Figure. 4. The polarization states of the proposed FPR system on the Poincaré sphere: the polarization path for right image (a) in horizontal direction and (b) in vertical direction, and that for the left image (c) in horizontal direction and (d) in vertical direction.
Figs. 4(c) and 4(d) show the polarization paths of the proposed FPR system for the left image in the horizontal and the vertical directions, respectively. In the case of the left image, the polarization state has the vertical polarization, which is position -S1. So, the difference of an optical axis between the left patterned half-wave A-film and the half-wave biaxial film has 90° in order to return to the starting position -S1from the position KH and KV. As a result, we obtained the optimized values of the final polarization position -S1: -0.9994 (B), -0.9996 (G), and -0.9997 (R) in the horizontal direction and -1 (B), -1 (G), and -0.9999 (R) in the vertical direction.
We calculate the light leakage and the 3D crosstalk and compare the conventional FPR system in order to verify the optical performance of the proposed optimal FPR system. Figure 5 show the calculated light leakage of the proposed FPR system for the entire visible wavelength range at normal, horizontal, and vertical viewing angles. The conventional FPR system experience serious light leakage as shown Fig. 1(b). However, the proposed FPR system maintains an excellent dark state for the entire visible wavelength spectrum in the normal direction as well as in the horizontal and the vertical directions.
Figure. 4. Calculated light leakages of the proposed FPR system in the normal and the oblique directions (θ = 70°, = 0° and 90°) for the entire visible wavelength spectrum.
The 3D crosstalk value can be defined as 
Here, L(R)_bright is the luminance of the left and the right eyes at the bright states, and R(L)_dark is the luminance of the left and the right eyes at the dark states. We confirmed that the proposed FPR system improved the left 3D crosstalk by about 91.87% in the horizontal direction and by about 97.22% in the vertical direction compared to the conventional FPR system. And the right 3D crosstalk obtained same 3D crosstalk value of the conventional FPR system.
In summary, we proposed optimal configuration for a FPR with a linear polarizer in a stereoscopic 3D display. In order to obtain excellent 3D image quality and vivid colors in all viewing angles, the proposed FPR system applied the wide-band and wide-view half-wave retarder, which consists of half-wave biaxial film and a patterned half-wave A-film. We demonstrated good optical performances of the proposed FPR system by effectively improving the 3D crosstalk in the oblique direction. We are confident that the proposed FPR 3D display will be an outstanding technology that will allow high-quality 3D images and wide viewing angles to be achieved.
This research was supported by LG display.
References S. Pastoor and M. Wopking, Displays 17(2), 100 (1997).
D. Matsunaga, T. Tamaki, H. Akiyama, and K. Ichimura, Adv. Mater. (Deerfield Beach, Florida.) 14(20), 1477 (2002).
J.-C. Liou, K. Lee, and F.-G. Tseng, in Digest of Technical Papers of the 8th International Meeting on Information Display (IMID, Ilsan, Korea, 2008), pp. 710-713.