David Bordwell's website on cinema   click for CV




Film Art: An Introduction

Christopher Nolan: A Labyrinth of Linkages pdf online

Pandora’s Digital Box: Films, Files, and the Future of Movies pdf online

Planet Hong Kong, second edition pdf online

The Way Hollywood Tells It pdf online

Poetics of Cinema pdf online

Figures Traced In Light

Ozu and the Poetics of Cinema pdf online

Exporting Entertainment: America in the World Film Market 1907–1934 pdf online


Hou Hsiao-hsien: A new video lecture!

CinemaScope: The Modern Miracle You See Without Glasses

How Motion Pictures Became the Movies

Constructive editing in Pickpocket: A video essay


A Celestial Cinémathèque? or, Film Archives and Me: A Semi-Personal History added September 2014

Shklovsky and His “Monument to a Scientific Error”

Murder Culture: Adventures in 1940s Suspense

The Viewer’s Share: Models of Mind in Explaining Film

Common Sense + Film Theory = Common-Sense Film Theory?

Mad Detective: Doubling Down

The Classical Hollywood Cinema Twenty-Five Years Along

Nordisk and the Tableau Aesthetic

William Cameron Menzies: One Forceful, Impressive Idea

Another Shaw Production: Anamorphic Adventures in Hong Kong

Paolo Gioli’s Vertical Cinema

(Re)Discovering Charles Dekeukeleire

Doing Film History

The Hook: Scene Transitions in Classical Cinema

Anatomy of the Action Picture

Hearing Voices

Preface, Croatian edition, On the History of Film Style

Slavoj Žižek: Say Anything

Film and the Historical Return

Studying Cinema


Book Reports

Observations on film art

Archive for the 'Film archives' Category

Silent frame rates and DCP: A guest essay by Nicola Mazzanti


1914 Kinetoscope Victor Model I Silent Film Projector by the Victor Animatograph Co. From Pinterest, identified by Robert Van Dusen.

DB here:

The first thing you learn about silent movies is that they weren’t shown silent. The second thing is that they weren’t shown as fast as we often see them. Screened at 24 frames per second (“sound speed”), many films look rushed and silly. Those were shot at slower frame rates, like 16 fps or 18 fps. By the end of the 1920s, though, some were shot faster than sound, up to 30 fps.

Most archives and some cinémathèques and repertory theatres owned adjustable-rate film projectors, so they could provide a smooth flow of motion for most silent films. But what happened when film projectors were junked and replaced by digital projectors? How could the Digital Cinema Package (DCP), that set of files containing the movie and a host of encryption devices, offer the various “silent speeds”?

I wrote a little about this problem in Pandora’s Digital Box: Films, Files, and the Future of Movies, but afterwards I continued to wonder how archives were coping with the matter once digital conversion was universal. The main problem is that digital projectors don’t currently provide options slower than 24 fps. Both the Hollywood studios that established the digital standard and the manufacturers that implement it simply ignored the history of cinema. Where does that leave archives, which restore, circulate, and show silent films on digital equipment?

So I asked our friend Nicola Mazzanti, Curator of the Cinematek in Brussels, about this problem. Nicola is one of the founders of the Bologna Cinema Ritrovato festival and a major consultant about digital cinema and the preservation of films in the new age. Here’s his typical provocative, pro-active answer.


Why can’t we ever make it simple?

Nicola 400

Nicola Mazzanti, Brussels summer 2011. The plaque reads: “Mr. Bill Clinton, President of the United States of America sat at this table on 9 January 1994. He drank coffee and chatted an hour with all present.”

When they try to sell you some new product there is always the sweet thrill that now, finally, all the unpleasantness and the drawbacks, the little things that bothered you in the past, will be gone forever. Then, you realize that the car you bought was a Volkswagen diesel.

With Digital Cinema we were all told that life would be happier once we jumped onto the bandwagon. Distribution of independent films would be easier. All sorts of catalogue titles would flow joyously from the archives to the multiplex. “Alternative content,” including slightly unsharp TV programs and opera with awful close-ups and great sound, would make theatre owners rich.

Many believed the PR. Some of this became true—the usual 10% of it.

Granted, it is easier and cheaper to produce and manage a DCP than a 35mm print. And some classics did find their way to independent or repertoire theatres (the few who survived the digital switch, that is). Several archives, ours included, are enjoying these benefits of the digital switchover. And advertising an 8K or 20K or whatever-K restoration can bring in audiences. But many of us also have tons of perfectly fine 35mm of classics, and they cannot be shown any more. Why not?

*35mm projectors are basically gone.

*Some catalogue owners permit screening only their “restored DCP” (and strictly with English or French subtitles – so much for cultural diversity!).

At the end of the day the number of available classic titles went down, not up.

We could talk about this situation a long time, but let’s home in on one effect of the digital transition. The story I want to tell today is one about those “technical details” that we seem never able to get right. Only the Gods of Cinema know why.

We could talk about screen ratios in this connection, but not today. Today we talk about frame rates.


Toward a standard

Film strip

Unlike video files but like a strip of film, a DCP consists of a bunch of individual frames that play nicely one after the other. The number of frames that are played back per second is defined by a simple line of code in an XML file. From the point of view of the technology in the server and the DCP format, there is no reason why we cannot play a film at any rate we like, say, 17fps. (Granted, we can’t really run it at 100fps, because of bandwidth, but you see my point.)

This is not the whole story. First of all, there is a need for a highly standardized system, or we would have the problem of a DCP playing well in one theater and at the wrong speed in the next. Then there are other issues, like the frequency of the projector or the link between the server and the projector. I will not bore you with the details. In sum, technical problems make it impossible to build what I had proposed in the first place: a hand-cranked digital projector. (I was serious.)

Yet all these technical problems could be solved, if we took a rather pragmatic approach. (Here, “pragmatic” is the PC term for “compromise.”) That’s what we archivists did, a few years ago.

At the time of publication, the Digital Cinema Initiative specs described something that did not really exist yet. The technology was only almost there. In the hurry to produce a standard that could work, it was perfectly sensible to start with only two frame rates: 24 and 48 (for 3D).

Some time later though, the issue of expanding the frame rates options came back with a vengeance. The strongest push was coming from European broadcasters who wanted to sneak their boring stuff to the cinema screens and thus called loudly for a 25fps frame rate. Then some filmmakers like Peter Jackson and James Cameron fell in love with the idea of higher rates, like 30fps and 60 for 3D (regardless of how this would balloon production budgets).

Soon the issue became geopolitical, with the Europeans pushing hard for 25fps, mostly because they considered it a victory to impose another standard on the Americans—even though they had lost the digital war as a whole. At the end, the Pandora’s box of frame rates had to be reopened, enthusiastically by some, reluctantly by others.

The Society of Motion Picture and Television Engineers (SMPTE) set up a workgroup where all strong views in favor or against were represented. For the historians among you, it was first called “DC28-10 Ad-hoc Group on Additional Frame Rates” and later “21DC.10 AHG Additional Frame Rates.” We archivists, following our typical model of guerrilla resistance, sneaked onto the panel.

Technically, the archival force was the Technical Commission of the International Federation of Film Archives (FIAF). Actually, it was Torkell Sætervadet and I who were volunteered for this “mission impossible.” Torkell is the world’s greatest expert on projection. He wrote the indispensable handbook for film projection, The Advanced Projection Manual (2006). His FIAF Digital Projection Guide, another must-read, was published in December 2012, the year film projection disappeared.

Torkell and I worked on the frame-rate problem for a couple of years. We mostly revised documents and conducted conference calls late at night (thanks to European time zones). We received some support from the more corporate members of the Workgroup. Many had never heard of lower frame rates, and the concept actually fascinated them. We also benefited from the support of Kommer Kleijn from IMAGO (the European version of the ASC), Paul Collard (then Ascent Media, now Deluxe, a profoundly competent person in the business), and Peter Wilson of High Definition and Digital Cinema Ltd.

The result: We succeeded to a great extent. We convinced the workgroup that silent rates had to be part of digital projection. True, the new standard allows only four other rates: 16, 18, 20, or 22. (To be extra-precise, due to technical issues, 18fps is in fact 18.18181818 and 22 is really 21.8181818.) But four options are quite a lot, and we decided that four were better than none.

So what we called the Archive Frame Rates standard (technically ST 428-21:2011) became part of the SMPTE standards for Digital Cinema. It can be found here. More or less in parallel, other frame rates were allowed: 25, 30, and 60fps. An announcement of this development, with commentary, can be found here.

Hurray! But don’t celebrate yet.

The Archival Frame Rates standard is not a mandatory one. The manufacturers of Digital Cinema projectors never implemented this particular standard in their machines. And archives all over the world never requested it before purchasing a projector. That, today, is the roadblock to proper screening of silent films on adjustable digital projection.


Home-made silent speed

Well, what are the alternatives?

One can produce, for example, a normal DCP at 24fps, and then use a simple piece of software (such as EasyDCP) that allows for different speeds if the DCP is played back from a computer. That feeds directly into the projector, bypassing the projector’s server. It works fine, although you have to try out several frame rates until you find the one that does sit well with the refresh rate of your projector. Otherwise, you might have some artifacts onscreen.

Alternatively, if your film’s chosen silent speed is 16fps you can easily triplicate each frame and create a 48fps DCP which would play in most theatres. 20fps would fit into 60fps, when available. For 18 and 22, however, this is no solution; they are tough frame rates anyway.

The problem with these as hoc solutions is that they are not universal. If an archive ships a DCP to another archive, someone acquainted with that file must be in the booth with a computer. Or you can just send the DCP and hope it works. Either of the two digital solutions can be used safely only in your own theatre, basically.

Hence my interest in pushing for a standardized solution. This is, I’m convinced, the best one for bringing silent films to broader public.

So I haven’t given up hope that one of the manufacturers might realize that the Archival Frame Rates standard would be a nice addition to their option package. Particularly if archives and other potential buyers start asking for it.


Does one speed fit all?

The silent frame rates are more complicated than what I’ve just outlined. Let’s go a little deeper, because it’s interesting.

When we refer to frame rates, we tend to use discrete numbers (16, 18…). But like all things analog, frame rates lay along a continuum. That ranged from 16 (sometimes lower) to 24 (sometimes higher). Anybody who worked as a projectionist knows that even with a modern projector equipped with speed variators, the actual speed can change according to the day of the week and the time of the day, depending on fluctuations on the grid.

The same variations occurred, naturally, with hand-cranked projectors and cameras. And early motor-driven machines used in the silent era also suffered unpredictable variations from moment to moment.

As a result, silent films were shot and screened at different speeds. Even the same film would have fluctuations in frame rates from scene to scene or shot to shot.

The best example I know personally is Murnau’s Faust (1926). When Luciano Berriatua and I were analyzing the existing negatives of the film, we were puzzled when we compared the two negatives from the original production. Each negative was shot with a different camera. As was common in silent film, the filmmakers provided one negative for domestic circulation and one for international distribution. In production the cameras yielded different angles on the scene–sometimes quite similar, sometimes different.

Faust comparison 3

Faust comparison 2 600

Nevertheless, the action each one captured was identical. The problem is that the two negatives were different in length!

Why? The cameras were operated by two different cameramen, each hand-cranking at a different speed. The assistant cameraman, undoubtedly younger and more enthusiastic, was cranking faster!

Such a disparity was not uncommon. On the contrary, any archivist confronted with the decision about “suggested frame rate” for a given film is confronted with a choice among many different frame rates. What one writes on the cans is always a speed that fits better on average, not necessarily on each individual shot or scene.

And the reality in archives, cinemathques and everywhere else, is this. Most silent film screenings, if not virtually all, are today executed at one speed throughout the whole picture,

There is strong, albeit anecdotal, evidence that projectionists in the silent era were encouraged to crank different sections of a movie at different speeds. Cues in some musical scores suggest that sequences may have been slowed or speeded up. Projectionists may also have chosen to accelerate chases and battles or slow down love scenes. But there’s no way to know to what extent this manipulation was a common practice.

Personally, I think that we are inevitably looking at a silent film with modern eyes. For one thing, we have a completely different approach to continuity. We ask that a restored silent film be perfectly graded, that its tinting and toning be consistent from shot to shot and scene to scene. But when we examine an original print, particularly from the ‘teens, we realize that our modern concept of pictorial continuity did not apply. It seems to have come into play in the late Twenties or later. Our assumptions about continuity make us consider narrative or visual discontinuity a technical error, or at least a trace of a more cavalier attitude towards continuity. I think that discontinuity of technique was often voluntary and sometimes deliberate.


Fun with frame rates

This point applies to frame rates as well. I am convinced that audiences in the silent era were much more relaxed towards something that we now consider so important. Otherwise, it’s difficult to understand why there so many variations in speed within one picture, and between different versions of the same film.

We should remember as well that we cannot really discuss anything “silent” without taking into consideration that it was a long and complex period of fast evolution and change. A film from the early ‘teens and one made ten years later are two different beasts, and different considerations should apply.

So, to return to our main theme: Are frame rates impossible to vary with a DCP? No—in theory.

If you decide to create your DCP at 24 fps, you can easily slow down each scene at a different rate, if you so choose. Alternatively, if the Archive Frame Rates standard were in use, you could apply your four fixed speeds (16. 18. 20. 22) to different sections of the DCP by simply creating separate “reels” and a correct playlist. This would be a bit like a DCP in which certain parts, such as the restoration credits, are actually a separate file to be played before or after the main body of the film.

Have I tried this? No. Could it work? Yes. Is it worth trying? I am not sure. Again, I am afraid we are looking for a precision that was not really part of a film screening in the silent era.


I remain strongly convinced that “old stuff,” including silent film, has an appeal for the public, and not just for few cinephiles. I also strongly believe that an archive like ours in Brussels should offer a wide range of films on DCP for screenings around the world.

In order to show a silent film in the present environment I have to do some nasty things to the images. I have to slow them down by means of software, even if some solutions work well enough to fool some archivists and historians.

If the alternative is never to show Maudite soit la guerre (Alfred Machin, 1910) to anybody, I’d rather use some tricks to make the screening possible in a modern DCP environment. That’s better than making the film invisible. I am also old enough to remember how horrible it was in the 35mm days when there was really no way to show silent films outside a cinematheque. While we wait for archives and manufacturers to implement the fuller range of options, we must keep our films alive.

So when today a nice audience shows up to see Maudite with a pianist in a multiplex in the middle of Flanders (or Ohio or Nebraska), I admit that I’d be unhappy that I had to trick the speed to 18 fps. Yet I still consider it a victory.

For discussion of silent-film frame rates the standard point of departure is Kevin Brownlow’s 1980 article, “Silent Films: What Was the Right Speed?” Brownlow suggests that many American silent films were intended to be screened at a higher rate than the cameraman had used during filming. See also James Card, Seductive Cinema: The Art of Silent Film (Knopf, 1994), 52-56, and  Jon Marquis, “The Speed of Silence” (2011). On the revival of interest in high frame rates, see this 2011 article and Pandora’s Digital Box.

Nicola has written about other aspects of film preservation in the digital age. See his recent contribution to“The Last Picture Show,” a symposium in Artforum (October 2015), 288-291.Tacita Dean and Amy Taubin contribute to this as well.

The magnificent Masters of Cinema DVD set of Faust includes both the domestic German version and the export version, along with a lengthy comparison of them, from which I’ve drawn my illustrations.

Maudite soit la guerre 500

Maudite soit la guerre (Alfred Machin, 1910).

An auteur, three archives, and the archivist as auteur

City of Sadness (1989).

DB here:

The books have been piling up again, and so I pass along some recommendations. This time the volumes are unusually handsome. Apart from what they say, these publications display how sumptuous a serious film book can be.

Particularly welcome, in light of the touring retrospective of Hou Hsiao-hsien films that began at MoMI on 12 September, is Richard I. Suchenski’s anthology of writings on and by the director. Hou Hsiao-hsien is packed with color illustrations and fully up to the standard of other publications from the Austrian Filmmuseum.

Some of our top critics and scholars have contributed essays. From Japan comes Hasumi Shigehiko on Flowers of Shanghai; from France, Jean-Michel Frodon on Hou’s collaboration with Chu Tien-wen; from Canada, James Quandt, eloquent as ever on Three Times. American scholars are here too. We have Jean Ma on The Puppetmaster, Abe Mark Nornes on calligraphy and frame space, Kent Jones on time in Hou, and James Udden on Dust in the Wind. Jim is author of the first English-language book on Hou and a contributor to this website. To these essays are added the unique perspectives offered by Taiwanese observers Peggy Chiao on City of Sadness and Wen Tien-hsiang on unmarried women in Hou. A wide-ranging introduction by Richard brings out virtually all the artistic, political, and cultural issues that have been raised by Hou’s body of work.

Unabashedly auteurist—and what’s wrong with that?—the collection adds even more value by including an interview with Hou and Chu, who supply precious information about the work-in-progress The Assassin. There are also statements by three distinguished directors: Olivier Assayas, Jia Zhang-ke, and Koreeda Hirokazu. Koreeda notes: “Life’s details (like eating) should be respected.”

We learn as well about the craft behind the artistry. We get a cascade of information from artistic collaborators, including cinematographers, sound designers, actors and others. For example, Tu Duu-chih, sound expert, says: “What [Hou] wants is a natural form of expression and he always manages the atmosphere to meet his needs. If he shoots a drinking scene, for example, he will select real dishes and alcohol and they must be delicious.” Delicious is a good word for this book too, an absolute necessity for every serious cinephile.


Birthday greetings to Vienna and Brussels

Speaking of the Austrian Filmmuseum, it celebrates its fiftieth anniversary this year. To celebrate, it has issued a heavyweight, unorthodox boxed set, The Austrian Film Museum at Fifty. Director Alexander Horwath has done a monumental job in bringing a great deal of material together, creating not only a history of the Filmmuseum but a real contribution to international film culture.

In volume one, Aufbrechen (Setting Out), Eszter Kondor provides a history of the institution, with emphasis on its emergence the 1960s and 1970s. Volume two pays homage to Béla Balázs with the title Das sichbare Kino (The Visible Cinema). This is a plump anthology of texts, pictures, and documents from the museum’s history. Given the fact that Peter Kubelka was curator for many years, you’re not surprised to find a letter from Michael Snow and an email from Ken Jacobs. But I didn’t expect an interview with Groucho Marx (letter appended) and correspondence from Don Siegel. This volume also includes a complete record of programs at the museum.

Volume three, Kollection, is a visual treat. It gives us a short history of film in fifty items. The survey includes The Unfinished Letter, a 22mm Edison film ca. 1911-1913, outtakes from Murnau’s Tabu, Morgan Fisher’s site-specific Screening Room (1968), Kurt Kren’s leather jacket, Chuck Jones’ What’s Up, Vienna? (1983), and frames from Norbert Pfaffenbichler’s dizzying Notes on Film 03: Mosaik Mécanique (2007).

The texts in all of the books are in German, but there are so many posters, photos, sketches, diagrams, and storyboards (one by Vertov) that you learn merely by browsing. These volumes are a must for research libraries with a focus on film.

No less idiosyncratic is another anniversary volume, this one from the Royal Film Archive of Belgium, now known as the Cinematek. It was founded seventy-five years ago and, as if by cosmic convergence, it holds 75,000 film titles. But the book, 75000 Films, isn’t a celebration of this particular museum. It’s a book about film archiving in general. As curator Nicola Mazzanti puts it in his introduction:

Film archiving is already changing and in a few years will be very different from what it is now. New skills, new machines, new people. As no book like this one exists about film archiving we wanted to make sure that at least one is published before the picture changes completely.

What will change? Chiefly, the sheer tactility of the work.

Inspecting a film to check its conditions and its history is all about a physical relation with the material. While winding through a reel of film your fingers run along the edges to feel imperfections and damages. One bends a piece of film to assess its brittleness, caresses a splice to check its resistance. . . .

In other words, it is a precise, careful, dedicated and highly specialized work that often is as tedious as rewinding hundreds and thousand of reels (by hand, as electric motors are often dangerous to the film). An uneventful process until you stumble upon a lost film or a camera negative or perhaps just a beautiful copy in which the colours are perfectly preserved. And the beauty of those images takes your breath away, the thrill of the discovery repays you for all the hours of boring inspections.

To capture both the routine and the exhilaration, Nicola let three photographers document, without constraint, what they saw behind the scenes of the Cinematek. Xavier Harck, Jimmy Kets, and Marie-Françoise Plissart took their cameras into the vaults, the work spaces, the projection rooms, even the loading docks. The images are gorgeous and radiate the touch and heft of reel after reel, can upon can, in profusion that evokes Resnais’ Toute la mémoire du monde.

Along with the photos are three essays about film archives. They are personal reflections on cinematheques and their place in film culture. Dominique Païni writes (in French) about how access to films changed during his years as director of the Cinematheque Francaise. Erich de Kuyper, filmmaker and novelist, reflects in Flemish on guiding the Amsterdam Film Museum and creating the programs there. I contributed a piece in English that tries to fit my personal research work into broader trends of film archiving. My essay is elsewhere on this site.


The age of impresarios

Langlois is the dragon who guards our treasures.

–Jean Cocteau

In spring of 1973 the New York Times announced that a city board had approved the leasing of a building that would house the City Center Cinematheque. This was to be the American counterpart of the Cinémathèque Française, and Henri Langlois was to be its director.

Langlois’ project was as vast and flamboyant as the man himself. An old storage building under the Queensboro Bridge, on First Avenue between 59th and 60th Streets, would be converted into a cathedral of cinema. I. M. Pei would design the new building. (One sketch is above.) There would be three auditoriums, a staff screening room, an exhibition space, a restaurant, a bookstore, and a flower market. Programs would draw upon Langlois’ archive of 60,000 titles, and there would be screenings from morning to midnight, offering as many as fourteen films a day. Langlois predicted that attendance would surpass a million a year.

My own encounters with Langlois, brief though they were, came at just this moment. I needed to see several French silent films for my dissertation work, and in 1972-1973 I wrote to Langlois asking for permission to visit his archive. Thanks to Sallie Blumenthal, we made contact and he gave me permission. In July, during the summer of Watergate, I arrived at the Cinémathèque. An expansive Langlois welcomed me and gave me a tour of the Musée du cinema, which had been somewhat prematurely opened. Jean-Louis Barrault’s costume from Les Enfants du Paradis was draped on a coat hanger nailed to a wall. During my two months in Paris, thanks to Langlois and Mary Meerson, I saw several rare films on Marie Epstein’s viewing table.

I have sometimes wondered if my stay there was an accident of timing. Perhaps as a young American I benefited from Langlois’ high hopes for his transatlantic alliance. Pressed by money problems in Paris, he could imagine that the American branch of the Cinémathèque would vindicate his vision of a motion picture museum without walls: films circulating everywhere, films shown all the time, no film too minor to merit attention. Had the City Center venue come to fruition, it would have overshadowed the Museum of Modern Art, Manhattan’s temple of cinema history. But the project collapsed fairly soon. It required private financing, and during the recession and the oil crisis money was hard to come by.

Langlois’ American adventure is just one episode in the tapestry presented in another archive-related anniversary volume: Le Musée imaginaire d’Henri Langlois. Edited by Dominique Païni, this de luxe production accompanied the Cinémathèque’s massive exposition from April to August. If Langlois were still living, he’d be 100 this year, and the sumptuousness of this catalogue is in part a testimony to what he accomplished. He helped make cinema equal to the other arts in cultural significance. But the enterprise is not too serious: Païni has made the volume’s emblem the famous shot, taken by an unknown hand, of Langlois in the kiss-my-ass salute.

The menu is familiar. There are informative essays by various hands, interspersed with illustrations and documents. But the execution is extraordinary. Reminiscent of 1920s publications, on rough paper and with decentered blocks of type, this square volume seduces you into sustained browsing. Moreover, it contains reproductions of artworks related to Langlois and his institution. We have works by Beuys, Chagall, Duchamp, Fischinger, Léger, Matisse, Miró, Picabia, Richter, Severini, and Survage—all connected, somehow to Langlois and cinema. There are frames from Le Métro, a 1934 film by Langlois and Franju., There are guest-book signatures, catalogue covers, correspondence, and much more.

The catalogue of the exposition is accompanied by a slender but no less ingratiating biographical chronology, festooned with still more images. Mais qui est ce Monsieur Langlois?  opens with a striking portrait by Henri Cartier-Bresson and ends with the telegram sent by Jean Renoir after Langlois’ death in 1977. “We have lost our guide, and now we feel alone in the forest.”


I met Langlois in the days of rivalry among archives, a good deal of it triggered by him. If you were welcomed to Archive X, and word got to Archive Y, that venue would shun you. Surveying these books, I was struck by their quiet assumption that archives must collaborate on projects and share their treasures. After Langlois, archiving became more cooperative, more routinized, more professional, and–Langlois and his allies would say–more boring. We may not need dragons now. Perhaps, though, we had to go through the turmoil of the Age of Impresarios in order to appreciate why collecting films was so important.

For some ideas on Hou’s visual style, see Chapter 5 of my Figures Traced in Light, this entry, and a discussion of the early films on this site.

Henri Langlois. Photo by Pierre Boulat.

Where movies live and breathe

Thanks to web tsarina Meg Hamel, I’ve just posted a new essay (you also see it on the left column). It’s a historical and personal discussion of film archives. It could also be titled, “How I Spent Nearly 40 Summer Vacations.”

It originally appeared in Nicola Mazzanti’s book, 75000 Films. More on that book and related topics soon! After I get done arm-wrestling Adieu au langage, the latest work of the world’s youngest 83-year-old filmmaker.

What’s left to discover today? Plenty.

Der Tunnel (William Wauer, 1915).

DB here:

Being a cinephile is partly about making discoveries. True, one person’s discovery is another’s war horse. But nobody has seen everything, so you can always hope to find something fresh. There’s also the inviting prospect of introducing a little-known film to a wider community–students in a course, an audience at a festival, readers of a blog.

A festival like Bologna’s Cinema Ritrovato (we covered this year’s edition here and here and here) offers what  you might call curated discoveries. Expert programmers dig out treasures they want to give wider exposure. Such festivals are both efficient–you’re likely to find many new things in a short span–and contagiously exciting, because other movie lovers are alongside you to talk about what you’re seeing.

A year-round regimen of curated discoveries is a large part of the mission of the world’s cinematheques. This is why places like MoMALACMAMoMIBAMTIFFICA, and other acronymically identified showcases are precious shrines to serious moviegoing.

But other discoveries are made in a more solitary way. Film researchers, for instance, ask questions, and some of those can really only be answered by visiting film archives. Sometimes we need to look at fairly obscure movies. And despite the rise of home video, there are plenty of obscure movies that can be seen only in archives. It’s here that the programmers of Ritrovato and Pordenone’s Giornate del Cinema Muto come to select their featured programs.

Archive discoveries aren’t predictable, and many are likely to be of interest only to specialists. Such was the case, mostly, with our archive visits this summer. But as in years past (tagged here), all our archive adventures yielded pure happiness.

This time I concentrated on films from the 1910s-early 1920s films because I hope to make more video lectures about this, the most crucial phase of film history. (One lecture is already here.) In our archive-hopping, we saw films I was completely unfamiliar with. I re-watched some films I’d seen before and found new things in them. I detected some things of interest in films I hadn’t known. Most exciting was our viewing of a major film that has gone unnoticed in standard film histories.


In the steps of Jakobson and Mukarovský

Love Is Torment ( Vladimír Slavínský, Přemysl Pražský, 1920); production still.

First stop was Prague, where I was invited to give two talks. At the NFA we saw two films on a flatbed: a portion of Feuillade’s Le Fils du Filibustier (1922) and a cut-down version of Volkoff’s La Maison du mystere (1922), the latter a big gap in our viewing. The expurgated Maison came off as rather drab, lacking nearly all the big moments much discussed in reports like James Quandt’s from a decade ago.  So we search on for the full version. . . .

As for the Feuillade: Le Fils du Filibustier was his last “ciné-roman.” Our two-reel segment, which seemed fairly complete, confirmed his late-life switch to fairly fast, Hollywood-style editing, with surprisingly varied angles.Again, though, we yearn to see the entirety of this pirate saga.

On another day the archive kindly screened several 1910s-early 1920s Czech films for us. Our hosts Lucie Česálková and Radomir Kokes translated the titles and provided contexts. Among the choices were Devil Girl (Certisko, 1918), with a protagonist who’s more of a tomboy than a possessed soul; and the full-bore melodrama Love Is Torment (Láska je utrpením, 1919). The plot, outlined here, involves scaling and jumping off a tower, twice. Once it’s a stunt for a film within the film, the second time (below) it’s the real thing.

Radomir explained to us that one of the co-directors,  Vladimír Slavínský, seemed in his 1920s films to specialize in building two reels (often the third and fourth) in a “classical” fashion before letting the other three become more episodic. And indeed, most of the late 1910s-early 1920s films we saw were up to speed with other European filmmaking, in their staging, cutting, and use of intertitles.

We look forward to viewing more Czech films as the opportunity arises. A culture that gave us Prague Structuralism is definitely worth getting to know better. In the meantime, the journal Illuminace, edited by Lucie, is injecting a great deal of energy into local film studies, and the archive is entering a fresh phase with its new director, Michal Bregant.


3D excavations

Der Hund von Baskerville (1914).

In Munich, we reconnected with our old friends Andreas Rost, now retired from administering the city’s cultural affairs, and Stefan Drössler, director of the Munich Filmmuseum. We also reunited with the stalwart archivists Klaus Volkmer, Gerhard Ullmann, and Christoph Michel. Talking with them, we realized we hadn’t been back for over ten years. Klaus and Gerhard were warm and helpful during our earlier visits.

One rainy afternoon, Stefan shared his research on the history of 3D. He presented a spectacular PowerPoint, with rare images and some truly startling revelations. He has given this talk at intervals over the years, but it grows and deepens as he discovers more. Accompanying it, he screened some Soviet 3D films, including the 1946 Robinzon Cruzo. This mind-bending item was made with diptych images, so that the projected image turned out to be slightly vertical. The soundtrack runs down the middle.

The director, Aleksandr Andriyevsky, made excellent use of 3D to evoke the stringy vines and protruding leaves of Crusoe’s island. Amid all the talk today about glasses-free 3D, it’s interesting to learn that Soviet researchers prepared such a system. Stefan’s archaeology of 3D, for me at least, was a pretty big discovery.

At Munich we also saw three silent German titles. Two were associated that resourceful self-promoter Richard Oswald. Sein eigner Morder was a 1914 version of Dr. Jekyll and Mr Hyde, directed by Mac Mack from Oswald’s screenplay. Shot in big sets, it spared time for the occasional huge close-up. The other film was Oswald’s semi-comic adaptation of The Hound of the Baskervilles (Der Hund von Baskerville, dir. Rudolf Meinert, 1914), which he had already turned into a play. The sleuth isn’t exactly our idea of Holmes (see above), and he isn’t as quick-witted, I thought, but it was an enjoyable item. Dr. Watson has a sort of tablet which picks up messages;  Holmes’ orders are spelled out in lights on a grid. Stefan rightly called it a 1914 laptop.

As for the third film: More about that coming up.


The shadow of Hollywood

Les Deux gamines (1921).

At Brussels, thanks to the cooperation of the Cinematek, I was able to see several items for the first time, and two held considerable interest. The short The Meeting (1917), by John Robertson, showed a real flexibility in laying out the space of a cabin both in front of the camera and behind it. Most interior scenes in 1900-1915 cinema bring characters in from a doorway in the rear of the set (as Feuillade does in his 1910s films) or straight in from the sides, perpendicular to the camera (as Griffith tends to do). The Meeting shows that the diagonal screen exits and entrances that we see in exteriors were coming into use in interior sets as well.

Another 1917 film, Frank Lloyd’s A Tale of Two Cities, was further support for the idea that American continuity filmmaking was well-established and already being refined at the period. Dickens’ classic tale is handled with dispatch–rapid exposition, smooth crosscutting to set up the plot lines–and the film makes dynamic use of crowds surging through well-composed, starkly lit frames. There are also some remarkably expressive close-ups, evidently made with wide-angle lenses.

To clinch a plot point, the resemblance of aristocrat Charles Darnay to British solicitor Sydney Carton, the star William Farnum plays both characters. Not much of a problem if you keep the characters in separate shots; the good old Kuleshov effect (aka known as constructive editing) makes it easy.  At this period, though, filmmakers were perfecting ways to show  one actor in two roles within a single shot. The most famous examples involve Paul Wegener in The Student of Prague (1913) and Mary Pickford in Stella Maris (1918).

Cinematographer George Schneiderman contrives some really convincing multiple-exposures showing Farnum as both Darnay and Carton. There are some standard trick compositions putting Farnum on each side of the screen, but several images take the next step and let the actor cross the invisible line separating the two halves. At another point, we get a flashy passage showing the two facing one another in court, followed by a “Wellesian” angle of the two characters’ heads in the same frame.


Hollywood’s pride in photorealistic special effects, so overwhelmingly apparent today, has deep roots.

Part of my Brussels visit involved checking and fleshing out notes on certain films I saw many years ago. Some were wonderful William S. Hart movies like Keno Bates, Liar (1915; surely one of the best film titles ever). There was, inevitably, Feuillade as well. The influence of Hollywood was powerful in the ciné-roman Les Deux gamines (1921). This baby, released in 12 parts originally, runs nearly 27,000 feet. At 20 frames per second, it would take six hours to screen. What with changing reels, making notes, counting shots, pausing to study things, and taking stills, Kristin and I took about ten hours to watch it.

Was it worth it? An adaptation of a popular stage melodrama, it can’t count as one of Feuillade’s major achievements. Two girls are left alone when their mother is reported dead. They are adopted by their gloomy grandpa and tormented by his overbearing housekeeper. They become the target of kidnapping by gangster pals of their father, who has divorced their mother and turned to a life of petty crime. Their allies are their young cousins, a wealthy benefactor, and their godfather and music-hall star Chambertin. Everything ends happily, if you count the father’s redemptive sacrifice on behalf of  a pregnant woman.

Les Deux Gamines is determined to delay its ending by any means necessary. Form here definitely follows format; Feuillade fills out the serial structure with plots big and small. (Shklovsky would love it.) There are incessant abductions, escapes, rescues, coincidental meetings, and timely reformations, plus at least three cases of people wrongly assumed to be dead. All of this is accompanied by an endless exchange of telegrams and letters. People are forever piling into and out of carriages, train cars, and taxis. Such material serves as makeweight for some genuinely big moments, including a cliff-hanging scene and a stunning climax in a smuggler’s warehouse stuffed with gigantic bales of used clothing.

Like Le Fils du Filibustier, this film shows Feuillade trying to change with the times. The supple long-take staging of Fantômas and Les Vampires and Tih-Minh mostly goes away, to be replaced by rapid editing. Feuillade employs standard continuity devices, as when the grandfather discovers that the kids have sneaked out at night and are trying to return by scaling the gate.


Feuillade varies his angles and lighting to accentuate the moment of visual discovery. Elsewhere, some appeals to “offscreen sound” (cutaways to doorbells and telephones) built up to a surprise effect.

But by the energetic standards of, say, Robertson or Lloyd several years earlier, Les Deux gamines is fairly timid. Feuillade doesn’t explore editing resources very much here, not even as much as in Le Fils du Filibustier. The fairly quick cutting pace stems partly from the stratagem of having dialogue titles interrupt static two-shots of characters talking to one another. This sort of proto-talkie-technique yields efficient storytelling but not much visual momentum. Feuillade tried flashier things in other films of the period (see here).


Hours and hours of nothing but Bauers

The Alarm (1917).

Yevgeni Bauer, one of the master directors of the 1910s, remains lamentably unknown. About two dozen of his over seventy films survive, but many of the ones we have lack intertitles. A few of his films are available on DVD (most obviously here; less obviously here). He died of penumonia in 1917, between the February revolution and the Soviets’ coup d’état in October. He was only 52.

My first archive-report entry back in 2007 recorded my interest in Bauer, and I’ve returned to his films over the years. Now here I was watching some again, confirming things I found of interest then, and discovering (that word again) new felicities. I hope to say more in those short video lectures on the 1910s, but I can’t leave without giving you a taste of his qualities.

Two of the films I saw this time were from 1917. The Alarm (Nabat) came out in May 1917, just before Bauer’s death in June. Originally running eight reels, it was cut down after the initial release, and that’s evidently the version we have. For Luck (Za Schast’em, September 1917) was directed by Bauer from his sickbed. Both films are fairly hard to follow. The Alarm lacks intertitles, and For Luck has many fewer than it had originally.

The two films are of exceptional interest, though. For one thing, there’s  the involvement of Lev Kuleshow, who at the age of eighteen served as art director for the earlier film and, apparently reluctantly, as an actor in the later one. More important, the films remain as beautifully designed, staged, and acted as ever.

The Alarm is a wide-ranging drama set before the February upheaval. The drama involves romantic intrigues among the upper class, interwoven with a workers’ rebellion against a master capitalist. The millionaire Zeleznov holds court in a vast office with chairs bearing ominous spires and spiky arches; the windows open onto a view of his factory. A long-shot view is above; here’s a sample of how Bauer shows off his decor in something akin to shot/ reverse-shot.


The idea of capitalism as an overreaching religious striving is evoked by turning Zeleznov’s headquarters into a Symbolist cathedral. And looking at the second shot, you wonder whether Kuleshov’s inclination to stage his own scenes against pure black backgrounds has its source in his work for the man he called “my favorite director and teacher.”

As ever, Bauer makes fluid use of depth. He choreographs meetings of Zeleznov’s brain trust in ever-changing arrangements, and he eases a man out of a boudoir through a mirror reflection over a woman’s fur-draped shoulder.


Compared to the scale of The AlarmFor Luck is decidedly low-key–a bourgeois melodrama that extends barely beyond an anecdote. Zoya has been a widow for ten years, and she hopes to marry the loyal family friend Dmitri. But Zoya’s daughter Lee hasn’t yet reconciled to losing her father. The couple hope that Lee has worked out her grief during her dalliance with a young painter (played by Kuleshov), but she reveals that all along she has hoped to marry Dmitri.

The Alarm used some extravagant sets, both for interior and exterior scenes, but a good deal of For Luck takes place in parks and terraces. The sincere Enrico sketches Lee in front of swans, and they steal some moments in a bower.


Still, there are some interiors boasting Bauer’s famous pillars and columns, which create massive, encapsulated spaces. Here Zoya looks off, in depth, at the ailing Lee, in bed on the far right.

Sharp-eyed Bauerians will notice the mirror set into the left wall, reflecting Zoya. Kuleshov, who did art direction on this as well as The Alarm, worried more about the trumpet-blowing Cupid floating between the pillars on the left. (“It turned out bad on the screen–incomprehensible and inexpressive.”) He did think that the tonalities of the set worked well: “As an experiment, I put up a set painted in shades of white that were ever so slightly different from one another.”

“Ever so slightly different” isn’t a bad evocation of the tiny variations of shape and shade, light and texture, that characterize Bauer’s ripe, sometimes overripe, imagery. This is a social class on the way out, but it leaves behind a great glow.


Tunnel: Vision

The Tunnel (1915).

In 1913, the popular novelist Bernhard Kellermann published Der Tunnel. It’s not quite science-fiction, more a prophetic fiction or realist fantasy in the vein of Things to Come. The book became a best-seller and the basis of a 1915 film directed by William Wauer.

The plot would gladden the heart of Ayn Rand. A visionary engineer persuades investors to fund building an undersea railway connecting France to the United States (specifically, New Jersey). No meddling government gets in the way of this titanic effort of will. Max Allan buys land for the stations, hires diggers from around the world, and risks everything he has. The obstacles are many. An explosion scares off workers; there is a strike; impatient stockholders raid and burn the company headquarters.

Max Allan moves forward undeterred, though he hesitates when his wife and child are stoned to death by a mob. After twenty-six years, the railway is opened. Max, along with his new wife (the daughter of his chief backer), proves it’s safe by taking the first transatlantic train. The event is covered by television, projected on big screens around the world (above). In the original novel, a film company was commissioned to document every stage of work.

The book skimps on characterization, and the film is even less concerned with psychology. Once the character relations are sketched, Wauer goes for shock and awe. The Tunnel‘s thrilling crowd scenes of work, fire, devastation, riots, and panic look completely modern. Bird’s-eye views of stock-market frenzy anticipate Pudovkin’s End of St. Petersburg, and Wauer creates an Eisensteinian percussion of light and rushing movement as workers flee the tunnel collapse.


For the sequences showing the tunnel construction, Wauer supplies violent alternations of bright and dark as men, stripped and sweaty, attack the rock face. The variety of camera positions and illumination is really impressive.


Comparisons with The Big Film of 1915 are inevitable. The intimate scenes of The Tunnel are far less delicately realized than the romance and family life of The Birth of a Nation, and the battle scenes in Birth have a greater scope than what Wauer summons up. But Wauer’s handling of crowds is more vigorous than Griffith’s riots at the climax of Birth, and his pictorial sense is in some ways more refined, even “modern.” There’s little in Birth as daringly composed as the static long shot surmounting today’s entry.

Wauer can handle small-scale action very crisply. The opening scene in an opera house creates low-angled depth compositions more arresting than Griffith’s depiction of Ford’s Theatre. Max’s wife, in one box, is watching his efforts to attract funding from the millionaire Lloyd. Wauer constantly varies his camera setups to highlight Macxs wife in the background studying Lloyd’s daughter, sensing in her a rival for her husband. Whether the angle is high or low, the wife’s presence in her distant box is signaled at the top of the frame.


The second and third shots above present similar but not identical setups, adjusted to reset the depth composition.

It was at Munich’s Filmmuseum a decade ago that I first encountered the brooding power of Robert Reinert’s Opium (1919) and Nerven (1919), the latter now available on DVD. I was convinced that Nerven was as important, and in some ways more innovative, than the venerated Caligari. Now the conviction grows on me that in The Tunnel we have another galvanizing, outlandish masterwork of the 1910s. I hope it will somehow get circulated so that wider audiences can discover it. Yeah, that’s the word I want: discover.

Without archivists, no archives. We’re grateful to Michal Bregant, head of the Czech Republic’s archive, for access to films and for his companionship during our visit. Thanks as well to Lucie Česálková, our host; her knowledgable colleague Radomir Kokes (who kindly corrected the initial version of this post); Petra Dominkova, our Czech translator; and Vaclav Kofron, editor of the Czech versions of our books. Lucie supplied the frame enlargement from Love Is Torment. As well: Good luck to the Kino Světozor!

In Munich, we owe a huge debt to archive chief Stefan Drössler, for his generous sharing of information and his and Klaus Volkmer’s rehabilitation of The Tunnel. Stefan also provided the images from Robinson Crusoe and The Hound of the Baskervilles. Coming up is his work for the annual Bonn International Silent Film Festival, 7-17 August. Thanks as well to Gerhard Ullmann and Christoph Michel.

In Brussels, as ever, the Cinematek has made us welcome, and we thank archive director and long-time friend Nicola Mazzanti and vault supervisor Francis Malfliet. Over the last thirty years, a great deal of our research has depended upon the cooperation the Cinematek leadership: Jacques Ledoux, Gabrielle Claes, and now Nicola.

My quotations from Kuleshov come from Silent Witnesses: Early Russian Films, 1908-1919, ed. Yuri Tsivian and Paolo Cherchi Usai (Pordenone: Giornate del Cinema Muto, 1990), 388-390.

There’s a chapter on Feuillade in my Figures Traced in Light, where Bauer is discussed as well. My essay on Robert Reinert is in Poetics of Cinema. 

Thanks to Antti Alanen for correction of a misspelled title. Speaking of discoveries, you’ll find plenty on his wonderful Film Diary site. During his recent trip to Paris, he’s writing about art exhibitions, Dominique Paini’s Langlois exposition at the Cinémathèque, and Godard’s Adieu au langage.

Screening at the Czech Republic’s National Film Archive. From left: Michal Večeřa, Tomáš Lebeda, Radomir Kokes, Lucie Česálková, and Kristin.

David Bordwell
top of page

comments about the state of this website go to Meg Hamel.