The eerie closing shot of
'The Devils' (1971)
is composed for 2.35:1 widescreen

Having decided what format to shoot the film on, the next crucial stage is determining the aspect ratio of the image - or its basic shape. This is expressed in the form of a width to height ratio, so if 1:1 is a perfect square, 2:1 will be a horizontal rectangle that is twice as wide as it is high.

Traditionally, virtually all films used to be framed in 1.33:1, or 4:3, producing a squarish rectangle. This was originally established in 1892, when Thomas Edison and William Kennedy-Laurie Dickson recommended that a film frame be the height of four perforations of a strip of 35mm film. Aside from some films shot in the Movietone process in the late 1920s and early 1930s (where the picture aspect ratio narrowed to 1.19:1 to accommodate the optical soundtrack) and a handful of widescreen experiments, 1.33:1 remained universal until the early sound era, and 1.37:1 thereafter (the latter generally known as 'Academy Ratio'). When television was introduced, it therefore made sense to adopt the same aspect ratio (almost universally referred to as 4:3, or "fullscreen").

However, it was television that provoked a change in cinema aspect ratios. After experimenting with 3-D and Cinerama, the film industry (represented, as ever, by Hollywood) began to favour 20th-Century Fox's CinemaScope process, which used an anamorphic lens to squeeze an image with a very wide aspect ratio (originally 2.55:1, though it narrowed to 2.35:1 when optical soundtracks were added) into a standard 1.33:1 frame. As a result, the picture would look vertically squeezed on the actual film image, but the application of a similar lens to the projector would recreate the original widescreen image. Rival widescreen processes to CinemaScope quickly emerged (Panavision being the best known), so for convenience's sake we shall refer to it as Scope and assume a ratio of 2.35:1.

The problem with Scope, aside from its incompatibility with television (see below) was that the image was so wide that it meant that cinemas often had to install new screens, with the result that traditional 1.33:1 images looked too narrow. Accordingly, even non-anamorphic films began to adopt widescreen aspect ratios, the most common being 1.85:1 (in the US) and 1.66:1 (in Europe). Because this system (sometimes referred to as "flat" widescreen, because it didn't require an anamorphic lens) made use of conventional 35mm film, the basic shape of the image was still 1.33:1. The widescreen ratio could be created in two ways, either by masking off the image on the print, by printing black bars at the top and bottom of the frame, or by printing a full 1.33:1 image and trusting the projectionist to frame it properly. However, an advantage of the latter system was that it allowed films to be framed in multiple ratios between 1.33:1 and 1.85:1, depending on what the cinema was capable of handling.

The need to remain compatible with 1.33:1 grew acute in the 1960s and 1970s, as television (and, eventually, video) became an increasingly important secondary market for films. Unfortunately, films shot in Scope ratios were impossible to screen adequately on television without either presenting it as a thin strip across the middle (generally considered unacceptable by broadcasters) or cropping the image at the sides. Originally, this process was crude in the extreme, as the telecine machine simply transferred a 4:3-shaped chunk of the middle of the frame, which in the case of a shot with two people occupying each end of the frame might well end up producing the effect of an empty room. A more sophisticated form of cropping came with the invention of the "pan-and-scan" process, whereby the telecine operator would move a 4:3-shaped frame across the original image in an attempt to capture its most important elements. Panning and scanning was almost universally used to present Scope films on television until the 1990s, though flat widescreen films were sometimes shown "open-matte" in the original 1.33:1 image. This was certainly preferable to cropping, though it created new problems in that the full frame sometimes revealed more than the cinematographer intended, such as boom microphones at the top of the screen that were not intended to appear in the widescreen version. It is usually possible to spot an open-matte transfer by the increased amount of headroom.

All these issues had an effect on image composition. In the 1950s and early 1960s, known as the "high Scope" era, cinematographers and directors took pride in exploiting the full width of the frame. From the late 1960s, and especially with the onset of video a decade later, cinematographers were much more careful about ensuring that the image wouldn't be completely wrecked if cropped to 4:3, a sensible precaution given that the majority of the audience would probably be watching the film in that aspect ratio. Accordingly, widescreen compositions in general and Scope compositions in particular tended to group important information together in such a way that it wouldn't occupy more than half the frame at any given time.

Things changed in the 1990s with the introduction of a wider television and video aspect ratio. Generally referred to as 16:9, this translates to 1.78:1, which is a convenient halfway house between the 1.66:1 and 1.85:1 flat widescreen aspect ratios. With the advent of DVD as the consumer video medium of choice, it is now far more common to present films in the original aspect ratio (using black bars at the top or sides where necessary), and consumer awareness of these issues has also grown.