Making digital video look like film is one of the issues at the forefront of every independent filmmaker’s mind when preparing a video production, and that is the way it should be, because the production value of a film makes a big difference to how the audience perceives it.
It is worth starting with an interesting but not entirely surprising fact: audiences, no matter how unfamiliar with the filmmaking process, can always judge the production value of a film. They may not be able to articulate it; they might not walk out saying how the color palette did not serve the story, or how the lighting was excessively high-key, but they instinctively feel that something (or a lot) wasn’t right. And that is a major barrier between the audience and the story you are trying to tell.
Why is the film look so desirable? 35mm film has a higher resolution than digital video, but that is probably the least important issue. The texture, color subtleties, grain structure, contrast rendition and motion characteristics of film are a far more important factor. Here is what Janusz Kaminski, DP on all of Spielberg’s films since “Schindler’s List,” has to say about it:
“I like looking at movies and seeing the movement of the grain. There’s a certain emotional response, whether viewers are conscious of it or not; we distance ourselves emotionally from images that are crisp and clear. Grain reminds people of their experience and gets them closer to the story.”
(From “Going with the Grain” – an article by John Calhoun)
This is partly why we perceive celluloid in one way and video in another. Kaminski is, of course, referring to digital video when he mentions “images that are crisp and clear.”
So how do we shoot a movie on digital video and approximate the look and feel of film as much as possible? Video currently cannot be a real substitute for film, but it is possible to shoot in such a way as to make it approximate the look of film. This article will deal with the salient issues and how best to tackle them.
Please note: This post will focus on (i) shooting with a shallow depth of field, (ii) tweaking camera settings and (iii) color correction / grading. I have dedicated a separate post to the big topic of lighting in digital cinematography.
Camera settings: gamma, sharpness, frame rate and shutter speed
Almost all professional-grade video cameras now have menus that allow the user to tweak the look they produce. This can be of great use in improving the look of the videos you shoot.
Many cameras allow the modulation of sharpness enhancement. Have you ever noticed how video tends to add an artificially enhanced edge around objects? This aims to make the footage look sharper than it it is, but it is misguided, because it actually makes the footage look worse. It is an ugly artifact. Many cameras allow the user to choose the level of sharpness enhancement. You should reduce the electronic sharpness enhancement or eliminate it completely. The resulting video will have a more visually pleasing and subtle look instead of the ugliness produced by the artificial edge enhancement.
Professional cameras also allow the modulation of gamma, which is essentially the rendition of the mid-tones. I recommend you shoot test footage of the same subject with exactly the same lighting, using a different gamma setting each time. Then you can choose which setting produces the look you want.
Many cameras these days also offer a preset referred to with names like “Cinelook”. These presets can work very well and are worth considering.
Many videographers claim that shooting at the slowest shutter speed, generally 1/30th or 1/60th of a second, produces the most film-like footage, due to the slight blurring. Remember that film has a shutter speed of 1/50th of a second, so this argument is a valid one.
That said, using a slightly faster shutter speed can produce a very slight strobing effect that is also very cinematic. I have personally used shutter speeds of 1/120th of a second and can confirm that the look is very cinematic – arguably more so than with slower shutter speeds, due to the slight strobing that we normally only see with motion picture film. As always, test everything to find what works for your tastes.
Motion picture film runs at 24 frames per second (24 fps). NTSC video has a frame rate of 29.97 fps and PAL video has a frame rate of 25 fps. In general, the closer the frame rate is to 24 fps, the more film-like the motion signature will be. NTSC video tends to look glassier and more video-like than PAL video, due to its high frame rate and lower resolution.
If you are going to shoot a project on video and are serious about transferring to film later, PAL is definitely superior to NTSC.
In recent years 24p cameras have been released. 24p stands for “24 frames per second, progressive” – in other words, instead of the interlaced footage of regular video, they take 24 discrete digital photographs per second, and therefore are the true digital equivalent of motion picture film. Please note, however, that the color and contrast rendition of all digital formats, including HD, is still nowhere near as good as that of motion picture film. The Panasonic DVX100 was one of the first 24p “prosumer” cameras to be released.
Deinterlacing video footage to simulate the look of film / 24p
The look of 24p can be faked in post-production by deinterlacing the video footage. Simple deinterlacing blends the fields, using interpolation, giving 30 fps with NTSC video and 25 fps with PAL video. Since half the lines are discarded and replaced by interpolation, you are essentially losing 50% of the vertical resolution. Nevertheless, the motion signature of deinterlaced footage is more film-like (less smooth) than that of interlaced video.
Some software packages convert interlaced video to 24 fps progressive scan footage, with interpolation, pull-down and other tricks. You are still technically losing resolution but the results are even more film-like than with straightforward deinterlacing.
That said, true progressive scan cameras, which eliminate the need to deinterlace the footage, are becoming increasingly common and soon 24p will be the new standard for most if not all video projects, as there seems to be universal agreement that the glassy, ultra-smooth look of interlaced video looks cheap and penalizes the project, whereas the slight strobing of progressive scan footage is associated with the look of movies and simply looks better, making the high-end look less elusive.
Shooting video with a shallow depth of field
Video cameras work by focusing an image on one or more CCDs (charged-couple devices), which then convert the resulting electrical signals into an image. The problem is that CCDs are smaller than 35mm film frames, and usually smaller than 16mm film frames, too. Therefore, for a given camera-subject distance and framing, the focal length used by video cameras to produce an image is shorter than for a 35mm camera, so the depth of field is greater. In other words, the background behind your main subject is sharper when you shoot on video.
The reason for which this is undesirable is, once again, perception: having the background out of focus makes the action more compelling and simply looks better. Above all, it draws the viewer a lot more into the world you are creating, because if the background is too sharp it competes with the foreground subject, massively weakening the effect of the shot. (Some people claim that shallow depth of field is only associated with film for historical reasons, but this is highly debatable, as the aesthetic and storytelling benefits of a shallow depth of field are obvious.)
To throw the background out of focus as much as possible, the focal length must be as long as possible (so move back and zoom in) and the aperture must be wide. You should also aim to use cameras with 2/3″ CCDs, as they are larger than 1/3″ CCDs and consequently require a longer focal length for a given camera-subject distance and framing, making the depth of field shallower. Cameras with 2/3″ CCDs also produce images of a higher quality.
Remember that depth of field also depends on the distance between the camera and the subject. The closer the subject is to the camera, the shallower the depth of field. This means that the longest focal length is not always the answer to achieving a shallow depth of field, because it forces you to place the camera further away from the subject. Sometimes an intermediate depth of field results in a shallower depth of field, simply because to keep the framing constant, the camera has to be closer to the subject.
Neutral density filters are vital when you are shooting outdoors and conditions are very bright. Neutral density filters are placed in front of the camera lens in order to reduce the amount of light that goes through, allowing the use of a wide aperture, which results is a shallow depth of field. In this way you can shoot with your camcorder lens at f2.8 even at midday in the middle of the desert. Some camcorders have inbuilt neutral density filters that can be engaged with a switch, but ideally you should have a range of neutral density filters, in one- and two-stop increments, to be used in conjunction with a matte box.
Having a shallow depth of field also enables the use of visual techniques such as rack focusing (dramatically shifting focus from one subject to another in the frame); as a filmmaker, you want every possible visual tool at your disposal, but rack focusing tends to look lame unless you don’t have a very shallow depth of field. Basically you want the subjects to be so far apart in terms of focus that the viewer has no choice but to look at whichever subject is in focus, simply because the other one is so blurred that the eyes naturally shift to the sharp subject. Even high-end movies get this wrong occasionally.
35mm lens adapters: Mini 35 by P+S Technik, Movie Tube, Redrock Micro M2, Letus35 and others
In recent years a number of 35mm motion picture lens adapters have appeared on the market. Examples include the Mini 35 adapter by P+S Technik, the Redrock Micro M2 adapter and the Letus35. Props to the maker of the Redrock, who made an excellent product that is cheaper than some other adapters, and without the ground glass trouble suffered by certain models. It is a straightforward fact that these adapters are the only way to achieve the depth of field characteristics of 35mm cinematography with a 1/3″ CCD camcorder. The tips described above will reduce the depth of field, but will not come close to the blurred backgrounds produced by these adapters.
These DV 35mm lens adapters allow the use of 35mm camera lenses with digital video cameras like the Canon XL2. They work by focusing the image on a ground glass and then projecting it on the CCDs. This results in video images that have exactly the same depth of field characteristics as 35mm film. These adapters can work very nicely indeed, but they will of course require a focus puller who is totally on the ball if the camera-subject distance changes even slightly during the take.
Video color correction: timing/grading
The color correction of digital video is crucial to approximating the film look. This issue is related to lighting but is so important that it deserves its own section. The hue and tint of a given shot is crucial to its visual impact and emotional effect on the audience; the color can be tweaked to make it look the way you want in the process known as color timing or grading. These terms actually come from the film process but are now freely used in digital filmmaking.
Whatever you do, do not leave a scene lit with white light, unless the sterile white look is what the scene needs. Every scene and shot requires its own color, although of course there will probably be recurring color themes in your movie. You can shoot it “white” and then grade it in post using the color-correction tools provided with your video editing program, or you can use dedicated software such as Apple Color, which is part of Final Cut Studio. If you shoot a scene “white” in post-production you can grade it to make it look yellower (warmer), bluer (colder), greener or whichever hue you want. Notice how a scene’s mood changes when you change the hue one way or the other!
A cheaper way to color-correct your project is to do lots of extensive tests and then get the look as you shoot. You do this through creatively white-balancing your camera: you place a colored filter over the lens, press the white balance button, and then take the filter off. The camera will now shoot footage that has a hue complementary to that of the filter. If you want a blue look, you white-balance with an orange filter on the lens; if you want a warm look, perform the white-balance with a blue filter; if you want a greenish hue, white-balance with a pink filter, and so on.
You can get a specific shade of blue by using a specific shade of yellow filter – you can determine which filter will yield the exact shade you want by experimenting during pre-production. Some cameras even allow you to save white balance presets, so that all you have to do once you have found the hue you want is save it as a preset and then simply recall it when you shoot, without having to mess around with filters on set (but you have to use light of exactly the same color temperature during filming, otherwise you will get a different hue).
Adding film grain in post-production to mimic the look of film
Many software packages, like Magic Bullet, can convincingly add film grain to your video footage. You can modulate the level of graininess, and this effect can work very well if used correctly. Film grain is visually pleasing if the story you are telling is pure fiction or if you want a high-end look, for the reasons so eloquently set forth by DP Janusz Kaminski in the quote at the beginning of this article.