Category: Film speed and duration

A further consequence of the fact that transfers from film to video are often made at an incorrect speed is that duration values can also be misleading.

For example, if a film shot at 16fps and lasting, say, for 12 minutes is transferred to video at 24fps, then on playback, the video version will last for only 8 minutes.

Although the playback speed of video is relatively fixed at 24 or 25fps, depending on the region of the world, for reasons explained here, it is not possible to be so sure of the frame rate at which the film passed through the camera in early cinema, particularly if they were hand-cranked.

However, 16-18fps was a widely recognised norm, so as a rule-of-thumb, if one comes across a video version of a film in which the movement is obviously accelerated, it is probably appropriate to add 30% to the given duration in order to calculate the playback duration of the film in its original filmic form.

In French, there is a convenient term for this process – la remise à cadence. No such specific phrase exists in English, but in the posts on this website, we shall refer to this process as “adjustment for speed“.

When early non-fiction films are transferred to digital media (and previously to analog video media), the movement of the subjects often appears to be too fast. It is often assumed that it is just in nature of early film to be like this. In fact, it is only too easy to become so accustomed to this accelerated movement that one no longer notices it.

But this effect usually arises because in making the transfer from film to video, insufficient compensation has been made for the fact that the typical frame rate in non-fiction film-making prior to the Second World War was 16-18fps, whereas the standard norm for video technology is 24-25 fps.

In effect then, the movement being shown in the digital transfers is often around 30% faster than it was in reality. Some video player apps, including VLC, allow one to adjust the speed of the playback to compensate for this, though this will play havoc with any voice-overs or music that might have been added at the same time as the transfer to video.

 

 

 

It is on account of this variability in the frame rate at which films were both shot and projected in early cinema that in many film catalogues covering this period, it is the physical length of the films that is indicated rather than temporal duration.

However, for most readers of The Silent Time Machine website, physical lengths will mean little. When one learns that the Austrian ethnographic film pioneer, Rudolf Pöch, recommended shooting sequences of no more than 20 metres, what does that mean exactly? Or when Margaret Mead and Gregory Bateson are said to have shot an unprecedented 22,000 feet of film, it sounds like a great deal, but how much is it really?

In order to get some sense of the duration of these examples, it is first necessary to be aware that the Pöch was shooting on 35mm film (in which there are 16 frames per foot of film) while Mead and Bateson were shooting on 16mm (in which there are 40 frames per foot of film). If one assumes that in both cases, the films were shot at 16fps, this means that the duration of the ideal sequence for Pöch  was 65 seconds, while the Mead and Bateson material totalled just over 15 hours of footage.

These calculations of duration are based on the assumption that the material was shot at a frame rate of exactly 16fps which, for the reasons described here, is far from being necessarily the case. However, provided this caveat is born in mind, temporal duration is arguably a more meaningful indicator of quantity than a bald statement of physical length.

When sound film was first developed in the latter part of the 1920s, the norm of 24fps (frames per second) became established for 35mm feature films. In film studios, where the cameras were by now driven by electric motors rather than cranked by hand, camera operators would ensure that the cameras really did run at this speed since this was essential to achieve synchronicity with the recorded speech of the actors.

However, most non-fiction films were still being shot at around 16-18fps. But even though many non-fiction film-makers were by now using clockwork-driven cameras rather than cameras cranked by hand, it was still not possible to guarantee that the norm of 16-18fps would always be exactly achieved. Some spring mechanisms were better than others; all spring mechanisms tended to slow down as the spring got towards the end of its unwinding.

It would not be until after the Second World War – by which time, many documentary film-makers were shooting on 16mm film – that 24fps would become the norm for non-fiction film-making as well.

In modern television and digital technology, the standard frame rate norm is a minimum of 24fps in the US and some other regions, while in Europe, it is 25fps. This poses a potential problem when transferring early film shot at 16-18fps to video.

 

The likelihood of deviation from the frame rate norm of 16fps during shooting was so great that the  projectors of early cinema were usually fitted with variable speed dials which allowed projectionists to adjust the speed at which the film passed through the gate of the projector. By this means, they could ensure that the speed of the projector matched the speed at which the film had been shot in such a way as to produce normal speed of movement on the screen.

However, they would often intentionally set the projector at a speed that was at variance with the camera speed. If they wanted to speed the action up, say for a Chaplin slapstick comedy, they would increase the speed of the projector. For a romantic love scene on the other hand, they could slow it down a touch. Some directors would even stipulate different projection speeds for different reels of the same film depending on the subject-matter.

At other times, projectionists would speed a film up for the banal reason that they knew that it had to fit into a particular time-slot in the cinema programme. Making these speed adjustments was regarded as one of the most important skills that a good projectionist should have.

In the early days of silent cinema, there was a norm that film should pass through the gate of the camera at a frame rate of 16fps (frames per second). This was only slightly over the threshold rate of 10-12fps above which the human eye can no longer discriminate between the various independent images in a sequence and sees them instead as a representation of continuous movement.

As there were 16 frames in a foot of 35mm film, which was the standard film gauge in early cinema, this meant that the film was being passed through the gate of the camera at the rate of one foot per second.

Prior to the 1920s, most cameras used in non-fiction film-making were cranked by hand. They generally worked on the principle that one crank of the mechanism would send one foot of film through the gate. There was therefore supposedly a neat correspondence between cranking and frame rate: one crank per second produced the standard speed of 16fps. Camera operators developed various verbal formulas, which they repeated as they cranked, that were intended to help them crank at the standard speed.

However, in practice, there was considerable variation in the speed at which operators actually cranked. A problem noted by the ethnographic film pioneer, Baldwin Spencer, among others, was that there was a temptation to vary the speed of cranking in accordance with the speed of movement of the subjects [cited in Dunlop 1979, p.112].

But if camera operators cranked too fast, when the film was projected at the normative speed of 16fps, this would have the effect of slowing down the movement on the screen. Alternatively, if they cranked at less than 16fps, this would have the effect of speeding the action up when the film was screened.

A further issue was that while 16fps was generally regarded as a norm, some professional production companies simply did not acknowledge it, with some shooting systematically at higher speeds, and others at lower speeds.

Today, when footage originally shot at 16fps is transferred to a digital format, it frequently happens that the transfer is made at the now-standard 25fps, with the result that the actions of the subjects appear too rapid.