Tuesday, May 31, 2011
Powerful processors and software enable DSLRs to do things never before possible in-camera
Nikon D5100 with the GP-1 GPS unit
A few digital cameras have built-in GPS units (including Sony's SLT-A55 DSLR), which can record the location where each image is shot (latitude, longitude and altitude). This information is recorded in the image's EXIF metadata. You can sort images by location, and plot them on digital maps. You also can get geotagging GPS units that enable this capability (Nikon's GP-1 for some of its DSLRs, for example).
For some time now, many digital cameras can detect human faces in a scene (including a number of DSLRs), and optimize focus and exposure accordingly. This simplifies photographing people or groups. An offshoot of face detection, Sony's Smile Shutter can detect when a subject is smiling and trip the shutter at that moment.
One bane of people photographers is the fact that people blink, often just as the shutter is tripped. Blink-detection technology can detect a blinking subject, and warn you that one or more subjects blinked in the last shot so you can reshoot on the spot.
Most DSLRs offer in-camera noise reduction (NR). It comes in two flavors: long-exposure NR and high-ISO NR.
When you make long exposures, more noise is produced as the exposure times lengthen. Most DSLRs have a long-exposure NR feature that takes a second "dark frame" after you shoot an image, activating the sensor for the same duration, but with the shutter closed. This records the noise pattern. The camera then subtracts the noise pattern in the dark frame from the data in the actual image to reduce the noise quite effectively.
When you increase the ISO setting of a digital camera, image data is amplified, resulting in increased noise of a different sort. High-ISO noise consists of two parts: chrominance noise (color blotches) and luminance noise (grain, like gray spots). This is different than long-exposure noise, and the camera deals with it in a different way. Exact details vary among camera manufacturers, but basically, the weaker settings deal with chrominance noise (which most viewers find offensive), while higher settings also work on luminance noise (and might reduce fine detail in the process). You also can use just the default setting, which works well for most images.
You also can apply NR when you process your images, of course, using your RAW converter or image-editing software, but it's very convenient to be able to do it as you shoot. The main drawbacks to in-camera NR are that it's not as effective as postproduction NR (your computer's processor is more powerful than cameras' built-in processors, allowing for more powerful NR algorithms), and it slows camera performance: Since the long-exposure NR "dark exposure" is equal in duration to the image exposure, it takes the camera twice as long to record each image; and high-ISO NR can reduce the number of consecutive exposures you can make in continuous-drive mode.
Sony's Multi-Frame NR shoots six frames in rapid succession, then combines the data into a single image with greatly reduced noise, providing ISO settings up to 25,600 in cameras that otherwise top out at ISO 6400.
Page 4 of 4