Chapter Five: The pursuit of high image quality.
1The APS-C size sensor – The best choice!
- The sensor for the X100 had to deliver beautiful images with high sensitivity while satisfying the desired camera dimensions. We undertook simulations in our labs and tested various sizes of sensors from 1/2 and 1/1.7 sensors used in conventional compact cameras to the Micro 4/3-size sensor. Based on the results, we determined that the APS-C size sensor with its superior S/N (high sensitivity/low noise) performance was ideal. (See Figure 1 and Figure 2.)
- Of course, a full-size sensor delivers even higher sensitivity and better S/N performance, but it requires more power for operation. In addition, measures to deal with the issue of heat generated by a full-size sensor were quite complex, and the large size would have a direct impact on the exterior dimensions of the camera. After giving careful consideration to finding a good balance between high image quality versus power consumption and factors like heat and how the bigger sensor affects the camera dimensions, we decided that an APS-C size sensor was the best choice for the X100.
- Our priority was not the number of pixels, but rather the size of the photodiodes. The larger the photodiode, the higher the sensitivity (the higher the number of electrons that can be held by the photodiode). The larger aperture of each pixel enables the sensor to collect light more efficiently.
- The reason for not using the EXR pixel array (array rotated 45°) in the X100 sensor is that the APS-C sensor without modification already provides more than ample signal information, making the EXR image capture approach unnecessary. The unique pixel compositing and blending technologies of our proprietary EXR technology are designed to get optimum image quality from sensors with small pixels.
2Best match and optimisation for the 23mm lens
- In order to tailor the APS-C sensor to the X100 lens and enable it to collect sufficient light in the sensor edge areas, the microlens array has been optimally scaled and customised. (See Figure 3.) The precision positioning of the microlens array was determined by simulations using wave optics and the calculation of colour wavelength characteristics and the precise path of light based on the pixel structure. The result is virtually the same light collection performance in both the centre and edge areas of the sensor. (See Figure 4.)
3New EXR processor – Designed to deliver higher speed and improved movie quality
- The previous generation processor was specially developed for the EXR-CCD sensor, but from the earliest stage of development, the new processor in the X100 was designed with the CMOS sensor in mind. The most important feature is the vastly improved processing speed. Compared with a conventional processor, it has been designed to deliver an overall improvement of 30% and twice the speed for some execution tasks.
- Camera responsiveness has also been improved. Re-assessment and re-engineering of image processing and the circuit configuration enables confirmation of the captured image in about 0.2 sec. after shutter release and shrinking the interval until the next shot to 0.9 sec. (compared with approx. 1.4 sec. with a conventional model.)
- In addition, in order to improve the EVF viewability of X100’s unique Hybrid Viewfinder, efforts have also been devoted to enhancing movie image quality. The improvement in image quality is also exploited in HD movie recording. As a result, the processor contributes to not only exceptional still quality but also excellent movie image quality.
4Slim optical low-pass filter exploits the full potential of the lens
- In order for the sensor to accurately reproduce the image from X100’s high-performance lens without loss of the high resolution performance, it is desirable to use an optical low-pass filter that is as thin as possible with a small separation width. However, when the separation width is reduced, false colours can be generated by sensor sampling. By optimising signal processing to minimise false colour generation while reducing the separation to the smallest possible width, the design lets the user take full advantage of the X100’s lens.
5The role of signal processing – Bringing out the essential beauty of the data
- The new EXR processor integrates two CPUs, twice as many as a conventional processor, to fully exploit the high performance of the sensor. This enables simultaneous processing of signals with different properties and contributes to high-speed processing performance. As a result, the EXR processor in the X100 contributes to improved AF speed as fast as 0.16 sec. with contrast detection-type AF.
- Managing the heart of the image processing system, the mission of the EXR CORE is to ensure that information received from the sensor is fully exploited without loss of any data; maximum resolution is preserved; detail and smooth tonality in highlights and shadow are revealed; and colour is naturally and faithfully reproduced.
- In addition, the EXR processor incorporates a reconfigurable processor. This processor that can dynamically adapt with rewritable circuits gives the EXR processor ample capacity to perform complex correction and processing tasks.