Comments: No Responses
A few notes on things that weren’t fully articulated in the video…
Correction About HDR Video
Actually some video cameras today do support HDR shooting. The most obvious one is the RED Epic, which uses their proprietary HDRx format. The camera records an additional data stream on top of the normal video to kind of bracket your shots. The effective dynamic range becomes 18 stops! Then in post, when you’re processing the RAW video files, you have the latitude to push or pull the overall exposure like you would when developing film.
Also, many pro camera use a logarithmic gamma curve, which essentially spreads out the exposure levels in the bright and dark areas to fit them into the sensitivity range of the camera’s sensors. So, if a scene has lights that go from 0 – 10, but the camera can only properly expose 3 – 7, the camera will spread out those 0 – 3 dark levels into the 3 – 3.5 range so the camera can see it. Likewise with the whites: the 7 – 10 range will be mapped on to the 6.5 – 7 whites so they’re not all just blown out.
And generally speaking, pro cinema cameras like the ARRI ALEXA, RED Epic, and Canon C300 all have modern sensors with dynamic ranges around 14 stops even without additional software tweaks. So digital filmmakers have much more latitude today, and can get away with less light than they used to.
Most cameras have an adjustable diaphragm that lets in variable amounts of light. It can expand and contract like the pupil of your eye. The bigger the hole, the more light comes into the camera. This also impacts the way the scene focuses. There’s a nice explanation of this in the Depth of Field episode.
f-stop is the ratio of the focal length of the lens to the diameter of the aperture. So, the larger the aperture, the smaller the f-stop.
Each lens is rated to a certain f-stop, which is the widest setting that the lens allows. For instance the 24-105mm Canon L-series zoom lens has a maximum aperture of f/4.
It’s more difficult to get all of the elements inside the lens to resolve an image when the aperture is wide open (especially in a zoom lens), so clever engineering and extraordinary precision is required to achieve f-stops of 2.8, 2.0 or wider. And there’s a huge premium for those last few stops. Adding a stop can double the price of a lens.
The f-stop dos not indicate an absolute diaphragm size; since it’s a ratio, it will change from lens to lens. But on a given lens, f-stops are marked at regular intervals that represent doubling or halving of the light. So, f/4 lets in twice as much light as f/5.6.
The f-stop is actually an ideal number based on the geometry of the aperture and the lens. But some light is lost inside the lens due to scattering and absorption. Cinema lenses are tested and labeled with an index that represents the actual amount of light coming through. Those numbers are called T-stops, and they give a more accurate estimation when trying to set exposure.
A quick note about light meters since I use one in the video. Digital still and video cameras all have some kind of light meter built into their imaging systems, but sometimes it’s nice to have a separate meter. It allows you to test the light in different ways and figure out lighting independently of the camera.
Light meters can measure in three basic ways. Many meters combine these in some fashion, so I’ll talk about them as functions.
The most intuitive is a reflected-light meter. It measures how much light is bouncing off of whatever it’s pointed at. This is how the sensor in your camera works. To use a reflected-light meter, you hold it near the camera position and read the overall scene.
A variation of this is the spot meter. The spot meter has a little eyepiece you look through, and lets you target a single spot in your scene to measure its reflected light level. Very handy if you want to determine the ratio of dark and light parts of a scene, or the lighting ratio of two sides of someone’s face.
The third way to meter is an incident meter. It measures how much light is falling on a particular area. To use your meter this way, you place it in front of the subject you want to measure and point to towards the camera. It tells you how much light is actually falling on that area. That’s how I use it in the video above.
Why does this make a difference? Well, let’s say you’re lighting a lump of coal and a snowball in the same scene. If you use a reflected-light or spot meter to adjust your lights, you’ll end up adding light to the lump of coal and reducing light on the snowball to get the exposure even. But then they’ll both look grey.
You actually want the same amount of light to fall on each of them (match the incident light meter readings), then the snowball will be white and the coal black.
Thanks for tuning in to The Light Journal.