directed by: Charles Chaplin


IMDB link:

Submitted by Yuri Tsivian on 2007-02-08

Yuri Tsivian's comment:

The audience composition: between 17 and 20 graduate and undergraduate students enrolled in my History of International Cinema (Silent) class. A few of them may have been familiar with the film (this is a repeat screening this week). The copy shown is a 16 mm print with added soundtrack shown at 24 fps (of the UofC Film Studies Center study collection).

The credits and the end title were not counted. I should have thought to program one key for the “end

Number of shots:
MSL 16.8 6 4.5
StDev 37.4 9.5 11.4
Min 2.1 1.3 1
Max 152.1 50.9 56.1
CV 1.04 1.06 1.47

Step: Vertical resolution: Height:
Degree of the trendline: Moving average : Color code?

Users' comments:

Author: Yuri Tsivian Date: 2007-02-08

The trendline at 6 gives a frequency picture of laughters.

Author: Yuri Tsivian Date: 2007-02-08

Change height to 500 to see Chaplin's laughter economy. Note that the hightest laughter intensity typically coinsides with the areas of the highest frequency. Am I imagining things, but this correlation between frequency and intensity reminds one of something I heard people say about coitus and orgasm. Cinemertics, brace yourself to measure hard porn!

Author: Torey Liepa Date: 2007-02-08

This is a fascinating alternative use of Cinemetrics. Particularly interesting is the fact that "Laughter" is nearly always bracketed by "Giggles," such that the audience rarely shifts from a "Chuckle" to outright "Laughter," but rather transfers progressively to "Giggles" first, though this could be in part a reflection of the measurer's perspective. I am assuming that "Giggles" are more emphatic that "Chuckles" though on the other hand each could refer to an entirely different mode of laughter altogether. Perhaps you could include "Chortle" in purple for those rare expulsions of cathexis found in slapstick comedy and This is Spinal Tap...

Author: Yuri Tsivian Date: 2007-02-08

By chucles I meant the low-level sounds, though as I heard them against the film I realized that these often are reactions to subtle jokes, visual gags, etc; not necessarily something inferior to louder giggles or still louder laughter. And yes, I notices that laughter must be built up by modest jokes.

Author: Kyle Westphal Date: 2007-02-08

A fascinating experiment...and one that I only figured out half-way through the film, much as I tried to look over Yuri's shoulder and understand why he wasn't clicking for each cut.

It also seems like the right way to approach Chaplin: he was a superb technician, but editing rhythm seems like one of the less interesting aspects of his work. Certainly if one is to think of his pacing and rhythm, the joke would be the more appropriate unit than the cut.

And for the record, I counted 23 students and I suspect that the print of A DOG'S LIFE was step-printed for 24fps.

Author: Gunars Civjans Date: 2007-02-08

This is one case where all degrees of trendlines give credible information. At degree 12 we just get a more detailed picture. Degree 6 doesn't fully register the "quickie" in the beginning, for example.

Author: Yuri Tsivian Date: 2007-02-08

On the other hand, Gunars, much as I struggled with the trendline for The Kid I could not get it to react to the two peaks of jokes towards the end of the film (perhaps these were, after all, more isolated laughs).

A quick clarification to Kyle's comments. Kyle Westphal who was looking over my shoulder yesterday was in a position to do so because he projected the films. He is also a student in that class but his contribution to the bursts of laughter does not count for he was in the soundproof booth. He is also a new Cinemetrics contributor.
I counted the number of students before the film, and yes, some came in later but I was too busy counting laughs to count heads.
I don't think the print was stretch-printed, one always detects the movement falsity, just speeded up to 24.

Author: Charley Leary Date: 2007-02-09

I'm a bit confused: does ASL mean anything here, i.e. is this measuring when a laugh begins and ends; or did you just click whenever you hear a laugh/giggle/chuckle?

On another note, coincidentally, I was teaching a class yesterday, and we watched one of my favorite Chaplin films: The Pawnshop. I could only count a handful of low-level chuckles. Viewing conditions were not ideal though...

Author: Yuri Tsivian Date: 2007-02-09

The ASL in this case spells ALF -- average laugh frequency, for I never clicked on cuts, only on audial sighns of appreciation. I'll go to Bologna this summer where the complete Chaplin will be shown at Cinema Ritrovato -- I plan to do this for the whole oeuvre with larger audiences, 35 mm restored prints and live music. I'll ask Gunars to design a special version of the Cinemetrics tool for this type of count, with less sonfusing terms. This was just a test, and by the way is worked when I showed the class the curve of their enjoyment. They clearly enjoyed studying their enjoyment.

Charley, don't despair at the lack of laughter for Charlie nowadays. He has had enough (though never thought it was enough). As I was measuring these two films for laughs, I realized that "chucles" are not necessarily inferior to "laughter" -- typically, they are a reaction to clever gags and sight gags. The Pawnshop (everyone's favorite) will yield less laughs than One AM or Work which are based on proven punches and falls (no less brilliant for that) -- for its best scene is based on transposition gags which ia an intellectual type of joke by definition. In addition, to fully understand what chaplin is doing there one must be familiar with the world of objects he is playing with -- and less and less students nowadays know that in 1916 clocks used to tick, empty cans were used for keeping angling worms, doctors used ears to listen to heart beats and other things that are lost on the young you teach. But that's what you are there for.

Author: Yuri Tsivian Date: 2007-02-09

Forgot to answer one thing: no, one click pes laugh, not the beginning+end of any laughter, which might be a thing to include in the future laugh-tool -- for indeed, some reactions had a duration to them.

Author: Charles O'Brien Date: 2007-02-12

The graph here, and also the one for The Kid, is extremely interesting! I think this particular use of cinemetrics is quite exciting because it may get close to how directors and editors of comedies actually thought about their work. In my research into early sound film I've come across trade press discussions of the need to hold off on introducing lines of dialogue so that the audience will have time to laugh at the joke. In too many films, as one French critic put it, "les acteurs n s'arretent pas pour laisser a chaque effet le temps de s'epanouir; ils continuent de parler, et le spectateur ne veut pas perdre une dizaine de repliques, il est bien oblige de reprimer les eclats de sa gaiete" (in Pour vous, 27 April 1933) [sorry, I can't get the accent marks into this text]. According to an American sound engineer, "A successful comedy is punctuated with laughs through which succeeding lines must penetrate. Thus dialogue must be recorded at a higher volume" (in American Cinematographer June 1934). If I recall, Lubitsch also discussed somewhere the need to pace the dialogue so that the periodic eruptions of audience laughter wouldn't obscure it. All of this suggests that the filmmaker is constructing the movie with a pretty clear idea of which moments will elicit laughter, and how much laughter, whether a chuckle, a giggle, etc..

Add your own comment:
Your name:

Enter code from above picture: