Color fields arrangement

The FSC-LCD is desirable for its high optical efficiency, low power consumption, and low cost. However, the color breakup is significantly observed on FSC-LCDs. We have proposed a 4-color field arrangement (4-CFA) method to suppress the artifacts for FSC-LCDs.

3.1.1 Concept

The two dimension time and location diagram was analyzed in order to determine an integration image of consecutive frames on FSC-LCDs. The fields of primary colors are respectively displayed in Fig. 3-1. When the observer views a moving image, the viewpoint follows the shift of the image, as indicates by the arrow A. Consequently, the observer recognizes a CBU image as rainbow color in Fig. 3-2

(a). The outline portion of image is determined by the order arrangement of color fields. Two other frame images with the color fields arrangement (CFA) were obtained. The integration of three consecutive frames is shown in Fig. 3-2 (b). The image on a retina will be compensated as a gray level image because of the viewpoint through the same ratio of each primary color. Consequently, the observer recognizes a non-CBU image with these color fields.

Fig. 3-1 (a) Color fields of RGB, GBR, and BRG in three consecutive frames (b) the relation between time and location of motional image.

(a)

(b)

Fig. 3-2 (a) Mechanisms of the color breakup when eyes follow the motional image, (b) an integrated image in three consecutive frames with the CFA method.

Although the CFA method can avoid the CBU of dynamic images, a flicker phenomenon will be an issue for the CFA method [84]. Because the human eye is more sensitive to green color, the lowest field frequency for green color to perceive invisible flicker is 50 Hz as shown in Tab. 3-1 [85]. However, the green field appeared at the first and third field in straight two frames. The green field frequency of 36 Hz in this worst case would be lower than the invisible condition. In order to suppress the flicker phenomenon, we proposed to insert a fourth color field to speed up the field frequency.

Tab. 3-1 The lowest field frequency that produces invisible flicker for each color

1st frame

3.1.2 4-CFA method with repeating color orders

The color fields with orders of RGBR, GRBG, and BRGB in three consecutive frames were proposed in the 4-CFA method. The color of the fourth field repeated the one of the first field as shown in Fig. 3-3(a). The motional image and simulation result of a moving image of white bar are shown in Figs. 3-3(b) and (c). The margin of perceived image is blurred due to differences in brightness but without color separations. In addition, the flicker phenomenon should be suppressed because of the color field frequency of 80 Hz, higher than the invisible condition of 50 Hz.

Fig. 3-3 (a) Color fields of RGBR, GBRG, and BRGB in three consecutive frames, (b) the relation between time and location of motional white bar with width of 8 pixels and velocity of 6 pixels/frame, and (c) the simulation image on retina with the 4-CFA method when eyes follow the image.

(a)

(b)

(c) (a)

(b)

(c)

3.1.3 Implementation of FSC controller

A timing controller, which is designed to convert the image data to a desired format in response to the timing signal, is a key component in an FSC system. The block diagram of the controller is shown in Fig. 3-4. This controller consists of the data convert, frame buffer, memory switch, and LED lighting unit. A graphic card is used to convert the input video rate to 60 Hz in advance. These 24-bit RGB input signals are stored into monochromatic red, green, blue data separately in the data convert unit. The function of the memory switch unit is to access the consecutive frame data alternately between two banks of SDRAM. As one is on reading operation, the other one is on writing operation. To avoid conflicts of the consecutive frame data during the operation, a frame buffer is as a register to translate the data into LCD drivers. At the same time, an LED timing unit determines the enable signals for LED drivers to switch corresponding LEDs on.

Fig. 3-4 The block diagram of the FSC controller.

Data Convert Memory Switch and Controller Data Convert Memory Switch

and Controller

A 5.6-inch OCB-LCD with QVGA resolution was used as a prototype to evaluate the properties of CBU and flicker phenomenon. The response time of OCB mode LC cell (t_LC) and the data addressing time (t_TFT) were about 1.0 and 1.2 ms in this study.

For a field frequency of 240 Hz, the illumination time of LEDs (tLED) about 1.9 ms was obtained as shown in Fig. 3-5. The backlight intensity of first and fourth fields was modified to meet the criterion of white balance. From the observation, there was no flicker as we predicted.

Fig. 3-5 The timing chart of the 4-CFA FSC-LCD.

3.1.4 Physical evaluation of CBU

The evaluation of CBU was carried out by a camera-tracking experiment. The perception of CBU was evaluated to verify the proposed CBU reduction method. The motion image was used as a test pattern to compare the image quality with the conventional and proposed methods. A stage with a high-speed camera, Phantom V5.1 by Vision Research, Inc.[86], was set up on a track and adjusted the velocity by a computer interface. In order to simulate the shift of the observer’s viewpoint, we synchronized the moving velocity of the bar on the image and the camera on the track, as shown in Fig. 3-6. The exposure time of camera was set to 1/20 second to integrate three consecutive frames while a frame frequency was 60 Hz. The experimental

RGB (n frame) RGB (n+1 frame)

results of perceived images are agreed with the observation, as shown in Fig. 3-7. The CBU is eliminated in the 4-CFA method with a blurring margin as predicted in Fig.

3-3(c). On the contrary, the conventional CBU was obvious for multicolor.

Considering the moving velocities, the CBU widths are proportional to the velocities of objects on a motional image. Similarly, the blurring margin of fast-motional image became wider in the 4-CFA method. Fig. 3-7 shows the comparison results with bar widths of 15 and 30 pixels. The widths of CBU and blurring margin were independent to bar widths in both driving methods. Moreover, cyan and yellow bars were tested in the 4-CFA method as shown in Fig. 3-8. Images with narrower and blurring margins were obtained simultaneously. Comparatively, field rate increasing [43], one of previous CBU reducing methods, could narrow the CBU width but not blur the margin at the same time. On the contrary, the intensity of color separation could be lightened but the CBU width could not be narrowed by multi-primary-color insertion method [44]. From the experimental results, the 4-CFA method is concluded as the practical method to eliminate the dynamic CBU phenomenon.

Track

High speed camera color bar

synchronized velocity

PC

moving velocity Panel

Track

High speed camera color bar

synchronized velocity

PC

moving velocity Panel

Fig. 3-6 The camera-tracking equipment

Width

Fig. 3-7 Photos of a moving white bar, which were taken by a tracing camera, with bar widths of 15 and 30 pixels, velocities of 3 and 6 pixels/frame.

Width

Fig. 3-8 Photos of moving cyan and yellow bars, which were taken by a tracing camera, with bar width of 15 pixels and velocity of 6 pixels/frame.

In addition to the dynamic CBU, the static CBU should be considered for evaluating the image qualities. The static CBU occurs when abrupt perturbation produces the saccadic movement. This movement gives a retina broken color sequence which is easily experienced while rotating the head or eyes. Therefore, the effective way to reduce this artifact is to increase the sequential frequency. For the still image, the field frequency of 4-CFA in three consecutive frames is 240 Hz.

Comparing the static CBU with the conventional and 4-CFA methods, the field frequency of 4-CFA is 1.3 times faster than that of the conventional one, resulting the slighter static CBU width. The results of our evaluation supported that the proposed 4-CFA method can reduce the visibility of CBU artifacts.