Matrix Metering, the metering system behind Nikon's acclaimed auto exposure control, was first introduced on the Nikon FA, in 1983. Since then, it has continued to evolve, and was incorporated as 3D Color Matrix Metering on the Nikon F5, in 1996, and as 3D Color Matrix Metering II on the Nikon D2X, in 2004. Today, its precision approaches the very highest levels of excellence. Nikon's proprietary 1,005-pixel RGB sensor, used in all Nikon flagship SLRs since the F5, continues to deliver scene analysis that no other camera can match. And through constant research, Nikon has been able to achieve a new level of precision auto control that makes the most of the sensor’s full potential. Nikon has devised a new system to read color information with much greater accuracy, and by incorporating this color information into the algorithm that evaluates the subject, the 1,005-pixel RGB sensor’s applications reach beyond auto exposure control to raise the accuracy of autofocus and Auto White Balance. We call this new technology, incorporated in both the Nikon D3 and D300, the Scene Recognition System.
This revolutionary technology was made possible by the sophisticated Nikon diffraction grating in the optical metering system. The incident light reaching the RGB color separation filter is first color separated by the diffraction grating as if it had passed through a prism. The color-separated light then efficiently passes through the respective color filters, enabling precise color information reading. This highly accurate color information is then utilized to the full with the comprehensive assistance of several elements, including the improved precision of the optical metering system, and the high-speed data processing handling information volume three times greater than conventional systems. Scene Recognition System is then able to maximize its ability to improve the accuracy of autofocusing, auto exposure, and Auto White Balance. Using ultra-fine fabrication technology of an order of 1/100 micron precision from the Nikon Core Technology Center, the diffraction grating boasts a fine structure of less than 1 micron — the same level as the wavelength of light.
The Scene Recognition System uses the 1,005-pixel RGB sensor’s color information to detect the subject in AF. The advantage of this system is most evident with the Dynamic-area AF 51 points (3D-Tracking) that automatically shift the focus points in accordance with the subject movement in the frame. Both the D3 and D300 employ Nikon’s new AF technology that uses 51 focus points. The more focus points, the more effectively the camera catches the subject. A higher number of focus points, however, requires greater proficiency on the part of the photographer. Indeed, operating the multi-selector while tracking the subject within the frame is not particularly easy. When the focus mode is set to Continuous-servo AF during 3D-Tracking, the subject’s color and brightness information are read with a user-selected focus point. Then, using the Scene Recognition System’s “Subject Tracking” information to track the subject's lateralmovement within the frame, the camera maintains a sharp focus by automatically shifting the focus points consistent with the subject’s movements. The photographer is thus not concerned with multi selector operation, and can concentrate solely on composition and decisive moments.
In Auto-area AF mode, using the Scene Recognition System’s “Subject Identification” information, which distinguishes human subjects from the background, the camera focuses on the person. Subject Identification uses color and brightness information to determine the background and foreground, and, for the first time in a digital SLR, can distinguish, by identifying skin tone, whether it is a human subject. Based on distance-to-skin information obtained by G- and D-type AF NIKKOR lenses, the camera is able to calculate proportions of and determine the presence and position of a person. This ability of the camera to automatically determine the focus point when there is no time to select a focus point for a spur of the moment shot, or when you’re shooting with Live View’s Hand-held mode without using the viewfinder and selecting a focus point is difficult, is extremely convenient. Subject Identification even makes it easier to focus on a human subject when the subject is not in the foreground.
3D Color Matrix Metering was first incorporated in the Nikon F5. After a process of steady evolution, 3D Color Matrix Metering II was first used on the D2X. The above-mentioned 1,005-pixel RGB sensor has played an important part in enhancing performance. Another improvement in performance is due to the database consisting of approximately 30,000 actual shooting situations.
Photographers have to shoot in various lighting conditions — front-lit, backlit, half backlit — and even in situations where lighting condition constantly changes. It demands them to have a vast amount of knowledge and experience in order to produce an accurate exposure under varying conditions. In order to automate the process and properly support photographers on location where every second counts, ever since the launch of Matrix Metering on the Nikon FA in 1983, the Nikon engineers developing the metering system have been consistently leading the test shooting in various kinds of lighting situations, and constantly upgrading the metering algorithm via incorporating the database of actual shooting situations.
Given the infinite variety of compositions and shooting conditions, it is impossible to record all the possible lighting situations. With their vast body of experience, however, Nikon’s seasoned engineers newly selected and judged approximately 30,000 of the most indispensable actual shooting situations. We then upgraded the database in the algorithm for improved application of the 1,005-pixel RGB sensor using the diffraction grating employed by both the D3 and D300, thereby ensuring further well-balanced exposure results.
With the D3 and D300, in addition to improving exposure control accuracy, Nikon has also taken on a new challenge - the Active D-Lighting that combines exposure control with image processing. With conventional systems, when the contrast in a situation was too high, either the highlights would be washed out or the shadow details would be lost; the photographer had to sacrifice one or the other. Now, Active D-Lighting solves this problem, working in combination with 3D Color Matrix Metering II to determine the optimal exposure. And in both the D3 and D300, the Scene Recognition System’s “Highlight Analysis” information delivers more accurate exposure control. Subsequent to shooting, image processing specific to Active D-Lighting is carried out to provide images just as your eyes see the scenes. Our database of actual shooting situations makes a major contribution here as well.
The Scene Recognition System has significantly improved the precision of Auto White Balance. Although Nikon's Auto White Balance systems have long been praised by leading photographers, with complex or particular lighting sources, in some cases the judgment of the system wavered. For a conventional AWB system, for example, under a mercury vapor light white tinged with green and the green of vegetation would appear exactly the same. How could Nikon deal with this? When the subject is green vegetation, the light source is most often natural daylight. Nikon felt that if the light source could be identified, the camera could accurately make decisions whether or not to remove green (with mercury vapor lights), or let the green remain (with natural daylight). Using a simulator, we analyzed approximately 20,000 images of actual situations and constructed algorithms that would dramatically improve AWB accuracy by determining the light source. Scene Recognition System treats the color and brightness of the scene read by the 1,005-pixel RGB sensor as characteristics extraction. The camera determines the light source based on the characteristics extraction and color distribution in the same manner as 3D Color Matrix Metering II. As a result, if you are shooting under tungsten light, for example, the system maintains the warmth of the color and reproduces the image as human eyes would see the scene. Although it may seem easy, we have succeeded in achieving a task of extremely difficult color reproduction. Even when shooting indoor sports and stage performances under lighting conditions that conventional AWB systems would have found difficult to cope with, Nikon’s new AWB correctly determines the light source and delivers high-precision Auto White Balance.
