Skip to content

Exploring the Speed of Silent Comedy Film – test #1

One of my fascinations with silent film is with how it is that it is shown faster than the speed at which it was shot, but doesn’t look it. Once I figured out that the secret is in adjusting the tempo of the performers’ physicality, the next step was to shoot tests to see if I could replicate what comedians and actors were doing over a hundred years ago to the same effect.

Over the last several years I have had the great fortune to meet, work and play with some of the many talented members of NYC’s clown and physical comedy community. I’ve participated in the weekly NYC Physical Comedy Lab, and have been in and involved with a few shows done by physical comedy theatre company Parallel Exit. This has informed me about what what happened in front of camera in the silent era in a way that merely watching the films over and over could never do.

Using the smartphone app Filmic Pro, over the last several years I’ve shot a lot of “undercranking” silent film tests with the clowns and physical comedians I’ve gotten to know. When we’ve gotten the taking speed and physical adjustments just right, the results look very much like what silent era silent film looks like.

Working with clowns and physical comedians has made understanding and replicating silent film comedy a lot easier. These comedians already know and understand the physicality of how to perform this kind of comedy. So it’s just been a matter of slowing it down a tad, and inserting pauses here and there, like physical commas, as needed.

Filmic Pro is the only app I’ve found that has the ability to shoot at variable speeds that emulates the way a motion picture camera would behave. When you film at 16 frames per second in Filmic Pro, the app actually only takes 16 images per second and also adjusts the shutter speed accordingly. The latter is something most apps do not do.

In this and the next four blog posts I’m going to share some short one-minute undercranking tests that were done a few years ago with some of the Parallel Exit folks. Joel Jeske, Danny Gardner, Mark Lonergan and I went to a rehearsal studio in NYC for a few hours, and cooked up these little short bits. Each of them employs the basic performance adjustments, the abandoning of logic, and the flaunting of gravity and physics that silent comedy did back in the day. These videos are not perfect, but they’re awfully close to the “look” of silent film.

I will also look for the tests and experiments I filmed with the NYC Physical Comedy Lab and at some point, will share those as well.

One key asset we had that no one had in the silent era was that because we were filming with an iPad and with Filmic Pro, we were able to watch each of our “takes” immediately after shooting, and could see where we needed to slow down and insert or lengthen pauses in physical movements so they would read and not zip by. We could identify the adjustments we needed to make, as well as see opportunities of where else we could take the business, and re-shoot.

I hope you enjoy these little 21st century Kinetoscope-length films.

This first one was basically about getting the tempo of movements right, making them slightly more deliberate or slower than real-time, and in breaking up movements with brief pauses that the undercranking would erase when sped up so they’d register and not look like sped-up film.


Next posts in this series:

Links:

7 thoughts on “Exploring the Speed of Silent Comedy Film – test #1”

  1. This is certainly a brand new idea. I’ve always heard the the silent films were originally shown at the “normal” graceful speeds at which they were shot. That later projectors showed revivals of silent films at faster speeds and created the false idea that was the original intention. You seem to be trying to recreate this newer false look.

    1. I’ve been researching the heck out of this for several years and have found much evidence to support it. The films were originally shown at frame rates that were faster than taking speed. The films were also shot at a variety of speeds. The running speeds were never enforceable and some theaters would speed up the films depending on if they needed to fit in an extra show or if the film was more than 7 or 8 reels. This seems to have begun in the mid-to-late ‘teens. Maurice Costello, a big star for Vitagraph in the mid-to-late ‘teens, told of watching himself in films in theaters and seeing the sped-up action, developed what he called a “slow motion acting” technique. I’ve just read a transcript of a speech given by Woodrow Wilson in 1916 at a convention of film exhibitors where he references having seen himself onscreen as being a very brisk walker. By 1922 there are running times for comedy shorts published that I’ve seen that indicate a projection speed of 24 fps, while comedy shorts were taken at a range of 12-18. Chaplin would have filmed City Lights at 24 fps if real-time speed was the norm, but he clearly didn’t and surviving shooting records for it and for Modern Times indicate taking speeds of 16 and 18. The Barrymore Don Juan was issued with a track at 24 fps and clearly was taken at 18 and 16 fps; the non-singing sequences in The Jazz Singer are taken at 18 fps. The error in earlier revivals is that films from the teens which should have been run at 20 fps were shown at 24 – because of a lack of variable-speed project or video transfer capability – and appear over-sped. The speed-up of silent film is not false; it was always part of the language. Films were being shown closer to 27 fps by the end of the silent film era, and faster than that in the small town theaters.

  2. This is a very nice experiment that you are doing. I like the music, too (I wonder who the musician was?). I remember reading about Chaplin sometimes designating a camera speed before shooting a take, “OK, Rollie, let’s make this a happy 12” (as in 12 fps), so it doesn’t surprise me that comedies would purposely undercrank.

    Occasionally, they would overcrank for certain effects, like in Charley Chase’s “Bromo + Juliet”, when they filmed a horse in slow motion to portray it as drunk, after drinking from the fountain that Charley threw the case of whiskey in.

  3. This is fascinating. On rare occasions, filmmakers appear to have used what might be termed extreme under cranking i.e. time-lapse, where a single frame is being taken every few seconds or so i.e. less than one frame per second. I think this is possibly how the effect in the opening shot of Reel 2 of Keaton’s “The Goat” was achieved, where the shot starts with Buster sitting on the front of a train in the distance, and ends with him in close-up. It looks like the train was actually moving forward extremely slowly (thus ensuring the smoke was moving in the right direction, and allowing it to stop more or less in the right position) while the cameraperson was taking a single frame every few seconds or so. A giveaway seems to be at the end of the shot, where Keaton’s head move appears quite jittery and unnatural – if single frames were being taken every few seconds, it would be very difficult to time a head movement to appear smooth.

Leave a Reply