Video has a look unlike film. For many it is too "real" and lacks the gloss, saturation and "mystic" quality that is associated with feature film production. In truth both mediums have their uses and video can be used to great effect to convince viewers of the reality of a subject. However, many film makers using video want to achieve a film-like look, and this article offers some help and advice in achieving it.
The primary technical differences between video and film are ...
This is the one area we can do little to change. All digital video has a fixed pixel resolution and film is continuous tone. We have to live with this limitation, but future improvements in video resolution (HD) will narrow the gap.
Standard 35mm film employs a frame rate of only 24 fps, whilst digital video uses either 25 (PAL) or 29.97 (NTSC). This difference means that movement in video is smoother than film. We associate this smooth movement with video not film. Although the difference between 25 and 24fps is negligible, it is worth converting NTSC to 24fps. Frame rates can easily be changed within an editing application.
During shooting, the length of time each frame of video or film is exposed for is determined by the shutter speed. The standard shutter speed for film is twice the speed of the frame rate; 24 x 2 = 48 = 1/48th of a second. Setting shutter speeds longer than this will increase the blurring of each frame image (especially for fast action) and make the pictures look less film like and more like interlaced video. If possible, try setting a shutter speed on your video camera which is twice the frame rate, so for PAL 1/50th of a second and for NTSC 1/60th of a second.
In a non-progressive scan video camera, each frame consists of 2 interleaved (or interlaced) "fields". In simple terms ... imagine each frame or picture on your TV is made from a number of horizontal lines drawn close together across the screen (actually it is, this is how cathode ray tubes work). If you were to number them from top to bottom 1, 2, 3, 4, 5, 6, 7 through to 576 (for PAL) then the first field would consist of lines 1, 3, 5, 7 etc and the second lines 2, 4, 6 etc. PAL DV has a resolution (frame size) of 720 x 576 therefore each field is effectively 720 x 288.
These fields are "drawn" onto the screen one after the other effectively making the frame rate of PAL DV 50 "half" field frames a second and NTSC approx 60 "half" field frames a second. It is this interlacing that is most responsible for video's "real" look.
The easiest way to avoid interlaced video is to use a camera that films in progressive scan mode! If you don't have access to such a camera you can de-interlace (remove one of the fields) once the video once it is captured. The problem with this is that you will lose half the detail. Therefore, many transcoding and optimising systems have a de-interlace option that you can try. The following are alternate possible solutions ...
In your video editing software create 2 video tracks and place copies of the same clip on both. Reduce the opacity of the top clip to around 25% and de-interlace the clip on the track below with a filter (FCP has one which discards one field and duplicates the remaining one). The top clip will replace some of the missing resolution and detail de-interlaced from the bottom but the overriding look should be more film-like.
Smart de-interlacing works by analysing each frame and working on only those parts of the image that won't produce visual side effects. This cannot be achieved manually and requires a dedicated filter (see later).
Many film looks are achieved by "grading" the film in post production. These can usually be achieved with video by changing the colour balance of clips with filters. Effects like de-saturate, colour correction, Curves, bleach etc will all help produce the look of film. Dedicated 3rd party filters can also help (see later).
Film also has a greater dynamic range (the range of tones from black to white) than video, and it's therefore worth experimenting with a filter (levels) to deepen blacks and whiten white highlights.
Film makers rarely use the 4:3 aspect ratio of standard TV so video makers should film and edit in wide screen or 16.9 anamorphic modes (most cameras and editing applications support this) or simple crop their video in the edit (obviously losing some resolution) Click here for more info.
In film, it is common for the cameraman to direct your eye around the frame by using a shallow depth of field and "pulling focus" ( a focus-pullers job!) to bring parts of the frame into focus whilst leaving the rest of the frame blurred. A classic example is when the director wants you to pay attention to something in the background and then switch to something in the foreground within a the same shot.
Some video cameras have either sensors (CCDs or CMOS which convert light into digital data) which are too small (resulting in all the frame always being in focus) or inappropriate lenses, which make achieving a shallow depth of field difficult. If your camera has an APS-C sensor, or is an SLR (typically with a larger sensor), you should be able to achieve a film like shallow depth of field.
However, even if you have a camera with a large enough sensor, changing focus manually without shaking the camera is challenging. In film production, a focus puller has a remote which controls the motor in the lens.
Apple's QuickTime Pro includes some effective colour and film distress filters such as "Film noise".
G Filters from www.nattress.com. This suite includes dedicated de-interlace and colour filters to help achieve a film look. There's a demo but it's well worth the £60.
Excellent bundle of plug-ins. A must for the serious video film maker who wants to achieve a great film look. www.redgiantsoftware.com/products/all/magic-bullet-suite/