Fujifilm Details Latest Evolution of 'Super CCD Honeycom'

Oct 2, 2008 21:04 Tomohiro Ootsuki, Nikkei Electronics

Fujifilm Corp exhibited its in-house designed CCD, "Super CCD Honeycom EXR," at photokina. We interviewed the company's engineer on the concept of the CCD.

Under the circumstances where cameras are required to have an effective pixel count of at least 10 Mpixels to be sold, Fujifilm developed the new CCD to further improve image quality.

The Super CCD Honeycom EXR can drastically enhance image quality by recording images using half of its effective pixels, if the user wishes. And its image quality is even higher than "F30," "F31fd" or "S6000fd," whose image qualities were once evaluated as the highest among the latest compact cameras (an example of a Website that gives high praise to the F31fd based on a test chart).

The new CCD is a 1/1.6 inch type. Its effective pixel count is unknown, but its pixel pitch is reportedly not very narrow. Cameras equipped with the CCD are expected to be released in early 2009.

In what follows, we will explain the aim of Fujifilm and the technology that realized the Super CCD Honeycom EXR.

RGB balance rearranged

First, for the purpose of preventing over exposure, etc, Fujifilm used "two-screen synthesis," a method of taking two picture at once and synthesizing them in the subsequent stage to extend the dynamic range. However, the method used this time is different from the method used for the "Super CCD Honeycom SR," a CCD that was embedded in the company's compact cameras in the past.

For the SR, two types of photo diodes that have different superficial measures and depths of depleted layers were used to take two different images at the same time. But those image pickup devices have low mass productivity, drastically increasing the cost of manufacturing.

This time, for the EXR, Fujifilm used photo diodes with almost uniform properties as in the case of the company's normal image pickup devices including the "Super CCD Honeycom HR."

As for the method to take two images at once, an image being taken is divided into two comb-like structures (side A and side B in Fig 2) and read out twice. Depending on the intentions of those who design the shutter speed and image processing, it is possible to enormously change the exposure times of the side A and side B.

As a result, the dynamic range of the EXR will possibly become eight times higher than the normal level, Fujifilm said. The dynamic range of the SR is four times higher than the normal level.

However, when an image is divided in such a way, the RGB balance of traditional color filters is inconvenient. Therefore, the company changed the arrangement as follows.

· Traditional arrangement

¡¡R B R B
¡¡ G G G
¡¡B R B R
¡¡ G G G

· New arrangement

¡¡R G R G
¡¡ R G R
¡¡G B G B
¡¡ G B G B

New pixel binning method applied

Furthermore, a new pixel binning method was employed to leverage the new arrangement. Pixel binning means unifying signal charges output from pixels in an image pickup device. It has an advantage of being able to enhance image quality of pictures taken in dark places.

In the pixel binning method for the traditional arrangement of color filters, signal charges are added up for same-color pixels aligned in every other longitudinal or horizontal line, lowering the resolution of images.

However, in the new arrangement, same-color pixels are diagonally adjacent to each other. Therefore, Fujifilm prevented the decrease in resolution by unifying each pair of those sale-color pixels into one. The recording pixel count becomes half of the effective pixel count.

Fujifilm plans to make it easy for users to utilize those technologies and functions. Though the details are not clear, it might be developing a function to automatically choose to use (1) the two screen-synthesis, (2) pixel binning or (3) neither of them (to record all of the effective pixels in a bright environment) based on the results of automatic image recognition by the camera.

24 Hours | 7 Days