Film vs. Digital Video
There are three
visual attributes that make DV look different than film: exposure
latitude, motion blur, and resolution. If you can
control these variables, they will be less obvious to the eye and your
DV will come closer to the elusive "film look":
Exposure Latitude
- A key difference between DV and film is exposure latitude, which
affects contrast and detail. Color negative has a usable exposure
range of 7 stops, with normal exposure approximately in the middle.
Most stocks provide 4 stops overexposure and 3 stops underexposure
where detail is still visible.
Video has a usable exposure
latitude of 5 stops, providing 2 stops overexposure and 3 stops
underexposure where detail is still visible. Exposure beyond the -/+
limits results in tonal compression and is reproduced as either pure
white or pure black, respectively. Obviously, there is a loss of
detail as well.
Since highlights tend to be
more troublesome than shadow areas, exposure
is generally geared to highlights, letting
shadows fall where they may. This avoids the "burn-out" often
associated with poorly shot video. Still, it's important to stay
within video's usable exposure and contrast range.
Motion Blur - Film yields a
slight blur in moving objects. This is known as motion blur
and it results in a distinct fluidity of movement-- a prime
contributor to the "film look." Motion blur is caused by film's relatively low frame
rate of 24 frames per second. A telltale sign of video is its
extreme sharpness and lack of motion blur.
NTSC video runs at 30 fps so how can such
a small
difference account for the radical increase in sharpness? The reason
is that there are two interlaced fields for every frame of video, so
the effective rate is actually 60 images per second (= 30 fps x 2
fields). This virtually eliminates motion blur, creating an
image that is a bit too sharp and devoid of fluidity (the dreaded
"video look").
The answer to this is a technical
breakthrough called progressive scanning, where each frame is scanned
once. In other words, the frame is scanned as a single field, with no
interlacing. The lower image rate reproduces motion blur comparable to film.
These cameras generally use frame rates
of 24fps to 30fps. 24 frames progressive, called 24p,
simplifies combining video and film footage because there is a one-to-one
frame relationship. The PAL version, 25p, matches the
European film speed of 25fps. Many state of the art cameras have switchable
frame rates and resolutions.
Another benefit of progressive scanning
is a dramatic increase in resolution. This occurs because progressive
scanning eliminates interlace artifacts (combed edges in movement) and
interline flicker (noise in fine patterns). There is a study by William E.
Glenn showing that perceived resolution in progressive scanning is 50%
greater than interlace scanning (Understanding Camera Resolution,
Broadcast Engineering, August 1999.
Resolution - The
final difference between video and film is resolution. Many filmmakers
erroneously assume that film is far superior across the board. The
truth is, HD has all but closed the gap. See
HD vs. 35mm
Film for more.
Despite this, film is still far
superior to standard definition (SD) video-- most of the formats in
the above chart. To minimize
this disparity you must make sure that nothing degrades the image
quality when shooting. For example, improper exposure will diminish
the apparent resolution of video by compressing tones and destroying
detail.
Arguably, the disparity in
resolution has less of an impact on the look of DV than exposure
latitude and motion blur. It is not noticeable to the average
audience, except of when aliasing rears its ugly head. Aliasing
can be minimized by avoiding fine patterns, particularly checkered and
striped clothing.
DV has an interesting
advantage over film that may, in part, make up for its lower
resolution. It can "see" in low light almost like the human
eye and captures beautiful images during sunrise and sunset. In fact,
gain can be boosted to +12 dB with minimal consequence. This is true
for Mini DV as well.
Transfer to 35mm Film
The ability to shoot on DV and
release theatrically on 35mm film has taken great strides, with some
interesting developments along the way. The trend started with The
Celebration in 1998 and it is picking up momentum. Nearly 50% of all IFP
submissions in 2000 originated on DV with hopes of theatrical release.
A combination of high quality
DV cameras (Mini DV included) and progressive labs like Swiss
Effects have produced truly stunning prints. Such transfers are made
with a laser recorder. This state-of-the-art machine uses three laser
beams to record RGB video directly to each frame of color negative.
If you anticipate transferring
DV to 35mm, there are certain decisions and settings that you must make
in terms of the camera, namely:
-
decided on 4:3 or 16:9
aspect ratio (discussed below)
-
switch on the dynamic
contrast control
-
switch off the digital
zoom
-
switch off the electronic
picture stabilizer
-
lower the detail level (or
sharpness)
Some labs have preferred
settings for different cameras. You should consult with the lab before
shooting begins and run a test to see that your expectations are met.
4:3 vs. 16:9 Aspect Ratio
A cameras aspect ratio is a
function of the CCDs shape. There are two video aspect ratios in use
today. The standard ratio is 4:3, which provides the square broadcast TV
image. The widescreen ratio is 16:9 and conforms to theatrical film and
HDTV standards. Most professional cameras allow one button switching
between ratios.
Some mid-range cameras have a
16:9 mode but do not have a true 16:9 CCD. Rather, they use a 4:3 CCD
and simulate the wide-screen aspect ratio by masking the top and bottom
of the image. This creates a widescreen effect, but uses less vertical
lines in the process.
4:3 Chip in 16:9 Mode
(shaded area appears black)
A true 16:9 chip
uses all vertical scan lines, so resolution is superior. It has side
panels that are used in the 4:3 mode and ignored in the 16:9 mode.
16:9 Chip
(uses all vertical scan lines, none masked)
You can tell whether the
camera has a true 16:9 CCD by observing the image while switching to the
16:9 mode. If the horizontal view expands, the chip is 16:9. If the
vertical view crops, the chip is 4:3 masked to simulate 16:9. In some
cameras the both the vertical and horizontal changes. If this happens,
the acid test is whether the horizontal view really broadens. If so, the
chip is 16:9.
Whether you use 4:3 or 16:9
mode the resolution inside the visible frame is the same (you simply
mask the top and bottom scan lines in the 16:9 mode). Consequently the
choice is really a matter of aesthetics rather than quality.
When transferring to 35mm it
is desirable to use all vertical scan lines (rather than mask some) to
maximize resolution. That's why a true 16:9 chip is important. In
the absence of a 16:9 camera, some filmmakers use an anamorphic lens to
squeeze all the vertical scan lines into to the 16:9 aspect ratio. You
should consult the lab when attempting this.
If a 16:9 camera or anamorphic
lens are not available, take comfort in the fact that the 4:3 chip
yields surprisingly good results in both 4:3 and 16:9 modes when
transferred to 35mm. Whatever approach you take, it is important to do a
test transfer beforehand. The test should include short clips under
various lighting conditions and ratios. Also, shoot different close-knit
patterns to see how your camera handles aliasing.
