On Christmas day, Hollywood’s newest movie musical will open at theatres nationwide. Les Misérables is based on Claude-Michel Schonberg, Alain Boublil, and Herbert Kretzmer’s hit stage musical, which of course was based on Victor Hugo’s 1862 novel. The English version of the stage musical opened in London in 1985, on Broadway in 1987, and was eventually translated into twenty-one languages and performed in over forty countries. Along with Andrew Lloyd Webber’s Cats and the Phantom of the Opera, Les Misérables can be considered among the most commercially successful stage musicals of all time.
Hit stage musicals have often been adapted for the screen. In recent years, Chicago, Hairspray, Rent, Dreamgirls, and The Producers were captured on film. It’s therefore no surprise that a mega-hit like Les Misérables was destined for the silver screen. However, one important aspect of this latest movie musical demonstrates a significant departure from most of its predecessors: the singing.
Historically, the vocal tracks for these movie musicals are recorded weeks or even months in advance of filming. This method gives the director, musical supervisor, sound editors and engineers a clean audio sample that sounds nearly perfect. Consider this clip from the 1944 musical Meet Me in St. Louis:
You’ll see that as Judy Garland dances, the heels on her shoes contact the wooden floor just beyond the rug. This dancing would surely create a “tapping” sound but of course we don’t hear this tapping in the clip because the audio was recorded long before the scene was filmed. Likewise, there are a number of actors in the shot; a single cough or sneeze from any of these actors would ruin the audio sample if it were recorded live. Additionally, if a musical number features an actor moving from place to place, the cameras also have to move–this means that the clanking of equipment or shuffling of feet could also destroy a live shot. Finally, it is physically difficult to sing and dance at the same time. Singers engaged in brisk physical activity, even light dancing, can easily find themselves out of breath. Recording the vocals in the studio eliminates all of these problems.
However, recording vocals in advance creates two important complications for the singing actor. First is the technical challenge of lip-synching. In the clip above, Garland must dance and engage with the other actors, all while she is attempting to flawlessly match the vocals she had recorded in a studio many weeks earlier. It’s difficult enough to sing, dance, and act at the same time–simultaneously lip-synching is a near-impossible challenge.
A more extreme example can be seen here, a dance number from High School Musical. The choreography is more intense than the Judy Garland clip and the viewer can easily see the singers struggling to mimic their prerecorded vocals.
Notice also at the end of the first verse, sung by Zac Efron, that he is doubled over, getting a playful rub on the head from another actor. Amazingly, his singing is unaffected by this physical posture. This example introduces the second problem with prerecorded audio for musical films: lack of spontaneity. If the actors behave spontaneously on camera, their physical action almost certainly will not match the audio; this creates the dubbing issues seen in the clip from High School Musical. On the other end of the spectrum, actors attempting to perfectly match the prerecorded vocals will almost certainly disengage from the acting moment and may even have to adjust choreography in order to truly match the vocals–the audio might sound “pretty” but the song will look dull on the screen.
For these reasons, director Tom Hooper decided to record the songs for Les Misérables live on the set. It’s an incredibly innovative approach that eschews nearly a century of movie musical tradition. This “behind the scenes” video provides a look at the techniques used to capture live audio.
As I read about the film and how the audio was recorded, I thought about the audio presentations my students give in Shakespeare Off the Page, one of the courses I teach in the UNCG BLS program. I ask students to give online presentations by recording their voice and submitting that audio file for evaluation. When I first taught the course, I was most interested in evaluating 1) Is the student’s argument convincing? and 2) Can I hear and understand the student’s voice? But as I’ve continued teaching the course, I’ve started thinking more like Tom Hooper. What I really want to hear in a presentation is the speaker’s personality–their interest in the topic, their engagement with the material, and their willingness to be spontaneous rather than “perfect.” After all, if one is simply going to recite text, what’s the point of speaking aloud? It’s a wonder that it’s taken Hollywood so long to figure this out.