-
Using a precisely-regulated imaging element and optics, we enable higher-quality imaging that was a challenge with conventional systems
-
We develop all vital systems internally, such as imaging lenses, sensors, and signal processors
-
We also provide total solutions, including post-imaging data analysis
Efforts to improve image quality
Large format image sensor
In digital photography, the size of the image sensor determines the amount of light that can be captured, i.e. the amount of information in the image. The larger the image sensor, the easier it is to obtain smooth tonal expression and high resolution. However, the size and weight of the camera body and the lowered processing speed due to larger image capacity become issues, making it more difficult to design and manufacture.
Fujifilm is one of the few manufacturers able to make cameras with large-format image sensors, which are even larger than the APS-C and full-frame image sensors used in typical interchangeable lens cameras. A lens with the resolution to capture the power of a 100 megapixel image sensor, precise assembly to maximize performance, and image processing technology that allows for easy shooting even when handling tremendous amounts of data. The combination of these brilliant technologies results in a camera that reproduces a breathtaking 100 megapixel world.
![The image sensor size is 43.8 mm x 32.9 mm.](https://asset.fujifilm.com/www/gr/files/2024-05/4381caaeed8f43c75459a4a8c28c91ec/fig_imaging_01.png)
![The APS-C image sensor size is 23.5 mm x 15.6 mm, while the 35 mm full-frame image sensor size is 36.0 mm x 24.0 mm. However, the sensor size of the FUJIFILM GFX100S measures 43.8 mm x 32.9 mm.](https://asset.fujifilm.com/www/gr/files/2024-05/8a96e0aa7f42e7bdc78abb8474906c1f/fig_imaging_02_en.png)
Image stabilization
While larger sensors have the advantage of producing high-resolution images, even slight shake can impact image quality. Therefore, it was necessary to take precautions against camera shake and vibration, such as fixing the camera to a tripod, making it unsuitable for casual photography in some cases. To solve this issue, Fujifilm developed the world’s first sensor-shift image stabilization function for large-format image sensors. We have achieved highly effective eight-stop image stabilization performance (equivalent to a 2 to the 8th power = 256 times faster shutter speed). This capability is supported by composite technologies brought together at a high level. This includes an algorithm that separates camera shake from the effects of the earth’s rotation and a high-precision mechanism that guarantees the positioning accuracy of the image sensor on the micron level.
![](https://asset.fujifilm.com/www/gr/files/2024-05/fb2ea026e01a04e8aedb4153d58f70cb/fig_imaging_03.png)
Pixel Shift Multi-Shot
In digital photography, the subject is captured as RGB color information by the image sensor. At this point, the color information for R (red) and B (blue) is not available in the G (green) pixels, so the missing color information is supplemented via calculations from adjacent R and B pixels. The same processing is done for R and B pixels, meaning 66% of a full-color image is calculated values rather than actual acquired data. Fujifilm has developed technology that uses an in-body image stabilization mechanism to reproduce true-to-life colors by capturing full pixel information for each RGB by shifting the image sensor position one pixel at a time. Furthermore, by shifting the image sensor by 0.5 pixels when photographing, the pixel count is effectively quadrupled. This makes it possible to precisely reproduce large subjects such as paintings and historic ruins, which were difficult to capture in detail in the past, with overwhelming tonal expression and three-dimensionality.
Principle behind the Pixel Shift Multi-Shot (Shifting image sensor)
This is achieved by taking 16 consecutive shots using the following two processes.
In order to obtain accurate RGB information for all of the pixels, four images are taken, shifting the image sensor pixel by pixel.
By repeating (1) four times while shifting by 0.5 pixels each time, the pixels are refined, achieving four times the resolution.
![(1) Obtains accurate RGB information from a total of four captured images: one in its initial state, one shifted one pixel to the right, one shifted one pixel up and one shifted one pixel to the left. (2) By shifting (1) to the right, up and left in increments of 0.5 pixels, the pixels are refined, achieving four times the resolution.](https://asset.fujifilm.com/www/gr/files/2024-05/06abf75f902abe5795266e61f5056f40/fig_imaging_06.png)