The Unknown (Tod Browning,1927) - framerate?

[Jonathan Greer]

Hello all,

We're due to screen this silent film and I'm trying to decide on what framerate to show it in.
I found conflicting information online from two different silent film festivals, San Francisco and Pordenone. San Francisco gives the running time as 75 mins - our print is 4435 feet so we'd have to project it at under 16fps which looks too slow to me to achieve this running time. It seems the Pordenone print is a longer restored version, 5235 feet which they played at 21fps for a running time of about 66 mins. This speed looks a bit too fast to me, I thought about 19fps looks "about right" but there's some leeway either side.
I don't suppose anyone has shown this film or might have information on the framerate? I saw in another thread (Position of optical sound on film in Switzerland and France) that Leo Enticknap mentioned

"Many if not most silent films were made in the absence of consistently applied technical standards determining the frame rate for shooting and projection. Often they were shot at one frame rate, with the expectation of being projected at another. Different shots within a movie were frequently taken at different frame rates. Some frame rates in shooting are physically impossible to reproduce in conventional projection (e.g. The Birth of a Nation, most of which was shot at 11-12 FPS).

Name me even one moderately well known feature made between around 1910 and 1925, and I could likely find you long running arguments between professors, archivists, and silent movie geeks as to what the "correct" projection frame rate should be, even though the correct answer is that there isn't one - just conflicting, informed opinions."

1927 is a bit outside this window so I'm hoping there might be a bit more consensus on the fps for this one...

[Randy Stankey]

I understand it the way you do. In the early days of cinema, there wasn't a set frame rate. Photographers and directors just shot movies at whatever frame rate they thought was good and there was little expectation that a movie would be projected at the same frame rate it was shot at.

In the early to mid 1920s, film without sound ran at 16 fps. Sound on film was being developed at around the time of the middle to late 1920s and, by the 1930s, was becoming common. During these times, film speeds were changing from 16 fps. to 24 fps. That means your movie comes in the "in-between" times. Because of that, I don't think that there is an official frame rate. I think your movie was still in the "whatever looks good" era even though film speeds were creeping toward 24 fps.

I like your idea of estimating the playback speed using the running time but, since so many old movies have been cut and re-cut, some multiple times, it's hard to know what the running time of your actual print should be. That puts us back to guestimating.

If I was in charge of showing your movie, I'd split the difference and test run the print at 20 fps. to see how it looks and make a judgement from there.

Other than that, do what you think is best and, if the film geeks complain, explain your reasoning but, if they don't like it, tell them to go get their own projector and show it the way they think is right. That's what theater owners in 1927 would have done.

[Jesse Crooks]

If you're lucky you might be able find the intended speed (in feet per minute, not frames per second) mentioned in contemporary trade publications. I've never done that research myself, but I've had accompanists point to that when requesting a specific frame rate.

[Jim Cassedy]

The Unknown (Tod Browning,1927) - framerate?[/h]

It's unknown! (sorry, I couldn't resist) I may be asking the obvious, but isn't the archive or library you got the print
from any help in giving you frame rate info?
As Randy said, I suppose you could get a rough idea from the running time- - but he also points out the 'gotcha's' of
doing it that way. I often like to browse through really old issues of Motion Picture News, or Exhibitor's Herald from
the late 1900's & early 20's, and in all of their reviews and listings of current releases, they never published actual
running times, they'd only say something like "3-Reels" or they would list the film length in feet, since there was
no real projection standard speed back then. However, in equally early issues of American Cinematographer,
as well as in some early camera manuals I read, they state that the "standard" recommended camera speed
was 16fps, but they's also state that there was a lot of leeway to vary from that depending on the 'nature of the
scene' and sometimes to compensate for tricky over or under lighting situations. I believe the the hand cranks
on the old Bell & Howell 2709 cameras (the workhorses of early Hollywood) were geared so that one turn of the
rank exposed 1 ft of film, and so all the early cameramen got quite qood at rotating the crand at 1 revoloution/sec,
which was more or less the standard speed, back when there weren't really any standards.

The last time I was faced with a silent film speed conundrum, I was saved by the fact that the movie was going
to be accompanied by a full friggin' orchestra. They had played the score for the movie in question on several
occasions, and it was the orchestra leader who helped me arrive at the projection speed during one of the
rehearsals, since they knew the score & the proper tempo for their music. (It turned ot to be 20fps)
I used to run a lot of silent stuff at several festivals, and one thing I learned from one of the maestros is that
if you are projecting with a live accompanyment - especially witha full orchestra, the projectionist almost
becomes one of the musicians. Sometmes you have to 'fudge' a change-over by a few frames, to have it
occur right on a partcular beat, or note. (With that thought in mind, I once suggested that in addition to my
IATSE membership, I should also join the musicians union to so thaty I could paid twice when running
accompanied flicks at festivals. That's probably why they don't hre me anymore...!)

[Phillip Grace]

HI Jonathan. Your quote from Leo Enticknap and your own approach to the problem are spot-on.
Very occasionally a print will arrive from a reputable archive with a recommended frame rate but those are the exceptions. In a more typical case I think it's all bets off until you see it on the screen and choose the appropriate projection speed yourself. I aim for the best (most plausible) human motion, and try to take into account the emotional intent of a series of scenes. It is generally based on one spool of a feature, screened in silence, and applied to the whole film. Not best practice, but rehearsals are unheard-of in my situation. If the film is to be screened with a score the music will influence the psychologial perception of motion. You may have to run a little bit faster than you first thought to achieve a pleasing match.
Just caught sight of Jesse's post. Certainly the musicians will much better acquainted with the film than the projectionist, and will be best placed to nominate a projection speed or speed range.

[Leo Enticknap]

In addition to all the above, there is a pragmatic consideration as well: if your projectors don't have a three-blade shutter, then depending on the blade size of your two-blade shutter, a flicker will become visible once you go below the 19 to 22 range, very roughly. It will certainly look seriously bad as low as 16.

As a very rough rule of thumb, the big Hollywood studios were cranking at 24 by 1925-26, whereas European ones stayed at 16-20 until the conversion to sound. The main reason was likely that by this point, large theaters were being built with big auditoria and long throws, in which the light loss from a three-blade shutter was becoming a problem. By 1927, Vitaphone was already established, and that was locked at 24.

The last time I seriously interacted with them (in the 1990s), the British Film Institute / National Film Archive had a habit of recommending frame rates for mid to late 1920s titles that looked pretty good for British, French, and German movies, but were clearly too slow for Hollywood. I have an unpleasant memory of once seeing The Wind (the 1928 Lillian Gish pic) at the National Film Theatre, in which, thanks to a frame rate that was clearly way too low, the climatic tornado at the end looked like not much more than a light breeze, and all the characters were staggering around like they were whacked out on some illicit substance or other (which, given Hollywood in the '20s, they likely were, but a higher frame rate would still have helped them). A lot more recently, when I projected The Black Pirate at the Egyptian in around 2016, the Fairbanks expert who introduced it instructed me that it had to play at 24. I was a little skeptical at first, but after looking at the first reel ahead of the show, agreed with her completely.

Unless your musician(s) have rehearsed to a given frame rate (per Jim's anecdote) and need it, my advice would be to play a reel and go with what looks right to your eyes, keeping in mind that in some shots, the filmmakers may have undercranked for dramatic effect.

[Steve Guttag]

My rule of thumb, in absence of any documentation to the contrary, or any timed (e.g. metronome to live accompaniment that dictates the speed), I look at people walking in the movie. It's hard to describe but one gets a sense of what looks normal. This, of course, does not include deliberately over/under-cranked scenes for "effect." A nice thing with inverter drive machines is the ability to make, on-the-fly changes while evaluating. Previously, split-sheave pulleys, and the like, could make things much more difficult to "find" the speed. Inverters also allow for changing speed mid-movie for over/under-crank cues.

To add to Leo's note about the slower speed and flicker...please consider, as you slow the projector down, you are leaving that frame exposed to whatever light source/heat you are using for a longer period too. Your chances of the various heat damage (embossing, curl, blister...etc.) all go up geometrically as the speed slows. So, those in the silent movie business should really be running 3-wing shutters or the equivalent (e.g. Kinoton "E" projectors that automatically goes to 3-wing mode below 20fps). Those with mechanical fire shutters may also have to allow for the slower speed (Simplex can use a weaker spring on the governor and in a pinch, remove one regular spring).

[Jonathan Greer]

Many thanks for all the useful replies. Apologies for the late response, at the other end of the technological spectrum Avatar 2 (and its 1000 different versions) has been a nuisance (Disney neglected to send 3D HFR test films to some independent cinemas including us - turns out this would have been useful! - sorted now).

I have heard back from the BFI: "21fps according to George Eastman House who restored it" - which is the speed Pordenone played it at. What puzzles me is that San Francisco give the running time as 75 mins, and the screening was accompanied by the same musician as ours. Could just be a typo of course, I've emailed him to see what he thinks.

[Leo Enticknap]

If this is a polyester print on 2302 stock (or equivalent), and has not been tinted or toned, I would doubt if the extra heat from a sub-24 frame rate with a two-blade shutter would be a huge worry, unless you have an enormous (as in, 6-7kW) lamp, and no heat mitigation such as a filter glass and/or liquid cooled gate in place. Polyester stock acts as a pretty efficient heatsink and is resistant to passing that heat through to the emulsion. My rough and ready heat check was to allow the film coming out of the bottom of the projector to pass lightly between the side of my index and middle fingers for a couple of seconds: if it felt more than barely warm, that indicated a problem.

Older acetate is obviously a lot more fragile.

Apologies for the late response, at the other end of the technological spectrum Avatar 2 (and its 1000 different versions) has been a nuisance (Disney neglected to send 3D HFR test films to some independent cinemas including us - turns out this would have been useful! - sorted now).

There can't be many projectionists who are having to deal with both sub-24 silent and dgital HFR frame rates in the same week!

[Steve Guttag]

Leo,

• In the silent era films, it isn't just sub-24fps to worry about...it is sub 18fps. The dwell time increases and the time it takes to damage film gets progressively easier.
• The emulsion faces the lamp so I don't see how the base (polyester) will help with emulsion related heat. Furthermore, if these are on B&W stocks, they are going to soak the heat on the emulsion even more.

[Leo Enticknap]

We are not talking about sub-18 for a Hollywood movie made in 1927. Even 20 is likely too slow.

I have never seen any embossing or other overheating damage on any 2302 print, even those of very early cinema titles that likely were projected at 16 or maybe even slower before I handled them. I've noticed that a lot on acetate b/w prints, though, even ones made as recently as around the turn of the (21st) century. That's not to suggest that common sense precautions shouldn't be taken, but in the absence of unusual conditions, I don't think that the additional heat when using a talkie era two-blade shutter going down to the 18-20 ballpark is enough to risk damaging a polyester b/w print, unless possibly if it's been tinted and/or toned (can silent era toning dyes be absorbed by polyester anyways?).

In my Egyptian days, there was talk of playing two reels of a surviving nitrate print of Traffic In Souls for a festival, that their owner (a private collector) insisted had survived in near new condition, shortly after the booth had been refitted and brought up to code for nitrate. As this would have necessitated going down to 16, on projectors with 6kW lamps (albeit with heat filter glass in the lamphouses and liquid-cooled gates in the projectors), we all thought that this would be a risk too far, especially given the unique and irreplaceable nature of the print.

[Jim Cassedy]

Several years ago, after running a restored silent film that was 'making the rounds' at one of the
aforementioned festivals, I got an urgent call from the projectionist at the next venue who was
having a brain-freeze trying to figure out how to handle the change-overs. He wanted to know
"where to thread up" , because the film was running slower, it meant that there was going to be
more than 8seconds between cues and so he figured that meant he was going to have to give
himself more leader-time. I tried explaining several times that he should just thread up normally,
since cues were placed by footage, not time, and that 10feet was 10feet, no matter what the
frame rate. It was still TEN FRIGGIN' FEET- - it was just going to take longer to get to the
change-over cue once you started the motor. When the call ended, he still didn't sound 100%
convinced that he didn't need to do anything differently. (on an Academy leader) - As I said, I
think he was overthinking the problem & having a bit of brain lock. About 30min after our call,I got
a text message that said something like " NOW I finally understand what you were trying to tell me"
I knew he'd figure it out sooner or later.

[John Eickhof]

I have run different silent films and if I was to say an average speed (at least for pre 1920) it would be about 18 frames / sec. It also must be considered to account for the musician or musicians as well as tempo and theme of the film as well as instrument(s) used, ie: pipe organ, pit machine or orchestral , and the complexity of the music / effects scripted. I used to install a Bodine brush style 1/4 hp 1800 rpm pony motor onto the existing soundhead coupled with gilmer belts and electronic speed control with pullies set for center film speed of 18 fps at 75% of full pony motor rpm, this would allow speed to increase or decrease 4-6 fps as was suitable after some rehearsals. I had a voltmeter to monitor motor voltage at correct speed for the average show, and the existing soundhead motor was not used to drive the machine, it was just an extra flywheel.

[Leo Enticknap]

...because the film was running slower, it meant that there was going to be more than 8seconds between cues and so he figured that meant he was going to have to give himself more leader-time.

The only gotcha I can think of there is if the motor's ramp-up time is not proportional when full speed = 24 versus a lower frame rate, which could be the case with a computerized inverter that allows you to specify the ramp-up time. If the target is 18 (say) and the same ramp-up time is specified as for 24, the threading sweet spot will be different. For example, at one place I worked, midway between 9 and 10 in the gate was perfect for 24 (assuming that the entire leader was intact), but if you threaded there for 18, you would see about half a second of black on the incoming leader after the changeover: 9 on the nail was the new sweet spot.

What can confuse projectionists unfamiliar with silent prints is if the leader does not indicate the track side, figuring out which way round to thread it. I've been in the audience for a couple of shows in which an intertitle has suddenly appeared in mirror writing immediately after a changeover.

[Phillip Grace]

There is another funny human quirk to do with placement of the changeover cue. According to some texts, at a notional projection speed of 16fps it was thought that 1/2 second was adequate reaction time for an alert operator to change - hence 8 frames of run-out after the last cue dot. At 24 fps Academy Standard cues have 18 frames of run-out, and the SMPTE Universal spec 24 frames - a whole second. So there is a tiny mental adjustment to make according to where the cues are placed (temporary or printed), the running speed, and how fast you need to react to the changeover cue. Too quick might be too soon.
For the sake of consistency I usually make up a show with temporary (chinagraph) cues at Academy spacing, and a temporary start mark on the head leader, and work out the running speed later.
Threading. In 35mm projection the emulsion always faces the lamphouse - in that case, the worst that can happen is that the picture appears upside down!