Identifying Fatty in a Shirt?

[Yuri Tsivian]

Hi, this is Yuri Tsivian of U of Chicago. Daria Khitrova and me spent 2 days at NFA in Prague measuring shot lengths for Czech release version of Keystone films. All Czech titles but one presented no problems identifying. The only one we are still working on is a Fatty film whose Czech title is "Fatti v kosil" (Fatty wearing a shirt). In it Fatty appears wearing a (lady's) swimming suite; later he is shown taking a roller coaster (with camera mounter on roller coaster trolley at times).
Any tips?
Another question, possibly for Paul Gierucki. The viewing table we used was set at 24 fps, which is probably too fast even for slapstick comedies. What speed did you choose to transfer Keystone films in your video-releases? Many seem quite right. Thanks!

[Richard M Roberts]

Hi, this is Yuri Tsivian of U of Chicago. Daria Khitrova and me spent 2 days at NFA in Prague measuring shot lengths for Czech release version of Keystone films. All Czech titles but one presented no problems identifying. The only one we are still working on is a Fatty film whose Czech title is "Fatti v kosil" (Fatty wearing a shirt). In it Fatty appears wearing a (lady's) swimming suite; later he is shown taking a roller coaster (with camera mounter on roller coaster trolley at times).
Any tips?
Another question, possibly for Paul Gierucki. The viewing table we used was set at 24 fps, which is probably too fast even for slapstick comedies. What speed did you choose to transfer Keystone films in your video-releases? Many seem quite right. Thanks!

FATTY WEARING A SHIRT sounds like CONEY ISLAND (1917), but that seems rather obvious. Do you spot Buster Keaton or Al St John in it? How long is the film?

24 fps is definitely NOT too fast for silent comedies, but is in fact, the correct speed for most silent comedies. That said, the best frame rate for most Keystone Comedies is around 20-22 fps, especially Keystone-Triangles, which seem to be really undercranked.


RICHARD M ROBERTS

[Yuri Tsivian]

Thank you. Actually, while I was writing this, Daria found the wanted title: The Bath House Beauty.
The speed problem we are trying to solve is to find a common frame-rate denominator for films measured on Cinemetrics. Some have been measured from transfers done (or viewing table run) at 24, some from transfers at 18 or even 16 fps. Paul Gierucki's Fatties, for instance, seem to move without much over-acceleration. So if we find out, what is a typical transfer films used for commercially released videos we will be able to equalize speeds and compare individual films' cutting rates.

[Bob Birchard]

Thank you. Actually, while I was writing this, Daria found the wanted title: The Bath House Beauty.
The speed problem we are trying to solve is to find a common frame-rate denominator for films measured on Cinemetrics. Some have been measured from transfers done (or viewing table run) at 24, some from transfers at 18 or even 16 fps. Paul Gierucki's Fatties, for instance, seem to move without much over-acceleration. So if we find out, what is a typical transfer films used for commercially released videos we will be able to equalize speeds and compare individual films' cutting rates.

There is no doubt that silent comedies (and dramas too, for that matter) were intended to be screened at a somewhat accelerated pace. The SMPE standard established in the mid 1920s called for shooting at 20 fps and projecting at 24 fps--but not everyone adhered to these standards. As late as 1926 Fox was still shooting at something in the neighborhood of 16-18 fps on pictures like "The Great K & A Train Robbery," while a number of late silent M-G-M films were intended to be run at 100 feet per minute--faster than the 90 feet a minute achieved at 24 fps.

Like D. W. Griffith, the Sennett Keystone studio tended to severely undercrank its films--In part this was to cram more story onto a single reel of film, but it was also to be able to achieve purposely accelerated action at typical projection speeds of the time. So, Sennett's camera operators were often shooting at speeds of 12 to 14 fps to achieve accelerated action at then (1914) common projecting speeds of 16 fps to 18 fps.

The Chaplin at Keystone DVD collection (all films shot in 1914) films were transferred at 16 fps or 18 fps depending on the film, and to my eye look about right with subtle actions able to be seen and appreciated in a way that they cannot be at higher frame rates. I know others may disagree. I'm not looking for an argument. I'm only passing on information. Let your own eyes and sensibilities be the deciding factors.

[Richard M Roberts]

Anyone who comes around here saying the speeds on the CHAPLIN AT KEYSTONE set look right to them is going to get an arguement whether they like it or not. 16 to 18 fps is too damn slow for the Keystones or nearly ANY silent comedy. The "subleties" one thinks they are missing at a faster frame rate are meaningless if the pacing and comedy timing of the whole short are ruined, as they are on the CHAPLIN AT KEYSTONE set. There is actually a pile of evidence to prove that most silent films, especially after the early 20's, and for comedies, pretty much after the mid-teens, were shown at speeds nearing 21-24 fps and by 1923-24, were run even faster than 24 fps in an interestingly sizeable number of cases. If your really interested, I will be happy to give you that proof privately as this is an old arguement that the folk around here have gone over many times.

Paul may chime in as to the Keystone set, but as Associate Producer of the set (and as a large number of the prints Paul used came from my collection), I can tell you that the speed used on most of the early Keystones was in the 20-22 fps range, and once we got to the 20's, the speeds went up to 24 for the most part. Paul is a stickler for getting it right, and I'm right behind him with a stick as far as speeds are concerned.

THE BATH HOUSE BEAUTY makes sense, and is indeed a rare Keystone title.


RICHARD M ROBERTS

[Yuri Tsivian]

Thank you, Richard and Bob, all this makes a lot of sense. Based on what you are saying, we will calculate average shot lengths at 20 fps.

Yuri

[Uli Ruedel]

For commercially released DVD transfers, would it not be possible to find a sequence with lots of movement, and go through the video frames step by step, while counting the actual/original film frames equivalent as evidenced by the motion? If this is done for a few seconds of video, maybe a reasonably accurate transfer speed could be calculated.

Uli

[Richard M Roberts]

For commercially released DVD transfers, would it not be possible to find a sequence with lots of movement, and go through the video frames step by step, while counting the actual/original film frames equivalent as evidenced by the motion? If this is done for a few seconds of video, maybe a reasonably accurate transfer speed could be calculated.

Uli

Measuring shot lengths? Oh God, don't tell me this whole conversation was done because someone is doing "cinemetrics" on these films.


RICHARD M ROBERTS

[Yuri Tsivian]

Exactly, this is what took Daria and me to Prague, as I wrote in the first post. One of the questions we wanted (and to an extant, managed) to figure out was what typically changes when an American comedy is released in Czechoslovakia, Soviet Russia, or France. Normally, these films were re-cut for ideological or (more often) commercial reasons. One general trend was to cut or shorten shots that appeared vulgar, nonsensical or politically incorrect. Of course, one could analyze these changes using the time-honored film studies prose, but this seemed a little too predictable to justify a project that involved many library hours/dollars (to browse and copy the Sennett paper collection at the Academy Library) and archival trips (UCLA and NFA in Prague are the first steps). What struck us in the Keystone continuity and cutting sheets was the meticulous attention to how many feet this or that "scene" or "leader" must take no to compromise the tempo of action (this tempo being something like a holy cow with Sennett). We copied copiously all these footage figures thinking that nothing that was so material for writers and directors at Keystone around 1913-1915 could be dismissed as immaterial by the historian of film. On the other hand, what can one do with all these numerous hundred-year old numeric data?
Perhaps it might make sense, we figured, to average these figures and compare these to average cutting rates habitual at other studios (and genres) active at the time. How fast or slow compared to Keystone were cutting rates of Pathe comedies with Max Linder? Were Sennett's comedies at Keystone faster or slower that his earlier comedies shot at Biograph? How does Sennett's cutting style compares with Griffith's, his filmmaking mentor's?
If these questions make sense (and to us they do) a good place to go to is Cinemetrics where they know how to handle shot-related figures, and have a solid database from which to pick existing data for comparison, for instance, Max Linder's comedies measured earlier by Charles O'Brien. Here, however, an obstacle arose which made my to subscribe to the Silent Comedy Mafia forum -- and I am glad I did, for now, we have an answer. The obstacle, as you know, is that the average shot length changes depending on the speed with which this or tan film was projected or video-transferred. To compare, say, cutting rates prevalent at Keystone and Biograph it makes sense to "normalize" the speeds choosing one as a common denominator for all numbers (I spare you the math of this, but this is doable knowing, as we do, the footage and the number of shots for every film in question).
Thank you all for helping us with this, 20fps makes sense a good compromise speed, and conveniently in the middle between 16 and 24, both too extreme! Cheers, mafia does good thinks too!

[Richard M Roberts]

Exactly, this is what took Daria and me to Prague, as I wrote in the first post. One of the questions we wanted (and to an extant, managed) to figure out was what typically changes when an American comedy is released in Czechoslovakia, Soviet Russia, or France. Normally, these films were re-cut for ideological or (more often) commercial reasons. One general trend was to cut or shorten shots that appeared vulgar, nonsensical or politically incorrect. Of course, one could analyze these changes using the time-honored film studies prose, but this seemed a little too predictable to justify a project that involved many library hours/dollars (to browse and copy the Sennett paper collection at the Academy Library) and archival trips (UCLA and NFA in Prague are the first steps). What struck us in the Keystone continuity and cutting sheets was the meticulous attention to how many feet this or that "scene" or "leader" must take no to compromise the tempo of action (this tempo being something like a holy cow with Sennett). We copied copiously all these footage figures thinking that nothing that was so material for writers and directors at Keystone around 1913-1915 could be dismissed as immaterial by the historian of film. On the other hand, what can one do with all these numerous hundred-year old numeric data?
Perhaps it might make sense, we figured, to average these figures and compare these to average cutting rates habitual at other studios (and genres) active at the time. How fast or slow compared to Keystone were cutting rates of Pathe comedies with Max Linder? Were Sennett's comedies at Keystone faster or slower that his earlier comedies shot at Biograph? How does Sennett's cutting style compares with Griffith's, his filmmaking mentor's?
If these questions make sense (and to us they do) a good place to go to is Cinemetrics where they know how to handle shot-related figures, and have a solid database from which to pick existing data for comparison, for instance, Max Linder's comedies measured earlier by Charles O'Brien. Here, however, an obstacle arose which made my to subscribe to the Silent Comedy Mafia forum -- and I am glad I did, for now, we have an answer. The obstacle, as you know, is that the average shot length changes depending on the speed with which this or tan film was projected or video-transferred. To compare, say, cutting rates prevalent at Keystone and Biograph it makes sense to "normalize" the speeds choosing one as a common denominator for all numbers (I spare you the math of this, but this is doable knowing, as we do, the footage and the number of shots for every film in question).
Thank you all for helping us with this, 20fps makes sense a good compromise speed, and conveniently in the middle between 16 and 24, both too extreme! Cheers, mafia does good thinks too!

This is a total and utter waste of time.



RICHARD M ROBERTS