Chapter 4.......................................................................................... 40
4.4 Discussion
3D displays will drastically enhance the viewing experience of future displays for many applications. Unfortunately, 3D displays generally have a lower resolution since the pixels are divided into multi view. Therefore, some 3D displays have the opportunity to switch between 2D and 3D mode such that either natural 3D images or high-resolution 2D images can be displayed.
Our task is to design a 2D/3D switchable display using liquid crystal lens with low crosstalk. Tab. 4-1 is the comparison of conventional LC lens and our novel design. The conventional liquid crystal lens suffers the issues of tardy changing ofn. Hence can cause high crosstalk phenomenon. Therefore, the application of Multi- Electrically Driven Cylindrical Liquid Crystal Lens (MeDLC) became more serviceable.
49
Tab. 4-1 Comparison of proposed and conventional methods.
O: High Δ: Low
4.5 Summary
In this section, a multi-electrode LC with closer lens- like distribution was investigated. The LC cell shows individual characteristics; that is, the refractive index distribution in conventional double electrode LC lens shows not easy to control the profile of n , in comparison with that of multi-electrode. Hence a simple structure with no LC alignment issue, smoother n curve and lower crosstalk phenomenon can be realized by MeDLC.
The MeDLC can be easily fabricated by semiconductor process. By using these well-developed fabrication processes, the designed MeDLC can be produced. The detail experimental results will be shown in next chapter.
Design
50
Chapter 5
Measurement Results and Discussion
5.1 Introduction
The objective of the measurement is to investigate the phase difference in LC layer and the beam size. According to the simulation result presented in Chapter 4, and the experimental result in this chapter, our design shows great performance than conventional one. Finally, the comparison of crosstalk of MeDLC and conventional LC lens will be shown in this chapter.
5.2 Measurement results and discussion
The measurement results can be categorized into three parts: refractive index profile reconstruction, beam size and crosstalk phenomenon. The optical performance of our device can be judged with these three parts. First, our MeDLC prototype were shown in Fig. 5-1
Spacer
Bottom glass Top glass
Side view
Top view electrode
LC director
Fig. 5-1 MeDLC prototype
51
5.2.1 Phase profile reconstruction
From the experiment setup of crossed polarizer we can capture a series of bright and dark lines. These were caused by the phase difference in LC layer, and n profile can be reconstructed by calculating the spacing between the lines. Eventually, consistency with the ideal lens-like distribution will be verified, as shown in Fig. 3-7.
Refer to Fig. 5-2, shows the conventional LC lens with double electrode. We can see that the n curve doesn’t change much with further increasing the voltage. And the curve is in a difference to the ideal curve. Therefore we can judge that the beam size of the conventional lens is large and can be expected a high crosstalk phenomenon, which will be further illustrated.
Fig. 5-2 Experimental result of double-electrode unit.
52
The result of seven electrodes and the ratio of WS :WE 1:1 were shown in Fig.
5-3. First, with our design not only the full effective aperture was kept but also the curve was very consisting with the lens- like distribution. Second, is to check the suitability of the electrode number with ratio : 1:1
E
S W
W as shown in Fig. 5-4.
From the result seven electrodes still has the lowest error function (EF). Final, the ratio of : 1:1
E
S W
W was yet found to be the most suitable ratio, as shown in Fig.
5-5. The sequence of optimized electrode value from outer electrode to center electrode is: 30v、8.6v、9.7v、0v. These measurement results were consisting with the simulation result and imply that to achieve a better performance of lens we must apply MeDLC. To be double confirmed of our design, the beam size will be illustrated at next section.
Fig. 5-3 Experimental result of MeDLC unit.
53
Fig. 5-4 Optimized electrode number
Fig. 5-5 Optimized
E
S W
W : ratio
54
5.2.2 Beam Size
The beam size image of the laser can be captured by CCD with polarizer parallel to the electrode direction, as shown in Fig. 3-9. The beam size was judged by the full width at half maximum (FWHM). Referring to Fig. 5-6 (a) and (b) shows the normalized intensity of conventional double electrode LC lens and the MeDLC Lens with both focal length about 3.3cm. A very sharp light peak point to the center was seen in Fig. 5-6 (b) whose beam size at full width at half- maximum is smaller than the conventional double electrode LC lens about 35%. A 2D mode can be seen after turning off the voltage, due to the laser were uninfluenced by the MeDLC as shown in Fig. 5-6 (c). These measurement results indicating a lens- like distribution was achieved and a 2D/3D switching images can be provided. Moreover, a smaller beam size can be obtained by using MeDLC lens which can result in a smaller crosstalk as well.
With this achievement, a comparison of crosstalk phenomenon of MeDLC and conventional double electrode LC lens will be demonstrated.
55
CCD image
0 255
CCD image
0 255
(a)
(b)
V=62V
Vmax=30V
(Outer electrode)
CCD image
0 255
(c)
V=0V
Aperture size : 1.5mm
Beam size : 80um
Beam size : 52um
Fig. 5-6 Normalized intensity of (a) Conventional double electrode LC lens (b) MeDLC when voltage ON (c) MeDLC when voltage OFF.
56
5.2.3 Crosstalk phenomenon
In the context of 3D displays, crosstalk is the amount of left image that leaks into the right view and right image that leaks into the left. When the crosstalk is too high, the brain no longer perceives two different viewpoints and therefore no disparity.
To investigate the crosstalk, the commercial software Light-Tools was used to calculate the crosstalk phenomenon of MeDLC and conventional double electrode LC lens by inputting the d profile. d is the height of the lens and is substituted by pixel 1 to pixel 9 at a time. For the eye distance of 65mm and a 9m viewing distance, the crosstalk phenomenon based on conventional lens and proposed design were shown in Fig. 5-7 and Fig. 5-8. The results demonstrate that 3D display with proposed design has less crosstalk than the conventional one about 43% as shown in Tab. 5-1.
Therefore, the image quality of each viewing zone can be improved due to lower cross talk.
57
View
N N+1 N+2 N+3 N+4 N-1
N-2 N-3 N-4
Fig. 5-7 Crosstalk of Conventional double electrode LC lens.
View
N N+1 N+2 N+3 N+4 N-1
N-2 N-3 N-4
Fig. 5-8 Crosstalk of MeDLC
Tab. 5-1 Crosstalk was simulated by Light-Tools
MeDLC Conventional
Double Electrode
Crosstalk (%) ~24 ~42
58
5.3 Numerical Aperture vs. Operation Voltage
To judge the operation voltage, we take the numerical aperture (NA) into account as shown in Eq.5-3, where n is the index of refraction of the medium in which the lens is working and D is the diameter of the lens. Here our aperture is 1508um. With same numerical aperture a lower operation voltage was obtained by using MeDLC Lens because it can control the liquid crystal profile by each segment electrode. The numerical aperture with same operation voltage of multi-electrode and double electrode lens were compare as well. From the result the multi-electrode is higher than the double electrode and that means it has higher optical power to banded or focus the light of 3D display. And the numerical aperture is about a factor of 1.66 higher than the double electrode as shown in Fig. 5-9. Therefore, a better optical power as a cylindrical lens and lower operation voltage was indicated. The crosstalk phenomenon can be described by numerical aperture as well. A small numerical aperture can also mean a large focal length; in this case the effect of crosstalk can be serious due to a close image of left and right as shown in Fig. 5-10 (a). As for the large numerical aperture which result in a small focal length can separate the image of both eyes widely, therefore can have a lower crosstalk phenomenon.
f
59
Fig. 5-9 Numerical Aperture property
High Crosstalk Low Crosstalk
f Left
eye Right
eye
Left eye Right
eye
f r (mm)
Normalizedintensity
0 1
r (mm)
Normalizedintensity
0 1
(a) (b)
Fig. 5-10 The crosstalk of (a) small numerical aperture (b) large numerical aperture.
60
5.4 Summary
For the lens- like distribution, beam size, crosstalk, numerical aperture and operation voltage of conventional double electrode LC lens and proposed design MeDLC are summarized in Tab. 5-2. The beam size is reduced by 35% and the crosstalk with proposed MeDLC is lower about 43%. Furthermore the operation voltage of MeDLC Lens is lower than the double electrode lens and have a numerical aperture improves by 66%. In addition, the results indicate that the 2D/3D switching display with proposed method not only smaller beam size and lower crosstalk, but lower operation voltage.
Tab. 5-2 Comparison of conventional and proposed MeDLC.
Δ: High O: Low
Decrease Factor
x0.65 1
Crosstalk (%)
~24 ~42
Numerical Aperture
Increase Factor
x1.66 1
Operation voltage
O Δ
61
Chapter 6
Conclusions and Future Work
6.1 Conclusions
In our opinion a viable 3D display should fulfill the following requirements:
1. The display should be auto-stereoscopic. It should allow freedom of movement with multi viewer potential.
2. The development of 3D display should motivated by trying to realize the image appears more natural.
3. The display should be able to display uncompromised 2D content.
Therefore, a novel active device of 2D/3D switching display has been demonstrated in this thesis. The principle of proposed method is to produce a natural 2D and 3D images by controlling the liquid crystal with switching ON and OFF the voltage. The simulation results indicate that a closer lens- like distribution can be achieved by our device. The measurement result also indicates that our device can be closer to the ideal lens- like distribution. And the numerical aperture (NA) of our device shows an improve by a factor of 1.66 which can result in a smaller volume of display and smaller crosstalk. Also the voltage requirement is much lower with same NA. In addition, due to a smaller beam size of proposed design the crosstalk of our device is lower than that of the conventional double electrode LC lens about 43%. In conclusion, the 3D displays with proposed method not only smaller beam size and lower crosstalk, but lower operation voltage.
62
The proposed design can also be applied for combining with a head tracking device. By tracking the position of the head and a LC lens which is able to move horizontally with respect to the viewer position. Therefore, a freedom of movement with wide viewing angle and high resolution 3D images can be perceived.
6.2 Future work
Currently although our operation voltage is lower than that of the conventional double electrode LC lens (current operation voltage of MeDLC lens is about 30 volt).
Nevertheless we think this is not lower enough due to a thick glass substrate (550um) between the electrodes and liquid crystal. In order to further reduce the operation voltage the thickness of the substrate between the electrodes and liquid crystal lens must be reduced as shown in Fig. 6-1. And our task in the future is to study and find a suitable thickness and parameter (ε: dielectric property) of a dielectric layer. And with more electrodes to control the lens-like profile which will also be an optimized design.
Fig. 6-1 Reduce the operation the voltage by reduce the substrate thickness with an optimized parameter and structure design.
63
Also from calculating the intensity distribution of MeDLC and without MeDLC when off state as shown in Fig. 6-2. The intensity distribution with MeDLC shows a diffraction pattern from ITO. This will reduce the image quality in displaying 2D images. To overcome this problem is to coating a material which has the same absorption and refraction rate corresponding to the ITO. In this case the phenomenon of diffraction can be reduce and without losing the quality in 2D mode. Therefore we will further study about this kind of material in the future.
CCD image
0 255
(a)
V=0V
Aperture size : 1.5mm
CCD image
0
V=0V 255
Aperture size : 1.5mm
(b)
Fig. 6-2 without driving voltage (a) MeDLC (b) without ITO pattern
64
Beside the issue of operation voltage and diffraction in off state. In this research even though the crosstalk can reduced to 24% only. However, the effect of crosstalk in multi- view autosterescopic 3D displays has been reported that the tolerance limit of visual comfort is around 10% [34]. Therefore our second task is to red uce the crosstalk to about 10%. At previous chapter we have described the relative between the focal length and crosstalk. Fig. 6-3 shows a simulation of the trend of the crosstalk versus focal length. From the result, in order to achieve a 10% crosstalk our focal length needs to be reduced to about 1.9cm. And this can be resolved by larger the cell gap or change a liquid crystal which has high Δn.
Fig. 6-3 Crosstalk vs. focal length
65
Reference
[1] J. Mansson, “Stereovision: A model of human stereopsis,” Lund Univ. Congnitive Science, Tech. Rep., (1998)
[2] E. Edirisinghe, J.Jiang, “Stereo image, an emerging technology”. SSGRR, L’Aquila,Italy, (2000)
[3] L. Hill, A. Jacobs “Invited paper: 3-D Liguid Crystal Displays and Their Applications” IEEE Vol. 94, No. 3, March (2006)
[4] J. S. Kollin, S. A. Benton, M. L. Jepsen,”Real-Time Display of 3-D Computed Holograms by Scanning the Image of an Acoustic-Optic Modulator,” SPIE Proceedings, Vol. 1212, pp.174 (1990)
[5] “Laser Based 3D Volumetric Display System,” US Patent No. 5,854,613 (1998) [6] K. Toyooka, T. Miyashita, T. Uchida, Proc. SID’01, pp.174 (2001).
[7] T. Sasagawa, A. Yuuki, S. Tahata, O. Murakami, K. Oda, ”Dual Directional Backlight for Stereoscopic LCD”, Proc. SID’03, pp.399 (2003).
[8] Y. M. Chu, K. W. Chien, H. P. D. Shieh, J. M. Chang, A. Hu, Y. C. Shiu, and V.
Yang, J. Soc. Inf. Display, No.13, pp.875 (2005).
[9] K. W. Chien and H. P. D. Shieh, Appl. Opt., No.45, pp.3106 (2006).
[10] L. Lipton et al., Proc. SPIE, No.4660, pp.229 (2002).
[11] G. Woodgate, J.Harrold, A. Jacobs, etc., “Flat panel autostereoscopic displays-characterization and enhancement”, Proc. SPIE, Stereoscopic Displays and Virtual Reality Systems VII, vol. 3957, pp.153-164, Apr. (2000).
66
[12] D. Takemori, K. Kanatami, S. Kishimoto, S. Yoshi, and H. Kanayama, Proc.
SID’95, pp.55 (1995).
[13] H. Morishima, H. Nose, etc., “Rear cross lenticular 3D display without eyeglasses”, Proc. SPIE, Stereoscopic Displays and Virtual Reality Systems VII, vol.3295, pp.193-202, Apr. (1998).
[14] Cees van Berkel and John A Clarke. “Characteisation and optimization of 3D-LCD module design”, Proc. SPIE, Vol.3012, pp.179-187 (1997)
[15] J. Young, B. Javidi, “Three-dimensional image methods based on multiview images”, J. Display. Tech., Vol. 1, pp.125-140, (2005).
[16] H. Isono et al. Japanese Pat. Appln.JP03-119889, (1991)
[17] M.G.H Hiddink, S.T. de Zwart, etc. “20.1: Locally switchable 3D displays”, SID pp.1142-1145 (2006)
[18] W.L. IJzerman, S.T. de Zwart, T. Dekker, “Design of 2D/3D switchable displays”, SID, pp.98-101, (2005)
[19] Hyung-ki Hong, Sung- min Jung, etc. “25.3: Autostereoscopic 2D/3D switching display using electric- field-driven LC lens (ELC Lens)”, SID pp.348-351 (2008) [20] G.J. Woodgate, J. Harrold, “A new architecture for high resolution
autostereoscopic 2D/3D displays using free-standing liquid crystal microlenses”, SID, pp.378-381, (2005)
[21] G.J. Woodgate, J. Harrold, “Key design issues for autostereoscopic 2D/3D displays”, Euro. Display, pp.1-4, 10 September (2005)
[22] S.J. Battersby, “Auto-stereoscopic Display Apparatus”, US6069650, (2000).
67
[23] G.J. Woodgate, J. Harrold, “LP.1:High Efficiency Reconfigurable 2D/3D Autostereoscopic Display”, SID Digest (2003)
[24] Stephen T. Kowel, Dennis S. Cleverly and Philipp G. Kornreich, “Focusing by electrical modulation of refraction in a liquid crystal cell.”, Applied Optics, Vol.23, No.2, pp.278-289 (1984)
[25] P. Yeh and C. Gu, Optics of Liquid Crystal Displays, Chapter 4, J. Wiley and Sons, New York (1999)
[26] GRINTECH, http://www.grintech.de
[27] E. Hecht, Optics (Forth Edition), Addison Wesley (2002).
[28] H. REN, Y.H. FAN, et al. “Tunable-Focus Cylindrical Liquid Crystal Lens”, Jpn. J.
Appl. Phys. Vol. 43 pp.652 (2004)
[29] E.J.A. Brouwer, “Frequency analysis on a 3D display”, Koninklijke Philips Electronics N.V., Technical note PR-TN 2005/00947, (2004)
[30] Y.H. LIN, at el. “Tunable-Focus Cylindrical Liquid Crystal Lenses”, Jpn. J. Appl.
Phys. Vol. 44, No. 1, pp.243-244,(2005)
[31] S. SATO, “Applications of Liquid Cristal to Variable-Focusing Lenses”, Opt. Re.
Vol.6, No. 6, pp.471-475 (1999)
[32] Y.P. Huang, C.W. Chen, T.C. Shen, “Invited paper: High resolution
autostereoscopic 3D display with scanning multi-electrode driving liquid crystal (MeDLC) lens”, SID, pp.336-339, (2009)
[33] T.C. Shen, L.Y. Liao, C.W. Chen, Y.P. Huang, H.P. Shieh, “Autostereoscopic 2D-3D display with multi electrically driven cylindrical liquid crystal lens”, OPT, (2009)
68
[34] T. Jarvenpaa, M. Salmimaa, “Optical characterization of autoste reoscopic”, J.
SID, pp.825-833, (2008)