A New Approach to Electro-Holography: Can This Move Holography into the Mainstream?

Holography holds great promise for use as 3-D displays, but to date several challenges have prevented the technology from becoming commercialized. But those obstacles are now starting to be overcome. Here, SeeReal Technologies describes its new approach to electro-holography.

by Hagen Stolle and Ralf Häussler

THE THEORY OF HOLOGRAPHY was discovered in 1947 when scientist Dennis Gabor was seeking ways to improve the resolution of an electron microscope. A hologram of an object was generated by interference of the electron wave reflected from the object. The object was reconstructed via subsequent illumination of the hologram with the reference wave. The same principle applies when using visible light instead of electron waves.

Holography quickly captured the imagination of scientists and, over time, society developed a thirst for high-quality holography to transform limited 2-D representations into much more tangible, life-like images. Holographic displays are superior to stereoscopic displays with respect to 3-D depth perception because holographic displays reconstruct an accurate light field with pixels placed at correct optical viewing distance. However, two inherent historical barriers have prevented real-time holographic displays for computer-generated holograms from becoming reality.

•Insufficient Display Resolution. In order to achieve a viewing angle of 60°, which is necessary to serve several users, a pixel pitch of about 1 wavelength is required. This means that for a 47-in. holographic display, for example, a resolution of 250,000 times that of HDTV is required.

• Inadequate Data Volume and Processing Requirements. The computation of each display frame requires significantly more steps for a holographic display compared to a 2-D display. Hologram computation involves calculations of Fourier transformations. This factor, coupled with the greatly increased number of pixels required, places a demand for enormous amounts of computational power. Thus, real-time video-quality holograms would typically require processing power up to several hundred Peta-FLOPS, i.e., approximately 10¹⁷ floating-point operations per second. This is far more than the current computation power of super computers.

The conventional approach to holographic displays reconstructs an object around the Fourier plane of a spatial light modulator (SLM). The size of the object reconstruction is limited to one diffraction order of the SLM because otherwise an overlap of multiple diffraction orders would be visible. The size h of one diffraction order is given by h = λd/p, where λ is the wavelength, d is the distance between the SLM and object, and p is the pixel pitch.

As an example, with λ = 633 nm, d = 500 mm, and p = 10 μm, we get h = 32 mm. A small pixel pitch of 10 μm is needed for such a small object reconstruction with a lateral size of 32 mm. The absolute number of pixels depends on the viewing angle, i.e., the angle from which the object can be seen. Prototype systems use 15 or 100 million pixels^1,2 or a high-frequency acousto-optic modulator.³Estimations result in a pixel count of 10¹² pixels for a full-parallax display with a 0.5-m lateral object size and ±30° viewing angle.²

The fundamental difference between con-ventional holographic displays and our approach is in the primary goal of the holographic reconstruction. In conventional displays, the primary goal is to reconstruct the object. This object can be seen from a viewing region that is larger than the eye separation.

In contrast, the primary goal of our approach is to reconstruct the light wavefront generated by a real existing object for the location in space of each eye.⁴ The reconstructed object can be seen if the observer's eyes are positioned in or close to at least one virtual viewing window (VW). A VW contains the Fourier transform of the hologram and is located in the Fourier plane of the hologram. The size of the VW is limited to one diffraction order of the Fourier transform of the hologram.

Figure 1 illustrates our approach and shows a Fourier transforming lens, a spatial light modulator (SLM), and an eye of an observer. Sufficiently coherent light transmitted by the lens illuminates the SLM. The SLM is encoded with a hologram that reconstructs an object point of a 3-D scene. A scene with only one object point and its associated spherical wavefront is shown. It is evident that more complex scenes with many object points are possible by superposing the individual holograms.

The conventional approach to holographic displays generates the wavefront that is drawn in red. The wavefront information of the object point is encoded on the whole SLM. The modulated light reconstructs the object point that is visible from a region that is much larger than the eye pupil. Because the eye perceives only the wavefront information that is transmitted by the eye pupil, most of the information is wasted.

In contrast, our approach limits the wavefront information to the essential information. The correct wavefront is provided only at the positions where it is actually needed, i.e., at the eye pupils.

Figure 1 shows a virtual VW that is positioned close to an eye pupil. The wavefront information is encoded only in a limited area on the SLM, which we refer to as the sub-hologram (SH). The position and size of the SH is determined geometrically by projecting the VW through the object point onto the SLM. This is indicated by the green lines from the edges of the VW through the object point to the edges of the SH. Only the light emitted in the SH will reach the VW and is therefore relevant for the eye. Light emitted outside the SH and encoded with the wavefront information of the object point would not reach the VW and would therefore be wasted. This is indicated in Fig. 1 by the green spherical wavefronts for the essential information and the red spherical wavefronts for the wasted information.

Managing Pixel Size via a Tracked Viewing Window

The requirements on the pitch of the SLM are significantly lessened by our approach. Our holographic displays are equipped with an eye-position sensor and a tracking system that always positions two VWs at the observer's eyes, i.e., one VW for the left eye and one VW for the right eye. This allows reducing the size of a VW to the size of an eye pupil. As noted above, the VW is the Fourier transform of the hologram. The pixel pitch of the SLM determines the diffraction angle, which in turn determines the size of the VW. A moderate pixel pitch p = 50 μm generates a VW with lateral size h = 20 mm at a distance d = 2 m, using h = λd/p with λ = 500 nm. These are typical values for a holographic display for TV applications.

These VWs, which may be generated by temporal or spatial multiplexing, can be increased in number to allow for multiple simultaneous viewers.

The size of the reconstructed scene is not limited by the pixel pitch but by the size of the SLM. The 3-D scene can be located anywhere in a frustum defined by the VW and the SLM. This frustum is indicated by the dashed blue lines in Fig. 1. The 3-D scene can be located in front of and behind the SLM. This contrasts with the conventional approach, in which the pixel pitch limits the size of the reconstructed scene.

Encoding Sub-Holograms in Real Time

The second major hurdle addressed was data volumes and computation speeds. Figure 2 illustrates how encoding the wavefront information of an object point in a small sub-hologram (SH) significantly reduces the computation effort.

Projecting the VW through an object point determines the size and position of its associated SH. This is indicated by the lines from the edges of the VW through each object point onto the SLM/hologram display. The hologram in each SH is a spherical phase factor that reconstructs its associated object point at the defined distance from the SLM. Super-imposing all the individual SHs yields the final hologram of the complete 3-D scene.

The SH size depends on the VW size and the object point position and is typically on the order of 10 mm (about 0.4 in.). This is much less than the total hologram size of at least 20 in. and more. Therefore, the computation effort is significantly reduced, as each spherical phase factor has to be calculated for the small SH only. In contrast, the conventional approach to holographic displays requires calculation of the wavefront informa-tion of each object point on the whole hologram.

In actual computational terms, this only takes single-digit TFLOPS to create full-HDTV 3-D images in real time. This will be well within the GPU domain in the very near future and already allows adaptation to existing ASIC architectures.

Fig. 1: Side view of a holographic display with lens, SLM, one object point, and observer eye.

Reducing Speckle and Higher-Order Effects

The reduction of speckle is carried out by conventional approaches as well as proprietary encoding techniques, again enabled by the VW concept. Because sub-hologram encoding requires only local coherence across the area of a SH, spatial and temporal coherence of the light source can be reduced without affecting the object resolution visible from the VW. Further speckle reduction is achieved by spatial and temporal averaging of the 3-D scene. These methods reduce speckle by more than 95%.

Because a holographic display is based on the diffraction of light at the SLM, higher diffraction orders of diffracted light occur. The use of sub-holograms already eliminates higher-order effects within the VWs because the VW size is limited to one diffraction order of the SLM. To assure noise-free views for all users, additional solutions have been developed based on simple modifications of hardware components that sufficiently reduce higher diffraction orders outside the VW.

Conclusion

The essential idea of our approach is that for a holographic display the highest priority is to reconstruct the wavefront at the eye position that would be generated by a real existing object and not to reconstruct the object itself. The tracked VW technology limits pixel size to levels already known for commercially available displays. Sub-hologram encoding brings computation into graphics card or ASIC range. SeeReal's new concept is applicable to desktop, TV, and mobile imaging.

While there have been impressive developments in 3-D display technology in the past decade, the remaining visual conflicts between natural viewing and 3-D stereo visualization have prevented 3-D displays from becoming a universal consumer product. In principle, the only 3-D display capable of completely matching natural viewing is an electro-holographic display.

SeeReal's new approach to electro-holography not only proves that it is possible, but that it is also closer to adoption than many experts imagined. The principles and concepts are in place. The checks and verifications are com-pleted. Prototypes are in full use (Fig. 3). The technology already exists – it is just a question of time for all the pieces of the puzzle to come together and for the first commercial real-time 3-D holographic displays to hit the market.

References

¹K. Maeno, N. Fukaya, O. Nishikawa, K. Sato, and T. Honda, "Electro-Holographic Display Using 15-Mpixels LCD," Proc. SPIE 2652, Practical Holography X, 15–23 (1996).

²C. Slinger, C. Cameron, S. Coomber, R. Miller, D. Payne, A. Smith, M. Smith, M. Stanley, and P. Watson, "Recent developments in computer-generated holography: toward a practical electroholography system for interactive 3D visualization," Proc. SPIE 5290, Practical Holography XVIII, 27–41 (2004).

³P. St.-Hilaire, S. A. Benton, M. Lucente, J. D. Sutter, and W. J. Plesniak, "Advances in Holographic Video," Proc. SPIE 1914, Practical Holography VII, 188–196 (1993).

⁴A. Schwerdtner, N. Leister, and R. Häussler, "A New Approach to Electro-Holography for TV and Projection Displays," SID Symposium Digest Tech Papers 38, 1224–1227 (2007). •

Fig. 2: An illustration of how sub-holograms are encoded in real time.

Fig. 3: SeeReal's prototype of a holographic display.