A lot of you who read this site make movies. Some are big and some are small. I make small movies, and I’ve found in my short time as a
So you want to make your movie more “cinematic.” In the first CHUDindie, I mentioned that there are a few variables you can manipulate within your image to achieve a more-filmic quality. This week I’m going to go a bit more in-depth on the topic of digital cinematography.
(Photo at right is of Lars Von Trier and a Sony F900 rig from Dogville. I’m loving the shit out of this.)
First let’s take a general look at a few of the things that make film look like film (assuming image resolution is a foregone conclusion):
24 frames per second
A shallow depth of field
Wide dynamic range
And then there’s video (non-progressive):
30 (NTSC) frames per second (60 interlaced fields)
Deep focus
Narrow dynamic range
FRAMERATE
Film is shot at 24 frames per second, an effect that’s created by a spinning shutter inside the film camera. The shutter is essentially a circle with 180-degree opening that lets light in at it spins, meaning that in order to create 24 separate frames in one second the shutter is open for 1/48 of a second. Video is shot at either 30 frames per second (60 interlaced fields) for an effective shutter of 1/60.
The progressive scan format – the “p” in 24p – allows a the lines of a given frame of video to be drawn in a single sequence, as opposed to the combining of two interlaced fields to create a single image in a standard video signal. Shooting video at 24p feels more like film because you’re looking at 24 complete images per second.*
With both film and video the filmmaker can use the shutter to create a more visceral visual experience. Closing the shutter to 1/120th of a second (at 30fps) or 1/96 of a second (at 24fps) will decrease motion blur and create a staccato effect seen in films like Gladiator and Saving Private Ryan and TV shows like The Shield.
DEPTH OF FIELD
Maintaining a controllable depth of field is key when you’re drawing attention to specific areas or subjects within a frame, and your depth of field is greatly influenced by the kind of glass you’re putting in front of the camera. One of the things that’s characteristic of 35mm motion picture lenses is a narrower depth of field, primarily due to the size of the lens and the image plane.** On your smaller, 1/3-inch or 2/3-inch chip/sensor video camera you’ll have a much wider depth of field, which make the 35mm lens adapters out there right now – Red Rock Micro, PS-Technik, or Brevis35 – popular choices for indie filmmakers looking to achieve that 35mm DOF.
The Redrock Micro M2
In most cases the adapters are configured in front of the stock lens and allow the filmmaker to take advantage of both 35mm cine lenses (PL mount) or 35mm still camera lenses (Nikon, Cannon, Minolta, etc). The result is a shallower depth of field and a more “cinematic” image.
While effective, the adapters do have some drawbacks:
– Longer set up times.
– Light loss (requires more lighting).
– Adds more weight to the camera.
– Extra costs not directly related to the adapter (i.e. camera support to cope with the added weight and extended camera body).
Other ways to knock down your depth of field:
– Your f-stop greatly effects DOF. Shooting wide open (at f2.8 or so) decreases your DOF. Conversely, shooting at an f11 or f16 will increase it. If you want to open up your iris but you have too much light then close your shutter or use ND filters to cut it down.
There are drawbacks to this method. Lenses have a “sweet spot” – an f-stop where lens performance is at its peak – that usually falls somewhere between f4 and f8; nowhere near wide open.
– Shoot at the long end of the lens. Literally, get way back and zoom in on your subject. You should have roughly the same DOF at a longer focal length as you would at a shorter one (provided the f-stop is the same), but because the image is compressed, there is the appearance of a shallower DOF.**
– At longer focal lengths, stack your shots – compose with fore, mid, and background elements to draw more attention to the subject in focus.
Luma (brightness) and chroma (color) are more difficult factors to manipulate because to do so requires knowledge of how luminance and color work in the digital realm. Also, you can look at every camera as a different kind of film stock with varying levels of light sensitivity and color response. What works for the EX3 may not work for the HVX-200, etc. Ultimately, how you control your image can be one of the most important elements of digital cinematography.
Dynamic range (in film) refers to the sensitivity of a given medium’s ability to define discernable image detail in the darkest shadows to the hottest highlights. Film has very wide latitude – around 14 stops. Video has more like 7 or 8 stops. There’s just nothing you can do to change how much light your camera can see, but there is something that you can do to affect how your camera sees light.
The way that film and video reproduce luminance and chrominance are very different. On an image density curve, film rolls off the exposure on the toe and the shoulder of the exposure, leading shadows gradually into the blacks, and highlights into white (this is called log space). The result is a low-contrast image. Video has a linear curve, starting at black and terminating abruptly at the maximum definable exposure, resulting in a high-contrast image.*** There’s almost no finessing of the high and low luminance values. The midtone values – called gamma – can be manipulated to draw out a less contrasty image. Keep in mind though, that video is linear by nature, and attempting to coax a low-contrast image out of a high-contrast curve can create muddiness and noise.
Log (film)
Linear (video)
Here are some tips to keep in mind when shooting video:
– Don’t crush your blacks. Unless you’re using a properly calibrated monitor, your flip-out monitor or eyepiece will lie to you.
– Expose on the nose. That is, try to capture your image at the proper exposure. If you have to, underexpose rather than overexpose. It’s easier to pull detail up than it is to bring it back down.
– Shoot with gain set at -3db. This improves your signal quality and helps reduce noise levels in your image. You’ll lose about 1/2 stop of light sensitivity, but if you’re underexposing and need to draw out more detail in post, this will help to keep your image from breaking apart in the shadows and lower midtones.
Another way to control your exposure and create a more cinematic look with video is to light it appropriately. Lighting for video will be the focus of the next CHUDindie.
(EDIT: I just realize I’ve blown past the subject of color sampling, which is pertinent to the subject of color space. Look for a follow-up on sampling soon.)
* Effectively. http://www.mandarinpictures.com/stephenzinn/ has an insane amount of detailed information on the subject.
** http://www.bluesky-web.com/
*** http://www.cirquedigital.com/howto/color_tutorial.html – this is one of the best resources for color space information on the web.