposed we departed from the narrative put forward in the documentary, creating a new experience that would deliberately steer away from explaining or contextualizing the footage. Perhaps we were lucky the story had already been told and we could now lavish on the footage, allowing it to claim its own presence, to play out in all its mystery and magnificence. Inglis brought in the electronic musician Rollo Smallcombe as someone who could write a new score while simultaneously editing the footage. We also agreed to include short excerpts of Bromberg’s voice (which we had newly recorded in collaboration with the Monocle podcast); we wanted to use them as minimal signposts—or you could say clues—throughout the film. Finally, we set a basic parameter that we would not manipulate any of the footage through added visual effects (not that this was ever needed!) or fades (this rule was broken but only on a handful of occasions).

Smallcombe approached the project with such precision that the idea of a live, improvised mix went immediately out of the window. As he later commented, the footage “came like a pack of cards that had been very well shuffled,” and I suspect he made it his mission from the beginning to make sense of these cards by putting them into some sort of ‘right’ order. Working together on this was as thrilling as it was delicate, and it certainly made me think in a new way about the boundaries, and the slippages, between curatorial and artistic engagement. Crudely speaking, Smallcombe was the artist and I was the curator—at least such are the labels we normally wear. But this divide quickly dissolved as we began to have intense creative tussles about the different visions we had for this project. Instinctively, Smallcombe was veering towards telling a story, and just as instinctively I was pushing against this. In retrospect, we both tried to make manifest different aspects implicit to the footage that spoke to us.

We started by pre-selecting some three hours of footage. It was vital that we both watched all the material sent over by Lobster because this gave us a sense of the nature of what we were dealing with. It made us see how obsessively scenes were shot and re-shot, and it gave us a feeling for the poetry and rawness of found material. It also highlighted the major themes that repeated over and over, across scenes and across reels. We were surprised how many bizarre discoveries were taking us away from the shots of Schneider, for which the footage is now notorious; the breadth and complexity of the material was truly astonishing. Smallcombe then organized the selections by creating a meticulous cataloguing system with dozens of visual themes, many cross-referenced under multiple categories. And it was in this process of categorizing that a sense of a narrative began to emerge for him.

My own position was that the cut must not look like a coherent edited film, that we have to let the footage just be, uncut, true to the archival state in which it was found. I pressed Smallcombe to show more of the mundane, let the repetitions show, keep the takes longer, let them become uncomfortable to watch... I suppose part of my anxiety was that in an overly polished cut, we were in danger of somehow reconstructing Clouzot’s original story and his intentions, or creating a long-form trailer for a film that never happened. In the end I think we achieved an ‘uncut’ feel, though with a lot of behind-the-scenes chiseling, some of which was also music-led. We stopped when it felt right: at some point we both realized that with the random and sometimes inexplicable material there is, creating order out of chaos was nigh on impossible, that our principal achievement would be in raising more questions than we could possibly have answers to. The film we created is not a reconstruction but rather an invocation. It does at times allude to what may have been, and I have no doubt now that it is better for it. Perhaps it even approximates a sense of frenzy and madness that Clouzot himself intended. In that sense, it is a tribute to his directorial vision and his crew’s extraordinary undertaking.

Don't miss our latest features and interviews.

Sign up for the Notebook Weekly Edit newsletter.


Henri-Georges Clouzot
Please sign up to add a new comment.


Notebook is a daily, international film publication. Our mission is to guide film lovers searching, lost or adrift in an overwhelming sea of content. We offer text, images, sounds and video as critical maps, passways and illuminations to the worlds of contemporary and classic film. Notebook is a MUBI publication.


If you're interested in contributing to Notebook, please see our pitching guidelines. For all other inquiries, contact the editorial team.