I’ve recently caught up with Charles O’Brien’s splendid, paradigm-shifting book, Cinema’s Conversion to Sound, subtitled Technology and Film Style in France and the US. Briefly, O’Brien combines thorough research into primary sources and empirical methods of analysing film style to critique the conventional idea that the coming of sound produced a homogenisation of cinema which spread abroad from Hollywood in the 1930s. In particular, this is not what happened in France, which developed a strong predilection for what was called the film parlant—the talkie which used direct recorded sound—as opposed to the film sonore—the sound film where the sound was post-synchronised, which fast became the Hollywood way.

Of course it’s not quite so simple. The French terms were introduced right at the start of the sound film, before the technology allowed post-production mixing, so you either had to use a synchronous recording (which on the model of radio could mix different sound sources together) or you had to lay the sound down against the image. Once post-production mixing was introduced around 1931, the two methods could be combined in various ways, and so they were. But what you get is a marked French preference for synchronous shooting and minimal post-synchronisation, while Hollywood opts for elaborate post-synchronisation, where synchronous dialogue was mixed with post-synchronised music and effects. The music was recorded to picture using click tracks to achieve proper timing; the effects were also recorded live to picture, but separately, in a special studio, following a methodology developed by one Jack Foley. Then the different tracks were re-recorded to produce the final mix. (Actually there was an intermediate stage, a music-and-effects mix, which allowed the dialogue to be dubbed in other languages.)

One of the striking things about this contrast is that it corresponds to the artesanal production methods of the French film industry on the one hand, and on the other, the Fordist organisation of production in Hollywood. It also produces big differences in style and aesthetics, even in what we might call acoustic ideologies. Hollywood privileged the intelligibility of speech; they learned to mount a single microphone on a boom which followed the actors movements, shooting with a single camera and repeating the action for different camera set-ups. French cinema privileged speech as performance, the better to exploit the talents of actors coming from popular theatre, and their skills in improvisation. They would sometimes shoot on more than one camera simultaneously; they also used multiple microphones which would be faded in and out as the actor moved around, a technique that allowed for longer takes and elaborate camera choreography. The great master was Renoir.

I confess I should have read this book already, particularly when I was writing the chapter on sound in my book on documentary. Although O’Brien doesn’t talk about documentary at all, his argument provides a crucial part of the context, mainly because he shows that the ‘experiments’ in direct sound filming by directors like Pagnol and Renoir, which I talk about, were not quite so out of the ordinary. My argument about documentary—that it was disadvantaged by the coming of sound, because the technology was designed for studio use, not location recording—still stands, but O’Brien’s book sheds new light and raises new questions.

Given that sound remains, within the academic field of film studies, the most neglected major branch of the production apparatus, this is a study in film history with far-reaching implications. O’Brien merely hints at this in his conclusion when he speaks of the nouvelle vague and their adoption of direct sound, now transformed by magnetic sound recording, as a retrieval of techniques practised in the 1930s but then rejected by the commercial industry.

The introduction of magnetic tape recording after the war—a fascinating story in itself*—would accentuate the contrast between le film parlant and le film sonore because of two factors. On the one hand, it allowed synchronous filming on location, which had previously been very difficult. On the other, it produced a new method of post-synchronising dialogue by means of looping: the edited film is broken down into short bits which are projected in a loop while the actors repeatedly speak the words until they fit exactly. This technique is used to accomplish ever more elaborate location shooting where capturing clean dialogue is almost impossible. What this means was encapsulated for me in a conversation a dozen years ago with the Colombian director Sergio Cabrera about his wonderful film La estrategia del caracol, when he told me proudly that 90% of the soundtrack was direct sound, and added, ‘but in Hollywood it’s the other way round’. Which in turn reminds me of a remark by Richard Attenborough that the single largest post-production cost in the making of Gandhi was bringing the cast back from wherever they’d dispersed to several months later for dubbing.

It was the French example, and its deliberate rejection of the industrial model, that provided the key for new wave film-makers in Latin America and other third world countries, for two powerful reasons: it helps in keeping production costs down to a manageable level, and it brings a strong reality effect to the screen. It also broadens the incorporation of non-professional actors and children—for which another Colombian film, Víctor Gaviria’s La vendedora de rosas, is a primary model, and so is the work of Kiarostami and other Iranian film-makers. In short, the difference between industrial and independent cinema throughout the world since the 1960s is in part defined by their different sound worlds. On one side, the artificiality of the acoustic cocoon constructed by Hollywoodian ideology—no blockbuster is nowadays complete without surround-sound. On the other, equally the result of careful fashioning, the soundtrack as the camera’s ear, the acoustic equivalent of point-of-view, which opens out towards the represented world with its little noises and resonances and sometimes echoes. But this initial formulation of binary opposition is only provisional. There are more questions to be asked here about other models for sound design and soundtrack composition, especially the whole issue of music and the different acoustic envelopes it creates.

And this would still only be a beginning, Historically there’s the question of sound in Soviet Russia, which O’Brien refers to in passing, and in general of language conventions and the history of dubbing, and the various types of disembodied voices (some of these are topics which have attracted intelligent individual studies). And then there’s the whole issue of digital sound. I expect I’ll return to some of these, because this term I’m teaching a new class on film sound, which is what finally prompted me to read this excellent book.

* See Repeated Takes