Sony DSC-W80/W Sony® Guide to Digital Photography (Spring 2007) - Page 14

Shooting the digital way CAMERA SYSTEMS The image processor ORIGINAL IMAGE MINUS CCD FIXED PATTERN EQUALS DIGITAL CLARITY The original recorded image (left) shows noise. The secondary capture (center) reveals the fixed-pattern noise of the CCD. Slow Shutter NR creates the resulting image (right) by subtracting the noise from the original picture! (Sample photos for illustration purposes.) Slow Shutter NR NightShot® system Outstanding clarity in low-light shooting. Image noise - unwanted flecks and specks of color - is always a concern in digital photography, especially in low-light shooting. In the dark, the strength of the desired image is relatively weak compared to the "fixed pattern noise" of the image sensor. The Sony Super HAD™ image sensor design works to overcome this problem with microlenses that maximize the incoming light. A second Sony technology, Slow Shutter NR, goes further still. After taking your pictures, the feature takes a second "exposure" of just the fixed pattern noise. By subtracting the noise from the desired photo, Slow Shutter NR achieves uncommonly clear shots at night! Captures shots where others cameras can't - in low light or no light at all! Most cameras are limited to visible light, but Sony's DSC-H9 has no such limitation. The NightShot system illuminates your subjects with infrared light, which the camera uses to record an incredible image. With the NightShot system, you can literally see in the dark. Anti-dust technology Extra protection for the image sensor of the a100. Digital SLRs have a dirty little secret: dirt and dust are attracted to the image sensor surface, degrading the picture. Once there, dust is excruciatingly difficult to clean. That's why the Sony a100 camera has comprehensive anti-dust technology. A special Indium Tin Oxide CCD coating helps repel dust. What little dust gets past this first line of defense tends to be dislodged by a unique CCD "shake" routine. The result? Minimum dust, maximum clarity! Slow Shutter NR is a key to great nighttime photography. (Sample photos for illustration purposes.) The image sensor converts incoming light into electrical voltage. But before it becomes useful, this voltage must first be converted to digital. Even then, the digital information goes through several processes before it becomes a digital picture. Most of these processes are undertaken by the "brains" of a digital camera, the image processing integrated circuit. Analog-to-digital conversion At first, the amount of light at each image sensor pixel is represented by a continuously variable, analog voltage. The voltage for each pixel is converted into a digital sample, made up of computer 1s and 0s. For each sample, more binary digits (bits) result in greater precision. Four bits per pixel result in samples such as 0101. This yields a maximum of 16 possible shades of color. Eight bits (samples such as 01010101) result 256 possible shades. Fourteen bits (samples such as 01010101010101) result in 16,384 possible shades! Even in total darkness (left), the NightShot system captures an image with infrared light (right). (Sample photo for illustration purposes.) 24 CAMERA SYSTEMS De-mosaicing Even with digital samples, we're still short of a complete picture. Because the image sensor typically uses a mosaic color filter array, each pixel is represented by a sample of a single color only (Red, Green, or Blue). To arrive at a usable picture, we need all colors sampled at all pixels. The image processor does this by calculating the likely value for the missing colors by looking at the adjacent color samples. This process is called de-mosaicing. Before de-mosaicing, the camera has just one color sample per pixel (Red, Green or Blue). After de-mosaicing, the camera has Red, Green and Blue samples for every pixel. Of these three, one is an original value and two are calculated. At 14 bits for each sample - Red, Green and Blue - each pixel is represented as one of 16,384 x 16,384 x 16,384 or 4.4 trillion possible shades. Conversion to a data file Before it can be recorded, the digitized, de-mosaiced data still needs to be configured as a computer file, typically in the Joint Photographic Experts Group (JPEG) format. This formatting is another function of the image processor. Analyzing the picture content The image processor is also required to evaluate the picture data to optimize various camera settings. For example, the processor evaluates the brightness of the image in order to set the proper exposure. It evaluates contrast in order to set the proper focus. And today's most advanced image processors can even identify subject matter in the image, such as faces, for high-precision adjustment of focus, exposure, flash and white balance. Setting ISO and white balance In film photography, light sensitivity and color characteristics are largely determined by your choice of film stock. In digital photography, these parameters are largely determined by computations performed by the image processor. Displaying the picture One of the joys of digital cameras is the ability to see the picture in the LCD monitor. The image processor supplies the right information both when you frame the shot and when you review pictures you've taken. The image processor can even convert the digital image into a video signal, suitable for display on a standard definition or even high definition television! The challenge of speed Because the image processor is responsible for so many functions, it plays a major role in the overall responsiveness of a digital camera. In fact, if you shoot sports, pets, kids or other fast-moving subjects, the image processor can actually make the difference between getting or missing the shot you want. The image processor is partly responsible for three of the most important speed requirements. • Start-up Time: The time it takes from pressing the power switch until the LCD display shows the scene. • Shutter Lag: The time it takes from when you fully press the shutter until the CCD captures the image. • Shot-to-Shot Time: The time it takes from the beginning of one shot to the beginning of the next. The challenge of noise In digital images, "noise" refers to mottling, flecks and specks of color that are unrelated to the original scene. While the image sensor contributes most of the noise, the image processor is another factor. In addition, smart processing can even reduce image sensor noise. Noise is critically important because it puts limits on your photography. If it weren't for noise, you could simply switch to a higher ISO setting to shoot in low light without a flash or to freeze fast action. Minimizing picture noise means maximizing your picturetaking opportunities. The challenge of blur Even when the camera is focused perfectly, the picture can still be a blur. The culprit? Camera shake. Any professional photographer will tell you that the ultimate solution to Continued on page 26. The image processor calculates "in-between" values to assign a sample for all three colors at every pixel. 25