cgardner
Offline
Dedicated FM
Upload & Sell: Off
|
 p.1 #5 · Thoughts on Live View? | |
My camera previous to my 20D was a Minolta D7Hi with an EVF which read off the sensor in real time. That gives a camera the ability to detect over and under exposure down to the pixel level in real time. For that reason many P&S cameras with LCD viewfinders do a better job at automatic exposure than a DSLR.
At some point Canon removable lens cameras will have EVF and I suspect Live View is a form of EVF "training wheels" for SLR user who can't imagine life without an optical viewfinder. But lose the pentaprism and flopping mirror and say good-bye to the x-sync limit, the 8 fps shooting limit, and most exposure woes resulting from doing the metering off the viewfinder before the actual exposure.
There are two definitions of "correct", "proper", or "optimal" exposure: technical and perceptual
The technical one is when the camera is able to reproduce the full range of scene with detail. That is possible with a wide range of scenes with B&W because the negative range can be changed to to compensate for scene range (i.e. zone system) or different paper contrast grades selected to match changing negative ranges (multi-contrast paper). Color emulsion contrast can't be altered in that way to fit the scene, either can digital.
What photographers started to do after the introduction of color film is to expose PERCEPTUALLY based on the midtones such as skin. If the scene range exceeded the contrast range of the print paper (the limiting factor) the photographer would aim for the middle and get it right and allow the ends to clip. But over time the new "correct" exposure came to be whenever what was most important was correctly exposed. That works when the areas clipped are insignificant, so photographers also learned to compose shots in ways to make the clipping less noticeable.
Thus its possible to have a photo which is correctly exposed in the perceptual sense, but not in the technical sense of having detail everywhere as in B&W.
Indoors in the studio the difference wasn't noticed much because lighting was used to fit the scene to the film. Even with digital if we start in a totally dark room and first add even fill to the point dark detail is revealed...
Then add background light for separation...
frontal "key" light to define shape....
and accent light from behind to complete the illusion of 3D shape
What we've done is change the range of the scene to exactly match the range of the sensor, achieving correct exposure technically and perceptually at the same time.
Outdoors the sky conditions and angle of the light affect contrast, but so does content...
Both shots above are exposed per the same criteria, retaining detail in the white highlights. But the relative amount of highlights and mid-tones in the shots influence perception of exposure making the top photo seem more "correct" in the perceptual sense. What flash allows us to do in that situation is balance the shady side with the sunny side...
at least in the foreground...
Again the context of the content will affect perception. Against the dark background, created as a result of keeping the sunny highlight below clipping, the foreground which is optimally exposed technically (detail everywhere on the scale) actually looks a bit overexposed perceptually! But take the foreground and crop it tight and it will look and be perfectly exposed in both the technical and perceptual sense.
Confusing? Well, understanding the difference between the way we see and the way a camera records images is the craft part of photography we who started in B&W and moved to color figured out along the way. Someone today starting with digital may not recognize how scene range and sensor range relate to each other or the need to change the scene with flash to fit the sensor both indoors and out.
Its not just the end of the scale in play either. As dynamic range of film got shorter, from 10 stops in B&W to about 5 with color the 18% gray, the perceptual mid-tone, stayed in the middle. What changed is how many stops under exposure to took to render white as middle-gray in a photo: 4 in B&W but only 2.5 in color. Digital cameras have a total range of about 8 stops, but only 6 which carry detail. As a result when exposure is set to be correct in the highlights as in this shot...
all the other darker tones look so much darker than by eye; middle gray is only about 3 stops away from the point of clipping. Getting a scene to match the sensor in sunlight outdoors requires an act of faith akin to bungie jumping. First we need to shoot into the shadows of the ambient because flash can't reduce contrast if it overlaps the shadows and highlights at the same time. Next we need to expose the brightest ambient highlights below clipping. The reason should be obvious; if they are blown without flash adding flash will just blow them more. Finally comes the act of faith, relying on the flash to bounce the shaded side back up to where it looks normal perceptually. The trick with flash is blending its direction with the direction of the natural light, which comes from overhead most of the time.
Its not possible to match the entire range of the scene to the sensor, but that's where the role of perception is used to fool the brain of the viewer. Cropping out the underexposed background makes it disappear -- get tuned out as unimportant - perceptually. Its only when there is more negative space than foreground content that the darker background begins to be noticed.
Chuck
|
