DE69019043T2 - Apparatus and method for a digitized video system. - Google Patents

Description

BACKGROUND OF THE INVENTION

1. Field of the invention

The invention relates to devices and methods for producing digitized video systems as well as to integrated circuits and devices intended therefor.

2. Description of the state of the art

The invention is applicable to the field of graphics and video display systems such as displays for computer systems, terminals and televisions. Recently, a demand has arisen for viewing areas that are larger and/or have higher resolution than is possible with conventional video display devices such as cathode ray tubes (CRTs) or liquid crystal displays (LCDs). This has been driven by consumer demand for larger television sets (TVs) and the need for presentation or computer-based screens at conferences that are designed for larger audiences.

LCDs, used for small computer systems and terminals, especially laptops and portable computers, use individual liquid crystal cells for each pixel on the LCD. LCDs are temperature sensitive, difficult to manufacture in large sizes, slow to change state, and require external light sources for viewing.

To overcome the dimensional limitations of LCDs, attempts have been made to develop projection systems that use an LCD as a spatial light modulator (SLM). Unfortunately, some problems remain. The LCD has an inherent slowness, so a rapidly changing image "shines". The resolution of the LCD is limited by the driver complexity. The driver complexity also requires that the size of the LCD is generally proportional to the resolution. This means that the optical projection devices for a high resolution system must be large and correspondingly expensive. Another problem is that the light transmitted by (or reflected from) the LCD becomes polarized. This can lead to a non-linear perception of brightness from the center of the field of view to the periphery.

The most common display system is the CRT (cathode ray tube). In a cathode ray tube, a scanning electron beam of variable current density is passed across a light-emitting phosphor screen. This light-emitting phosphor screen is struck by the electron beam and produces light depending on the magnitude of the electron beam's current density. These tubes can be used in direct viewing or projection mode. However, they have numerous disadvantages. The first of these is cost.

The higher cost is dictated by the difficulty of manufacturing large display tubes (currently 45-inch tubes are manufactured). Another reason for the cost is the large amount of raw materials required (particularly glass). This results in a very heavy display that is not easily transportable.

Resolution is a problem for CRTs. There are two main reasons for this. The first concerns the shadow masks used in color CRTs. A shadow mask is used to separate the color phosphors used to create the three primary colors (red, blue, and green) and to help guide the electron beam used to excite the phosphors. The brightness of a pixel is related to the size of the phosphor dot. However, with a larger phosphor dot, the shadow mask must be made larger and it becomes more visible. The brightness is also related to the energy of the electron beam. If the energy increases, so does the brightness. Unfortunately, the shadow mask is also sensitive to the electron beam and is thermally deformed at high energy. The image then becomes blurred, both because the shadow mask becomes more visible and because the electron beam is deflected towards an undesirable phosphor.

The second limitation on resolution is the rasterization. All pixels to be illuminated are scanned sequentially by an electron beam. This beam is moved back and forth across the phosphors in a raster pattern. In general, the beam is turned off when it is returned across the phosphors (known as the retrace time), and it is also turned off when it returns to the starting point (vertical blanking interval). This is not a theoretical limitation (all phosphor dots can be accessed). It is rather a practical limitation. This is because the fluorescence of the phosphors decreases as the electron beam moves to the next location. The electron beam must return before the human eye notices the decrease, otherwise the display will flicker. To compensate, longer duration phosphors can be used, but these have the disadvantage of a smearing effect when the display data changes.

Rasterization has another typical side effect. It sets an upper limit to the perceived brightness of a display. As explained above, a phosphor can only be exposed for a very short period of time and then it begins to fade. If the phosphor is exposed hard, it begins to grow larger (ie it begins to excite neighboring pixel locations) and blur the display. If the phosphor is excited continuously for a longer period of time, it appears brighter than in the Case where it is only excited during the raster period. This is due to the fact that the human eye requires an integration time of about .1 seconds for bright light sources and .2 seconds for dim sources.

Projection CRT systems do not have the shadow mask problems. However, they are expensive because they typically require three CRTs (one each for red, blue and green). They also have the serious disadvantage of low brightness (because the image produced has to be expanded). This is particularly true when a single CRT is used in a projection mode. Each type has all of the other problems related to the raster. Moreover, they are very large when used in rear projection configurations due to the complex optical paths required.

Another disadvantage of conventional display systems is that they are primarily analogue. Even if the information to be displayed is stored in digital form, for example in a computer, it must be converted into an analogue raster before it can be displayed on the cathode ray tube.

Other spatial light modulators have been used in projection displays. For example, the use of a spatial light modulator control for a display system is described in U.S. Patents 4,638,309 and 4,680,579 issued to Ott and incorporated by reference. In Ott, a deformable semiconductor mirror array is used in conjunction with Schlieren optics to form the spatial light modulator. Deformable mirror arrays are described in U.S. Patents 4,441,791, 4,710,732, 4,596,992, 4,615,595, and 4,662,746 and post-issued U.S. Patent No. 5,061,049 to Hornbeck, all of which are incorporated by reference.

Another display using a light tube is shown in U.S. Patent 3,576,394 to Lee, which is hereby incorporated by reference. Various types of human factors information on the critical flicker frequency are provided in "Applied Optics and Optical Engineering" (1965), Volume 11 (The Detection of Light and Infrared Radiation), by Rudolf Kingslake, which is hereby incorporated by reference. Acousto-optic spectral filters are shown in I.E.E.E. Transactions on Sonics and Ultrasonics, vol. su-23, No. 1, January 1976, pages 2-22, which disclosure is hereby incorporated by reference.

A HDTV (High Density TeleVision) system is shown in U.S. Patent 4,168,509 to Hartman, which patent is incorporated by reference. Various types of electronic TV tuners are shown in U.S. Patents 3,918,002, 3,968,440, 4,031,474, 4,093,921, and 4,093,922, which are incorporated by reference. Various multi-frequency responsive display materials are shown in SPIE, Vol. 120 (Three-Dimensional Imaging, 1977), pages 62-67, "PRESENT AND POTENTIAL CAPABILITIES OF THREE-DIMENSIONAL DISPLAYS USING SEQUENTIAL EXCITATION OF FLUORESCENCE" by Carl M. Verber, and IEEE Transactions on Electron Devices, Vol. ED. - 18, No. 9 (September 1971), pages 724-732, "A True Three- Dimensional Display" by Jordan D. Lewis et al., which are incorporated herein by reference. One type of display is shown in Information Display, November/December, 1965, pages 10-20, "Three Dimensional Display Its Cues and Technigues" by Petro Vlahos, which is incorporated herein by reference.

Laser (Light Amplification by Stimulated Emission of Radiation) based projection systems are well known in the art. These systems can also use fluorescent dyes with invisible laser light. This follows from SID INT. SYMP. DIGEST, Paper 10.1, May 1983, "Projection Display of Radar Image using Gas Laser and Organic Fluorescent Pigment Screen" by H. Yamada, M. Ishida, M. Ito, Y. Hagino and K. Miyaji, which is incorporated herein by reference. Further details on various dyes are given in CHEMISTRY AND CHEMICAL INDUSTRY, Vol. 23, No. 3, 1970, "Increasing Application Field for Fluorescent Pigment" by R. Takano, which is incorporated herein by reference.

The operation of laser-based displays is based on deflecting a beam of coherent light produced by a laser to produce an image. The deflection units include devices such as rotating mirrors and acoustically modulated deflectors. These projectors suffer from a number of problems which have prevented their commercial adoption.

The first of these problems is flicker, which also places an upper limit on the resolution that can be achieved (i.e. the number of pixels that can be displayed). Only one point of light (pixel) can be displayed at any given moment, due to the nature of the deflection devices. There is also no lighting duration in this display, since these projectors generally direct the light onto a diffusing surface and no means are provided to continue to emit light after the light has been deflected. This means that all points to be displayed must be illuminated within a period of time that is less than the critical flicker frequency (CFF) of the human eye.

A second problem is laser speckle. This is considered to be a random interference pattern of intensity resulting from the reflection or transmission of highly coherent light from or through an optically rough surface (with depth-given local irregularities greater than one-quarter of a wavelength). This phenomenon is addressed in the JOURNAL OF THE OPTICAL SOCIETY OF AMERICA, vol. 66(11), 1976, page 1316, "Topical issue on laser speckle" by N. George and DC Sinclair; APPLICATIONS OF OPTICAL COHERENCE (WH Carter, Ed.), 1979, pages 86-94, "Role of coherence concepts in the study of speckle" by JW Goodman and COHERENT OPTICAL ENGINEERING (FT Arecchi and V. Degiorgio, Eds.), 1977, pages 129-149, "Speckle interferometry" by AE Ennos, all of which are incorporated herein by reference. Techniques for reducing speckle are also provided in JOURNAL OF THE OPTICAL SOCIETY OF AMERICA: PART A. Vol. 5(10), 1988, pages 1767-1771, "Effect of luminance on photoptic visual acuity in the presence of laser speckle" by JM Artigas and A. Felipe and OPTICS COMMUNICATIONS, Vol. 3(1), 1971, "Elimination of granulation in laser beam projections by means of moving diffusers" by E. Schroder, all of which are incorporated herein by reference.

Another problem was the production of color images. This requires the use of multi-color lasers. There are great technical difficulties both in aligning multiple deflection units and in keeping them synchronized so that the different colors are imaged simultaneously at a given pixel location.

In JP-A-55-65940, small vibrations of a screen are used to eliminate laser speckle from an image projected onto the screen by laser light.

EP-A-0 139 991 describes a projection display device using an array of individually controlled deformable mirror cells, in which the degree of deformation of a cell determines the light intensity at the corresponding pixel of the displayed image.

SUMMARY OF THE INVENTION

An object of the invention is to reduce the formation of spots in a display.

According to one aspect of the invention there is provided a projection light image display system with reduced staining comprising:

a coherent light source for generating a light beam,

a screen,

a spatial light modulator responsive to signals representing an image to be displayed for directing light from the source to the screen to display the image on the screen,

this system being characterized in that a transducer is coupled to the screen and serves to generate acoustic surface waves in the screen where the image is displayed.

According to a second aspect of the invention there is provided a method of reducing speckle in a projection light image display system in which light from a coherent light source is directed through a spatial light modulator onto a display screen to display an image on the screen, and characterized in that speckle in the displayed image is reduced by generating surface acoustic waves on the screen.

The preferred embodiments shown here illustrate various concepts in the field of digitized video systems for two- and three-dimensional images. A deformable mirror assembly is shown that can receive an image or a portion of an image for display while another image or another portion of the image A true three-dimensional video display system has a display that can be solid or gaseous.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be described with reference to the accompanying drawing, in which:

Figures 1a, 1b, 1c and 1d show a two-dimensional digital video system using a deformable mirror array;

Figures 2a, 2b and 2c show a two-dimensional digital video system using a deformable mirror array and a laser;

Figure 3 is a block diagram of the electronics for a digitized video system;

Figure 4 shows a two-dimensional digitized video system for producing a color image;

Figures 5a, 5b and 5c diagrams and a color wheel;

Figure 6 shows a two-dimensional display;

Figure 7 is another view of the display of Figure 6;

Figure 8 is a plan view of a spatial light modulator;

Figure 9 is a circuit diagram of a memory cell of a deformable mirror arrangement;

Figure 10 is a flow diagram for the electronics of Figure 3;

Figure 11 a three-dimensional digitized display system;

Figures 12 and 13 show a pointer for the system of Figure 11;

Figure 14 shows a pointer for the system of Figure 11;

Figure 15 shows a pointer for the system of Figure 11;

Figure 16 shows a processor for a multi-dimensional arrangement; and

Figure 17 shows a display with a movable surface.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

Figures 1a, 1b, 1c and 1d show a preferred embodiment of a two-dimensional digitized video system 75 comprising an imaging system 1 and a screen 2. The screen 2 may be a relatively flat plate made of a suitable material, or it may have a curved configuration to focus reflected light toward a viewer. The screen 2 may be translucent to allow for rear projection. In reflective mode (or front projection), the screen 2 may be composed of a rigid material such as plastic, metal, etc. and have a reflective surface. The surface may be frosted or have a lenticular pattern as is well known in the art. In the rear projection mode, the screen 2 may be made of glass or a translucent plastic and may have a patterned surface such that light incident on it from the imaging system 1 is scattered. Rear projection screens of this type are well known in the art.

A light source 10 is used to generate light energy to ultimately illuminate the screen 2. The light source 10 may be of conventional construction such as incandescent, halogen or arc lamp type or any other suitable form. In addition, the light 9 generated is concentrated and directed by a mirror 11 onto lenses 12. The mirror 11 may have any suitable shape such as parabolic, ellipsoidal, etc.

The lenses 12, 13 and 14 form a beam splitter which serves to form a light slit 8 from the light 9. This is done to concentrate the light energy and to increase the efficiency of the entire system. A beam bending mirror 20 is used to direct the slitted light 8 via a path 7 to a spatial light modulator (SLM) 15. Of course, other structures can be used to direct a source of concentrated light energy to an SLM 15 without departing from the scope of the invention.

The SLM 15 serves to selectively direct portions of the light from the path 7 back to a magnifying lens 5 and onto the screen 2 to form an image. In the preferred embodiment of the invention, the SLM 15 is of a known type known as a deformable mirror array (DMD), which is discussed in more detail below. Other SLMs (such as Bragg cells, LCDs, etc.) can be used in either reflective (as shown) or transmissive mode, provided the individual light beams can be redirected at a sufficiently high rate. The importance of the need for high switching speeds will become apparent later.

The SLM 15 has a surface 16 onto which light from path 7 impinges. The surface 16 has a number of switchable elements (as at 17) which are controllable, to redirect light to the magnifying lens 5. For example, if the element 17 occupies one position, a portion of the light from path 7 is redirected along path 6 to the lens 5 where it is magnified or spread along path 4 to impinge on the screen 2 and form an illuminated pixel 3. If the element 17 occupies another position, the light is not redirected to the screen 2 and the pixel 3 thus remains dark.

A computer 19 controls the operation of the SLM 15 via a bus 18 to produce the image on the screen 2 by selectively redirecting portions of light on the path 7. The bus 18 provides the necessary control signals and image information from the computer 19 to the modulator 15. The computer 19 may be, for example, a digital signal processor (which is explained in more detail below).

In the preferred embodiment of the invention, the surface 16 contains an array of deformable mirror cells. Deformable mirror cells suitable for use in the array 16 are shown in Figures 1b, 1c and 1d. Four cells 17, 27, 37 and 47 are shown in Figure 1b. The mirror 32 of the cell 17 is connected to the modulator 15 by a pivot 30 at approximately the location shown in Figure 1b. When the cell 17 is actuated, the mirror 32 is pulled downward from its position shown in Figure 1c to its position shown in Figure 1d. When the cell 17 is in the lower position, it directs light along the optical path 6. When the mirror is in the upper position shown in Figure 1c, the portion of the beam from the light path 7 is diverted away from the light path 6 toward the display 2. The other portions of the assembly 16 that are not pivotally mounted, such as the surface portion 34, also do not direct light toward the display 2.

In the currently preferred embodiment, it can be seen that light is directed to the screen only when a mirror (such as 32) is in the lower position. The This is because other parts of the assembly 16 (such as the surface portion 34) may also be reflective and could add visible noise to the display. In the preferred embodiment, the light path 7 is substantially perpendicular to the SLM 15 and the light path 6 is inclined. Of course, other embodiments are possible without departing from the scope of the invention. For example, the light path 7 could form an angle with the SLM 15 and the light path 6 could be formed when a mirror element such as 32 is in the upper position (or perpendicular to the surface portion 34). A Schlieren aperture could then be inserted in front of the lens 5 to block out unwanted light such as that from the surface portion 34.

Figure 2a shows another preferred embodiment of a two-dimensional digitized display system of the invention. This includes an imaging system 50 and a display 51. The display has a pixel 52 which is one of many pixels that make up the image to be displayed on the screen 51. The pixel 52 is arranged along a light path 53 from an image magnification system 54 which may be of any suitable type such as a lens system, an optical fiber array, etc.

A laser 61 generates a beam of light 62. A beam expander 60, which may be of any suitable type, serves to expand the small diameter beam of light 62 from the laser 61 to a large diameter beam 63. This beam is then deflected by a beam bending mirror 55 along path 64 to the SLM 56. In this currently preferred embodiment, the SLM 56 has a different structure (as will become apparent below from the discussion of Figures 2b and 2c) than that shown in Figures 1a, 1b and 1c. Mirror elements of the array 57 such as 58 are movable between two positions. In the "on" position, element 58 deflects a portion of the light from path 64 along path 66 to the magnifying lens 54 where it is magnified (spread out) along path 53 to impinge on the screen 51 and form the illuminated pixel 52. In the "off" position, the light is directed along path 65 where it does not impinge on the screen 51.

The screen 51 may be of conventional construction such as screen 2 of Figure 1a. However, as discussed in the BACKGROUND, unless the screen 51 is optically smooth, laser speckle is likely to occur. For practical reasons such as manufacturing difficulties, this is unlikely. Furthermore, some type of scattering surface (such as ground glass) should be used to increase the viewing angle. This will increase the amount of speckle unless the degree of scattering almost completely dephases the incident coherent light. Unfortunately, this has the side effect of blurring the image and thus reducing the apparent resolution.

The invention overcomes this speckle problem in several ways. Images formed on the screen 51 are composed of several rays of light simultaneously striking the surface. A given ray of light will strike the screen at a slightly different angle than any of the others due to the effect of the magnifying lens system 54. Furthermore, each ray of light will have a slightly different light path length. This is further increased by the spacing between the respective mirror elements of the array 57. Due to the differences in the relative angles and path lengths of the simultaneously striking rays of light, the relative phase of the rays will be different when they strike the screen 51. This has the effect of reducing the overall coherence of the light image and thus reducing visible speckle without reducing resolution.

The screen 51 also has a transducer 90 coupled thereto. The transducer 90 establishes surface acoustic waves 91 that pass the screen 51 with substantially parallel waves. An optional wave absorber 92 absorbs waves that have passed the screen 51 to prevent reflections. In other embodiments, the transducer 90 may generate arc or circular waves. The transducer 90 may be of conventional construction and is preferably used to generate signals in the ultrasonic range, which are beyond the range of human hearing. The amplitude of the surface acoustic waves 91 is greater than a wavelength of light. This serves not only to bring the relative light rays out of phase, but also to scatter a single beam. The reason for this is that the angle at which the beam strikes the display surface 51 (and hence the angle at which the beam is reflected or diffracted) changes due to the surface acoustic wave. The screen 51 can be used in either front projection or rear projection mode.

In another embodiment of the invention, the screen 51 may contain or be coated with materials known in the art (such as uranyl ions, lanthanum ions, erbium ions, organic fluorescent daylight dyes, etc.) that exhibit optical fluorescence when excited by non-visible light. The laser 61 generates either UV or IR light (depending on the materials selected). Preferably, the screen 51 is used in the rear projection mode. The side of the screen facing the viewer (the side opposite the side facing the imager 50) is additionally coated with a material that absorbs the light from the laser 61 so as to minimize visible interference. Optionally, the coating may reflect the light. For reasons discussed below, the fluorescent material has a decay time that is significantly less than phosphors used in conventional CRTs.

The SLM 56 is coupled to the computer 59 via a control bus 70. The computer 59 controls each mirror element (such as 58) of the array 57. These mirror elements are switched between the "on" position and the "off" position to produce the image on the screen 51.

In the additional embodiment of the invention, the computer 59 is also coupled to the laser 61. The laser 61 is of the multi-mode or pulsed type well known in the art. The computer 59 controls the output light frequency of the laser 61. By techniques discussed below, a color display can be formed by sequentially timing the frequencies generated by the laser 61. Similarly, the screen 51 can comprise a plurality of materials, each of which fluoresces at a different color depending on the frequency of the incident light rays.

A portion of the assembly 57 is shown in Figure 2b with deformable mirror cells 48, 72, 73 and 74. Only cell 48 will be discussed in detail. The mirror 110 of cell 48 is movable about pivots 112 and 113. The movement of the mirror 110 is best shown in Figure 2c. The mirror 110 is movable about an axis 116 from the position shown by a dashed line 118 to a position shown by a dashed line 134 with respect to the normal surface 120.

In the "on" position, the edge 120 of the mirror 116 contacts the landing electrode 122. The mirror 110 is moved to the "on" position by applying the appropriate voltage to the control electrode 124. The voltage at this electrode is applied to the positive electrode 128 and through an inverter 129 to the negative electrode 130. A differential bias is applied to the mirror 130 via an electrode 132. In the position shown by line 118, the mirror 130 directs a portion of the light beam from the light path 64 along the light path 66, through the system 54 along the light path 53 to the pixel 52 on the display 51. When a negative voltage (the "off" position) is applied to the electrode 124, the mirror 110 rotates to the position shown by the dashed line 134 in Figure 2c and directs a portion of the light beam along the light path 65.

In Figure 3, a signal source 140, which may be, for example, a TV tuner such as one of those included above, is connected to an electronic system 144 via a bus 142. The bus 142 is connected to an analog-to-digital (A/D) converter 146. The analog signal received by the A/D converter 146 from the bus 142 is converted by the A/D converter 146 into digital codes on the bus 148. These digital codes are representative of the chrominance and luminance information of each pixel of an image. Via the bus 148, the digital codes from the converter 146 are transferred to a buffer memory 150. The digital codes from the converter 146 are stored in the buffer memory 150. In another mode, digital codes or information may be loaded into buffer memory 150 from, for example, a computer or graphics system via bus 152. However, in some embodiments of the invention, all information may come via bus 152 and signal source 140, bus 142, converters 146 and 148 are not required.

Buffer memory 150 may be a single port random access memory (RAM). In this case, bus arbitration between buses 148, 156 and 152 may be performed in a conventional manner by direct memory access (DMA) controllers, as is well known in the art. First priority is given to bus 148 (since there is no practical way to stop data from arriving from signal source 140). Second priority is given to bus 156 because it must receive display data very quickly. Alternatively, buffer memory 150 may be of the two-port or three-port type for increased throughput. The construction of multi-port memories is also well known in the art.

The digital codes or information collectively represent an image to be displayed. The digital codes in the buffer memory 150 are transferred to a central processing unit (CPU) 154 via bus 156. The CPU may be a conventional microprocessor such as a TMS 99000 (manufactured by Texas Instruments) or, more desirably, a digital signal processor (DSP) such as the TMS 320C10, TMS 320C20, TMS 320C25 and TMS 320C30. Details on the use and construction of these DSPs can be found in Digital Signal Processing - Products and Applications "Primer" (1988), TMS 320C1x User's Guide (1987), TMS 320C10 User's Guide (1983), TMS 320C2x User's Guide (1987), TMS 32020 User's Guide (1986), TMS 320C25 User's Guide "Preliminary" (1986), and TMS 320C30 User's Guide (1988), all of which are incorporated by reference.

The CPU 154 is connected via bus 158 to a video memory 160, which is preferably made of video DRAMs (VRAMs) such as the TMS4461 from Texas Instruments, Inc., Dallas, Texas. Preferably, multiple VRAMs are used, with the high-speed serial outputs of each VRAM corresponding to some of the columns of the generated display on the screen 178. This is done to increase the load bandwidth of the planar projector 172. The reason a large load bandwidth is required is discussed below.

The CPU 154 decodes the video information, including the chrominance and luminance, within the information from the memory 150. The CPU 154 is programmed to extract an image from this information and store this image including the chrominance and luminance in the video memory 160. The image may also be modified by the CPU 154 under commands via bus 162 or under the control of its program. The buffer memory 150 and the video memory 160 include a memory 164 and may be constructed as a single memory. Moreover, the video memory 160 may be loaded directly from a graphics system or a computer 166 via bus 168.

The electronics system 144 and the projection system 172 include an imaging system 174 in which the imaging system 1 of Figure 1a may be used, with the spatial modulator 15 connected to the video memory 160 via bus 18. Similarly, the memory 160 may be connected to the modulator 56 in Figure 2a via bus 70. In other words, the imaging system shown in Figure 3 contains more electronics details, and Figures 1a and 2a contain more optics details, and any of the various embodiments disclosed herein may be combined using the techniques presented herein to construct various digitized display systems. The image stored in memory 160 is transferred via bus 170 to projection system 172 and to display 178 via light path 176 for display as in the systems shown in Figures 1a-1d and 2a-2c.

From the above descriptions, it is apparent that the projector system 172 is a planar projector. In other words, all of the pixels to be displayed on the screen 178 are illuminated simultaneously rather than sequentially by a raster. Moreover, data supplied to the projector 178 comes from the memory 160, while incoming data is temporarily stored in the buffer memory 150. Thus, the transfer rate for data to the projector 172 is completely independent of the transfer rate of data received from the signal source 140. Thus, the present invention is independent of the raster speed.

In the United States, the television broadcast standard is NTSC. This requires an interline pixel rate of 60 hertz. In other countries, pixel rates may be 50 hertz. The raster speed is, of course, directly related to the pixel rate. Since the present invention is independent of the raster speed and pixel rate, it can be freely used in any country. The CPU 154 only has to detect the frequency of the synchronizing signal (or use the color burst signal or other known techniques) to determine how to store the data in the memory 160. In an alternative embodiment, a switch is provided to allow the user to select the standard used. In another embodiment, the signal source 140 may provide the CPU 154 with a signal associated with the broadcast frequency to which it has been tuned to inform the CPU 154 of the broadcast standard used.

In the presently preferred embodiment of the invention, the SLM used in the projector system 172 contains a 1280 by 800 cell array. Thus, the display image on the screen 178 has a resolution of 1280 by 800 pixels (each pixel corresponding to one cell). Of course, any size can be chosen, depending on the application.

The NTSC broadcast standard has an image resolution of approximately 320 by 200 pixels. A simple way of displaying NTSC data on the display of the present invention would be to control a subarray of 16 cells (i.e., a 4 by 4 subarray matrix) for each of the pixels. However, this would result in an enhanced projected image appearing very coarse. In the presently preferred embodiment, the NTSC data loaded into the buffer memory is processed by the CPU 154 before being loaded into the memory 160. This Processing uses the well-known nearest neighbor algorithm to interpolate intermediate display pixels in both the horizontal and vertical directions, since at least an entire frame of data (or at least several lines of data) may be stored in the buffer memory 150. Thus, for each pixel of NTSC data, 15 additional pixels may be calculated by the CPU 54. Additional and/or alternative processing including algorithms other than the currently preferred nearest neighbor algorithm may be used. This serves to smooth out the image and increase the visible resolution.

Similarly, if the incoming data has an image resolution greater than that which can be displayed by the projector 172 (i.e., greater than the number of cells in the SLM), the CPU 162 applies processing routines to the data in the buffer 150 before storing the results in the memory 160. This processing again relies on looking at a matrix subset (i.e., horizontally and vertically adjacent pixels) of the image data in the buffer 150. A calculated pixel is then output to the memory 160 for display by the projector 172 on the screen 178. Since each pixel to be displayed is calculated, it is not necessary for the resolution size of the incoming data to be an integer multiple of the display resolution. Thus, the invention is completely standard independent in terms of both data transfer speed and resolution.

Figure 4 shows another embodiment of the invention. In Figure 4, the imaging system 210 has an electronic system 212 and a projection device 214. This is a color system with a green laser 216, a red laser 218 and a blue laser 220. The green laser 216 generates a beam of green light along the light path 222. The red laser 218 generates a beam of red light along the light path 224. The blue laser 220 produces a beam of blue light along the light path 226. A combination mirror 228 passes the green light from the light path 222 through a light path 230 and redirects the red light from the light path 224 to the light path 230. Another combination mirror 232 is arranged along the light path 230 to pass the green and red light from the light path 230 through the light path 234 and to redirect the blue light from the light path 226 to the light path 234.

A beam expander 236 expands the light from the light path 234 into a larger diameter beam along the light path 238. The light from the light path 238 is redirected by the mirror 240 to a spatial light modulator 242, which may be a deformable mirror arrangement such as the device shown in Figures 2b and 2c. The mirror 240 redirects the light from the path 238 to the light path 250. The light from the path 250 strikes the spatial light modulator 242. When a positive voltage has been applied to the deformable cell 246 with the control signals from the memory in the electronics system 212 (only the computer system 243 and the bus 244 are shown), it rotates to the left (as shown in Figure 4) or to the right when a negative voltage is applied. When a negative voltage is applied to the deformable cell 248 using the control signals from the memory in the electronic system 212 (only the computer system 243 and the bus 244 are shown), it rotates to the right (as shown in Figure 4), or it rotates to the left when a positive voltage is applied. This is explained in more detail below with reference to Figures 8 and 9.

As shown in Figure 4, the portion of the light from the light path 250 associated with the cell 246 is redirected along the light path 252. The light path 252 does not intersect the projection system 254 or the display (or screen) 256. The portion of the light from the path 250 associated with the cell 248 is redirected along light path 258. Light path 258 intersects projection optics 254. The light beam redirected by cell 248 along light path 248 is expanded and directed by the projection optics onto display 256 along with any other cells (not shown) on spatial light modulator 242 that are rotated to direct light to projection optics 254. This produces an image on display 256 which is arranged along light path 260 from projection optics 254. Lasers 216, 218 and 220 are activated sequentially by electronics system 212. Computer 243 loads the appropriate information for each color. For example, if green laser 216 is to be activated, information for the green portion of the image is loaded into modulator 242. This is explained in more detail below. The projection optics 254 may be, for example, a lens system or an optical fiber arrangement.

The invention is concerned with the critical flicker frequency (CFF) of the human eye. The CFF is the frequency at which the eye is no longer able to perceive a flickering (i.e. rapidly changing between dark and light) image as various flashes, but only perceives it as a continuous light source. This frequency changes with the intensity of the source. For example, at low brightness (retinal illumination of 1.6 photons) the CFF is about 5 Hertz. At higher luminance levels (retinal illumination of 5 photons) the CFF is greater than 60 Hertz. The eye has an integration time of 100 to 200 milliseconds (depending on intensity) before it can perceive the relative intensity of a light source. Therefore, the brightness for each displayed pixel can be changed by the inventive system by rapidly modulating a constant light source.

In a similar way, the eye also has an integration time for color. This means that when several colors are seen, the eye merges them into a single color. For example, if the primary colors red, green, and blue are presented in rapid succession, the eye sees a white source. By changing the intensity of each primary color (including through the time modulation discussed above), any color can be selected.

The SLM of the invention can be modulated at a very high frequency. For example, the mirror cells have a switching time between on and off of about 10 microseconds. Likewise, the array can accept control data at a very high rate. How this is achieved is shown in Figures 8 and 9. In the present embodiment of the invention, the entire mirror cell array of the SLM can be charged and each cell switched during a time period of 20 microseconds.

Because of this high switching capability, the SLM of the invention can modulate each pixel at a data rate that is 833 times faster than the CFF for a bright image. At lower levels of desired luminance, the deviation rate is of course much higher. With the invention, a wide dynamic range can be achieved in both chrominance and luminance due to this speed.

An example of how different luminance levels are achieved is shown in Figure 5a. For simplicity, each time period (T1, T2...T25) is assumed to be 4 milliseconds, although as stated, the SLM of the invention operates much faster. This is done only to show that different luminance levels can be achieved even at these low rates and the speed is still higher than the critical flicker frequency. Each large time increment (corresponding to 4 time slots of 4 milliseconds each) represents the same pixel location on the screen and the desired intensity obtained from the signal source. The horizontal lines represent the signal provided by the SLM. For example, if the horizontal line is in the lower position, the SLM will not direct light to the screen, whereas if the horizontal line is in the higher position, the light will be directed to the screen.

The first large increment (with time slots T2 to T5) shows that the SLM is not directing any light towards the screen. Thus, the displayed location of the pixel is not illuminated. During the second increment (with T6 to T9) the pixel is fully illuminated and assumes its state in which it is visibly brightest. The next increment (represented by T10 to T13) shows half the intensity. In other words, for half the total time no light is directed to the screen, whereas for half the time light is directed to that screen.

The next two large increments (represented by T14 to T17 and T18 to T21) are of equal intensity. This means that the perceived intensity of the light is brighter than the previous increments, but less than the second. However, the eye can perceive the increment from T18 to T21 as just brighter than that from T14 to T17. This is because the eye has an integration time for intensity. Since there is an extended period (T15 to T20) when the light is on, the eye begins to integrate this as a high level of brightness and can therefore perceive the period T18 to T21 as brighter. Changes in the pattern can thus be used to further increase the apparent dynamic range of brightness.

Figure 5b shows how this dynamic range is further extended. During each of the large time increments (T2 to T5, T6 to T9, etc.) the light source itself was modulated in intensity. While the pattern shown is of a sawtooth type, any pattern such as a logarithmic, exponential, etc. Looking at Figures 5a and 5b together, it can be seen that the perceived brightness of the increments T10 to T13 is greater than that of T22 to T25, although both increments have two on states and two off states.

Modulation of the light source can be achieved in a variety of ways. For example, in the laser-based projection system of Figure 2a, the laser can be controlled either by the computer 59 or by other circuitry to rapidly change its intensity of generated light. For conventional light source generator systems as shown in Figure 1a, a variable pitch wheel as shown in Figure 5c can be rotated in the light path in front of the SLM 15. This rotating wheel would preferably be placed between the light source 10 and the lens 12 to help maintain the slit beam.

Color can be added by providing different primary colors (e.g. red, green and blue) in succession for a time interval less than the CFF time. The eye will temporarily integrate different colors into a single color if they are provided in rapid enough succession. In the system of Figure 1a, color can be added by rotating a wheel such as that shown in Figure 5c in the light path leading to the SLM 15. By using the wheel of Figure 5c in this way, each major section thereof acts as a different color filter. For example, one section could filter all but red, the next all but blue, and the third all but green. Thus, a single wheel would provide both brightness control and color control. The wheel shown in Figure 5c is for illustration purposes only, as the wheel could be divided into multiple color sections, or contain multiple parts, or have other suitable configurations. In other embodiments, other filter systems may be used. such as acousto-optical spectral filters.

Color can be added to the system of Figure 2a by using a multimode or pulsed laser 61. Each color could be selected by tuning the laser 61 to a different frequency at a relatively fast speed. The system shown in Figure 4 has three color lasers (216, 218 and 220). These would be sequenced. Alternatively, all pixels of a given color (derived from various combinations of the three lasers 216, 218 and 220 simultaneously producing light at different intensities) could be displayed simultaneously. The next group of pixels of a different color or intensity would be displayed, and so on. Of course, pixels of different brightness, but of the same color, could be handled by the modulation technique of the SLM within the time slots discussed above.

Consequences other than those discussed above are possible. For example, blue light sources (such as laser 220 of Figure 4) tend to be more expensive than red or green for the same light output. It would then be better to stick to a pattern such as red, blue, green, blue, red, etc.

Another embodiment of the invention, in which color is added, is to optically combine three projectors such as those shown in Figures 1a and 2a to produce a single image. Each of these projectors would respond to only one of the primary colors.

Figures 6 and 7 show another example of a useful projection optics 310. A light source 312 generates a beam of substantially parallel light along a light path 314. A spatial light modulator 316 is disposed along the light path 314. As with the other spatial light modulators, some of the light is directed along a light path 318 to the display 320 to form an image to produce a beam of light, while the remaining light is not. In Figure 7, this remaining light is directed along the light path 322 and is intercepted by a plate 324, not reaching the display 320. A serrated lens 326, which may be made from a single piece of molding compound, is disposed along the light path 318. The light is reflected parallel along the light path 328 as received light, but the image is magnified in the vertical direction as shown in Figure 7. Note the differences in width between the dashed lines of the light paths 318 and 328. The light reflected along the light beam 328 is directed to a serrated lens 330. The light reflected from the lens 330 to the display 320 is still parallel. However, it is magnified in the horizontal direction as shown in Figure 7. Lens 330 is disposed behind display 320 and only a portion is shown in Figure 7, which is better illustrated in Figure 6. In Figure 6, projection system 310 of Figure 7 has been replaced with a planar projector 332. Lens 326 magnifies light from light path 318 in the X direction and lens 330 magnifies light from light path 326 in the Y direction. Light is reflected from lens 330 in the Z direction to display 320. Lenses 326 and 330 are serrated or stepped as best seen in Figure 6 from the side view of lens 326. Each step has a reflective surface, such as surface 340, and a non-reflective surface, such as 342. The reflective surfaces may be straight (as shown) or optionally curved. A curved reflecting surface will spread out or magnify an incident beam of light. If the curved design is used, it may not be necessary to break a beam into several smaller beams, since the desired magnification still occurs. The lenses 326 and 330 are therefore narrow at one end (the lower end, as shown in Figure 7) and thicker at the other end (the upper end, as shown in Figure 7).

When slit light is used in the display system shown in Figures 6 and 7, the magnification actually results in several spaced smaller beams of light striking the display surface. In the preferred embodiment, the display surface is a highly scattering surface (such as frosted glass) which serves to blur the beams to form a larger pixel. In an alternative embodiment, the display surface may be coated with dyes as discussed above which fluoresce when excited by the incident light.

As shown in Figure 8, a spatial light modulator 410, which is a deformable mirror array 412, is constructed on a single substrate. Details of the construction of deformable mirror arrays are shown in the patents and patent application incorporated above. The timing control for the array 412 may be located at one or more locations, such as in the timing circuit 412. A rectangular array 416 of deformable mirror cells is shown. The cells may be those shown in Figures 2b and 2c, arranged in rows and columns. The array shown contains 1280 by 840 cells, but it may be of any shape, such as rectangular or circular, and of any suitable size, such as 320 by 200 cells. A register 418 is disposed between the timing circuit 414 and the array 416 along the top (as shown in Figure 8) of the array 416. The register 418 may be a shift register and may consist of several different registers. The register has a number of taps 420, for example 10 or 100, as required to ensure that the register can be loaded at the required speed. The taps are connected to a bus (such as bus 170 in Figure 3). The timing circuit may provide most of the address signals required to load the information into the Arrangement 416, or this can be done via the bus.

A decoder 422 is provided along another side of the array (the left side as shown in Figure 8). The decoder 424 provides the necessary control signals to select the appropriate row in the array 416 in response to an address for the information in the register 418 to be loaded. The decoder 424 is provided along another edge of the array 416 of Figure 8 at the bottom. The decoder 424 provides the necessary control signals to select one of several memory cells associated with all or at least the deformable mirror cells in the array 416. A counter or switch in either the timing circuit 414 or the decoder 424 provides the address of the correct memory cell to be selected. For example, a pixel code loaded into memory cell 1 of 3 memory cells associated with each deformable mirror cell may be displayed by the selected memory cell 1 of all or at least most of the deformable mirror cells, for example, memory cell 426 of Figure 9. It should be noted that decoder 422 must not only determine which row of deformable mirror cells to select, but must also determine which of the mirror cells associated with those deformable memory cells to select.

As shown in Figure 9, row select lines 428, 430, 431 and 432 from decoder 422 are connected to the gates of NMOS access transistors 436, 437 and 438, respectively. Data lines 440, 446, 447 and 448 from register 418 are connected to one side of the source-drain paths of transistors 436, 437 and 438, respectively. The other side of the source-drain paths of transistors 436, 437 and 438 are connected to the inputs of CMOS inverters 454, 455 and 456, respectively. The inverters and access transistors may be replaced by conventional DRAM cells or SRAM cells. Only the inverter 454 will be discussed in detail. The inverter 454 comprises a PMOS transistor 460 and an NMOS transistor 462. The gates of the transistors 460 and 462 are connected to the input of the inverter 454. One end of the source-drain paths of the transistors 460 and 462 is connected to the output of the inverter 454. The other end of the source-drain path of the transistor 460 is connected to the supply voltage (Vcc) while the other end of the source-drain path of the transistor 462 is connected to ground.

The outputs of inverters 454, 455 and 456 are connected to one end of the source-drain paths of transistors 468, 469 and 470, respectively. The other sides of the source-drain paths of transistors 468, 469 and 470 are connected to node 472. Node 472 is connected to the input of inverter 474. Inverter 474 has a PMOS transistor 478 and an NMOS transistor 480. The gates of transistors 478 and 480 are connected to the input of inverter 474 and node 472. One side of the source-drain path of transistors 478 and 480 is connected to the output of inverter 474, which is the input of the deformable mirror cell, such as control electrode 124 (Figure 3). The gates of transistors 468, 469 and 470 are connected to memory cell select lines 484, 485 and 486, respectively, which are 3 of the select lines 488 from decoder 424. It is seen that decoder 424 provides the read function of the memory cells and decoder 422 provides the write function. It is convenient to provide at least 2 memory cells per deformable memory cell and while 3 memory cells are shown in Figure 9, any number required may be provided.

In operation, the select line 436 is driven high and the transistor 436 is turned on so that a high level ('1') or a low level ('0') is present at the input of the inverter 454, supplied by the line 446. can be stored. The select line 436 goes low and the transistor 436 is turned off, storing the applied voltage at the gates of the transistors 460 and 462. When the signal is high, the transistor 462 is on, ground is applied to the output of the inverter 454 and the transistor 460 is turned off. When the signal is low, the transistor 460 is on, the supply voltage is applied to the output of the inverter 454 and the transistor 462 is turned off. Thereafter, if the pixel of information stored in the inverter 454 is to be displayed, the select line 484 is brought high and the inverse value of the '0' or '1' is stored. The inverter 474 inverts its input and the '1' or '0' from the output of the inverter 474 is applied to the deformable mirror cell.

While information is being displayed on inverter 454, one or both of inverters 455 and 456 may be loaded with information. Furthermore, inverter 454 may be loaded while information is being displayed on any of inverters 455 and 456. It should be noted that lines 446, 447 and 448 may be made as one line connected to transistors 436, 437 and 438. It should be noted that the circuits of Figure 9 may be implemented in NMOS, PMOS, CMOS, GaAs, bipolar, CCD or any other suitable technology. Thus, a preferred embodiment of a display cell 490 is shown in Figure 9 with its inverters, access transistors and selection transistors.

Figure 10 shows a flow chart for the operation of the CPU 154 of Figure 3. The flow chart begins at step 510 and proceeds via line 512 to enter logic step 514. In step 514, the information in memory 150 is checked to determine which broadcast standard was received, for example, HDTV, NTSC, PAL, etc., or to determine what computer display information was received over bus 152 from, for example, a graphics adapter, an enhanced color graphics adapter, or a video graphics device. This may be done by locating the vertical and horizontal synchronization pulses, as well as the color and brightness components of the information. The logic then advances along line 518 and enters logic state 520. In logic state 520, a test is made to determine if the standard has been determined. If the standard has not been determined, the logic advances along line 522 and re-enters step 514. If the standard has been determined, the logic exits state 520 and advances along line 524 to logic step 526. In step 526, the logic locates the information within memory 150 for a frame, and preferably the start of a frame.

Once the standard is known, the data can be checked for the vertical sync pulse, for example. The logic then exits step 526 via line 528 and enters logic state 530. If there is an image, the logic exits state 530 via line 532 and enters step 534. In step 534, the image is stored in the appropriate format in memory 160. The logic then advances from step 524 via line 538 and re-enters step 526. If there is no image, the logic exits state 530 and re-enters step 526 via line 540. Under certain conditions, such as a channel change, the logic exits state 520 via line 543 and enters step 544. Step 544 could cause the logic to re-enter step 510, or it could perform some other operation such as displaying the new channel number.

Furthermore, a switch could be provided to allow the user to manually set the standard. Also, the standards for many channels could be stored in the EPROM, which is read by the CPU 154 using the channel information provided by the signal source 140. Therefore, the switch or EPROM would be checked in step 514 and the standard would be determined in state 520.

In Figure 11, an actual three-dimensional digitized video system 610 is shown. The system 610 includes a display 612 containing at least one multi-frequency responsive material and two spatial modulators 616 and 617. One or more energy beams are provided along each light path 620 and 621 impinging on the modulator 616 and 617, respectively. These beams are preferably substantially parallel and beam expanders (not shown) expand the beams to intersect the lens systems 628 and 629, respectively. The lens systems 628 and 629 redirect the beams into substantially parallel beams into the display 612. A depicted hand pointer 635 enables interaction between the user and the computer controlling the display, for example, to magnify a selected area, to position or move an object in the display 612, to create a selected point for rotating the image in the display, or to perform any of the tasks provided in connection with computer-aided design systems.

The modulators 616 and 617 may be any of the various ones described herein. However, the modulator 617 may be of a different type in that it provides an output beam along a horizontal line (as shown in Figure 11), for example line 638. The modulator may also be a horizontal scanning beam. The modulator 617 is controlled either internally or externally to provide the lines in a known sequence synchronized with the information loaded into the modulator 616. This known sequence may include all odd and then all even, or all even and then all odd, from top to bottom or bottom to top. Thus, the line of energy provided by the modulator 617 defines a plane within the display 612, for example, plane 640. The voxels to be displayed are those for which an associated cell in the modulator 616 directs an individual portion of the incident energy beam along the light path 624. Thus, one plane of the image 650 is displayed at a time.

If the display 612 is monochromatic, two beams of energy of different frequencies may need to be provided along the light paths 620 and 621. If color is to be provided, several beams of different frequencies should be provided in a sequence. For example, beams along the light paths 620 and 621 could cause a multi-frequency responsive material within the display 216 to glow red in the plane where both beams are present. These single or multi-frequency beams could cause a multi-frequency responsive material to glow blue, and other beams could cause a multi-frequency responsive material to glow green. Then the next plane in the sequence would be provided with red, green, and blue information. Of course, all planes of the display 612 could be provided with red information followed by green and blue information if desired. If intensity information is provided, several different intervals for displaying red-green-blue information could be provided for each of these planes, as needed. Additional memory cells could be provided in addition to cell 426 to store the additional intensity information. Thus, image 650 is generated one plane at a time to display a complete three-dimensional image. System 610 of Figure 16 can be modified to provide all color information simultaneously. Three spatial modulators can be provided with the red, green and blue information respectively. A suitable optical system would be provided to direct all three beams along the light path 624. The appropriate additional energy beams would also have to be provided along the light path 625. This would allow a complete plane with all color information to be displayed simultaneously.

Although two-dimensional television signals could be displayed in the display 612, the digitized video system of Figure 11 could be further modified in at least two ways to provide a two-dimensional display. First, the display could be made thinner in the horizontal direction to form a thin plate opposite the lens system 629. The modulator 616 would be similarly thin, leaving one or more rows of cells. If different rows are provided, some or all of the rows could provide redundancy if some of the cells in a row fail. If different rows are used, the intensity could also be changed according to the number of cells in the column being displayed. For example, if four bits of intensity information were provided, if maximum intensity was desired, the four cells in that column could be all on, and all four voxels would be displayed at the appropriate time. The modulator 617 would operate as explained above. Second, the display could be made thinner in the vertical direction to provide only a thin region closest to the lens 628. The modulator 617 could be replaced by an energy source that provides a plane of energy across the display. The entire two-dimensional image is then displayed simultaneously.

The beam expansion units and the lens system of Figure 11 could be replaced by the toothed mirrors of Figures 6 and 7. The sensors 655-660 are intended to detect signals originating from or generated by the pointer 655. radiation. More or fewer sensors may be provided as needed. In one embodiment, the pointer produces a beam of energy at a frequency which interacts with another beam of energy in the display. This interaction is sensed by sensors to determine the line of the pointer above the display 612. In another embodiment, the emission of the pointer is directly sensed by sensors 655-660 to determine the line along which the pointer 635 points into the display 612. The system 610 of Figure 16 is useful for a hostile environment such as aircraft or tanks. The display 610 could be solid or gaseous, and could be provided with a three-dimensional housing. The display could be of any suitable shape, for example spherical. The display 610 may include a coating on both its interior and exterior surfaces between the user and the display that is transparent to visible light generated in the display and absorbs or reflects the energy rays emitted to produce the luminescences. The coating and multi-frequency responsive material may be used as the display 320 shown in Figures 6 and 7.

One type of pointer suitable for use with the digitized video system 610 of Figure 11 is shown in Figures 12 and 13. The pointer shown in Figures 11 and 12 has a plurality of buttons 712-714 which can be used to provide an encoded input to the display 612. The front of the pointer 710 is triangular in shape and is provided with beam generators 718-720 which can be IR devices. The output from the generators 718-720 can be sensed by sensors 655-660 to determine the line along which the pointer is directed into the display 612. Buttons 712-715 can provide information which is then encoded for transmission by the generators 718-720, for example by tuning the generators in different sequences or at different frequencies or both when a particular button is pressed. Buttons 712 and 713 may indicate that the distance of interest is greater or smaller in the display, respectively. Button 714 could indicate that a drag is to occur, and button 715 could lock or release the cursor. Handles 725 could be provided. The cursor provided in the display could be the entire line of the pointer across the display with a highlighted dot to indicate the distance. The cursor could be a visible XYZ axis with an axis along the line of the pointer or a visible XYZ axis aligned with the display.

In Figure 14, another pointer 730 is shown which is rotatable about a pivot 732. Handles 734 may be provided. Knobs 736 similar to knobs 712-715 are provided. Beam generators 739 and 740 are provided at the ends of segments 743 and 744. Segments 743 and 744 are pivotable about pivot 732, allowing the point at which beams 446 and 447 coincide to be moved. Thus, beams 466 and 467 may be of different frequencies, causing multi-frequency responsive material to glow, which is sensed by sensors 656-660.

Another pointer 750 is shown in Figure 15. This pointer 750 has an actuating lever 752 and an elongated body 753. Buttons 754 and 755, respectively, comparable to buttons 714 and 715 are provided. Handles 758 may be provided. The actuating lever 752 changes the angle of the beams 760 and 761, thereby changing the distance at which the beams coincide. For example, the closer the actuating lever is pulled to the body 753, the closer the beam intersection is moved to the body. The beams are generated by beam generators, with only beam generator 763 being shown in Figure 15. The beam generators may be similar to the beam generators 739 and 740 of Figure 14.

Figure 16 shows a multi-dimensional array processor (MAP) 808 coupled to a host processor 800 and a three-dimensional display 801. The host 800 may include any type of computer system such as a personal computer, a minicomputer, or even a broadcast information receiver system. The 3D display 801 may be of the type described above, which is described below in connection with Figure 17, or of another suitable display type.

In the presently preferred embodiment of the invention, the display 801 displays the 3D image as a number N of XY planes, thus creating an XYZ image. Of course, other embodiments are possible without departing from the scope of the invention. For example, the display may include a number of XYZ blocks with one (or more) MAPs per block. Alternatively, each processing cell (e.g., 810, 820, and 830) of the MAP may operate on an XYZ basis rather than the currently preferred plane basis.

The MAP 808 is generally comprised of a plurality of processing cells (PCs) 810, 820, and 830. As indicated, in the presently preferred embodiment, each processing cell operates on an XY plane of the N-plane image. For illustration purposes, only three processing cells 810, 820, and 830 are shown. PC 810 is for the first plane, PC 820 is for the second plane, and PC 830 is for the Nth plane of the image. In other embodiments, each PC may be responsible for different planes of the image.

Each processing cell 810, 820 and 830 includes a processor (811, 821 or 831) coupled to a level memory (812, 822 or 832) via a bus (813, 823 or 833). The processors 811, 821 and 831 may be of different types. For example: digital signal processors (DSPs) such as the TMS 320C30; conventional microprocessors such as the TMS99000; or graphics signal processors (GSPs) such as the TMS34020, etc. may be used. In the presently preferred embodiment, GSPs are used because they are not only programmable, but are also designed to manipulate two-dimensional (XY>) data. Likewise, the memory used may be of various types. Video RAMs (such as the TMS4461), DRAMs, SRAMs, or other memories may be used. Preferably, the memory architecture is of a two-port type. In the present embodiment, VRAMs are used because they are inherently two-port, with the conventional memory interface used for the processor buses 813, 823, and 833, and the high-speed serial interfaces used to interface with the display bus 803.

The host 800 communicates with each processor 811, 821 and 831 simultaneously in the MAP via the high level descriptor language (HDL) bus 809 and the control bus 804. The control bus 804 is used to signal when the host 800 can send another HDL command (as described below) and to synchronize the PCs with the display 801. Optionally, each processor can communicate with its neighbors via other buses. For example: the processor 811 and the processor 821 communicate via the bus 805; the processor 821 communicates with the level 3 processor (not shown) via the bus 806; and the N-1 processor (not shown) communicates with processor 831 via bus 807. As can be seen, the communication provided between the processors is useful in some manipulations of a three-dimensional image.

In an alternative embodiment (not shown), the host 800 would only communicate with the processor 811. All instructions to other processors (such as 821 and 831) are allowed to "ripple" over the inter-processor buses (805, 806 and 807).

The display bus 803 is used to provide display data to the 3D display 801. In the currently preferred embodiment, each PC (810, 820, and 830) can sequentially put an entire level of display data on the bus 803 before the next level. For example: the PC 810 puts the first level of image data on the bus 803; next, the PC 820 puts the second level of image data on the bus 803; each subsequent PC (not shown) also puts its data on the bus 803; finally, the PC 830 puts its level on the bus 803. The process then begins again. Of course, other sequences can be used. In other embodiments, each PC can be responsible for a block of XYZ information and put the entire block on the bus 803, or interleave levels as required by the display 801. In any event, all of the image data is present on the bus 803 at some point in time. Furthermore, since the display 801 must be refreshed frequently (preferably faster than the critical flicker frequency), the image data is available for a relatively short period of time.

As can be seen below, each PC must have access to the information that is present in other layers. If this were to be done via the interprocessor buses 805, 806 and 807, it would be relatively slow, especially if the information must be supplied from different layers. Therefore, each processor (811, 821, 831) is also connected to the display bus 803 via buses 816, 826 and 836, respectively. In the currently preferred embodiment, the buses 816, 826 and 836 are coupled to the processor buses 813, 823 and 833, respectively. In another embodiment, each processor may have an additional port coupled to the display bus 803. In another embodiment, the plane memories 812, 822, and 833 may be used to buffer data from the display bus 803.

The High-Level Descriptor Language (HDL) used by the Host 800 to control the MAP PCs can be divided into three basic types. These include structural formation types (which perform arithmetic and logical operations on a predetermined set of points), fill types, and translation types (including rotations). These are all executed in parallel, as will be apparent from the following description.

A structure formation may include types such as line drawing, fractal generation, logic operations such as XORs, frame drawing, etc. For example, if a line is to be drawn for display on display 801, the following sequence occurs. Host 800 issues an HDL command specifying the instruction type (i.e., line drawing) and a formula specifying the parameters of the line. Each PC of MAP 808 simultaneously receives this command over HDL bus 809. All parallel PCs evaluate the formula to decide if the level for which the PC is responsible is affected. If so, the PC holds a common control line of control bus 804 low until it terminates. This prevents host 800 from sending another command until all processing is complete. Any PC that needs to change a point in its assigned level memory will do so and then release the control line. When all PCs have finished, the control line can go high and the Host 800 can issue another HDL command.

In the preferred embodiment, the control line is treated slightly differently. All PCs hold the control line immediately after receiving a command, and they release it when the PC is either finished or determines that the PC is not affected by the operation.

Fill instructions begin in the same way. The host 800 issues a fill HDL command to all PCs over the bus 809. All PCs that immediately determine that they need to change their respective levels hold the control line of the control bus 804 at a low level. In this case, however, since this is a three-dimensional fill, as the fill progresses in one layer, "holes" or "leaks" may be opened in another layer, forcing a PC that previously did not need to change its layer to do the same. There are two currently preferred ways to inform the previously unaffected PCs that they need to change their layer memories.

The first uses the interprocessor buses 805, 806 and 807. When a given PC determines that it needs to change its allocated level memory, it informs the neighboring PCs via the allocated interprocessor bus that it is doing so. Then the neighboring PCs also hold the control line low. When the given PC is finished, it informs the neighboring PCs of this fact and releases the control line. The neighboring PCs hold the control line low and they then check to see if they need to change their level memory. To determine this, they will either check the given PC's level data as it passes through the display bus 803 or they can request data directly from the given PC via the interprocessor bus. If the neighbors determine that they have no changes, they release the control line. If, however, changes are necessary, they inform their respective neighbors and the process described is repeated.

The second technique avoids the need (at least in this case) for interprocessor buses 805, 806 and 807. Each PC only needs to be able to pull down the control line of control bus 804, but it must also be able to read the status of the control line. This is done as follows. If a given PC determines from the HDL instruction (or, as will be seen, from data passing over bus 803), that it must change its level memory, it will pull the control line low and change its memory. When it is finished, it must keep the control line low for at least one full cycle of all image data passing over bus 803. This allows the given PC to determine if it is really finished. In other words, if another level has changed, the given PC may also have a change. If not, the given PC releases the line. At any time the line is held low after an HDL fill command, all PCs must continually sample their neighbors' level data passing over bus 803 to determine if they need to begin the change. If so, they hold the line low and repeat the above process.

The last class of HDL instruction is that of translations and rotations. The instructions start in the same way. The host 800 sends the instruction and all affected PCs hold the control line. In the case of a mere XY translation or rotation (i.e. no translations or rotations between planes), all processors can simultaneously execute the translation in their assigned plane and release the control line.

In the event of any translation or rotation affecting the Z axis (i.e., requiring an interplane translation), all affected PCs hold the control line for at least one full frame cycle of data traveling over the display bus 803. This is done so that the associated processors can buffer all the interplane data they require. After the full frame cycle, the PC's processors can change their associated plane memory and then stop holding the control line. No data must be changed before a full frame cycle occurs, so as not to affect data required by another PC.

As can be seen from the above discussion, the architecture for a processor for a true three-dimensional display is quite different from that required for a processor designed to display a three-dimensional image on a two-dimensional display. Computational power is not so critical, but rather bus data flow is more critical. Consequently, the processors used in the PC can be simpler and smaller, making it easier to place them on a single chip with the plane memory. In fact, for some true 3D displays, the entire MAP can be implemented on a single substrate.

One type of three-dimensional display is shown in post-issued U.S. Patent No. 5,042,909 to Garcia and Williams, which is incorporated herein by reference. Referring to Figure 17, a surface having a 360° helix 900 is rotated about an axis 902 to create a three-dimensional cylindrical space 904. The surface 900 extends from the axis 902 to the other corners of the space 904 as shown in Figure 17. As the surface 900 is rotated, each point in the space 904 is intersected once during each rotation. The surface 900 may have any suitable shape, for example, a circular disk. A beam of light is projected along the light path 906 such that it intersects the surface 900. The image information may be provided by any of the systems shown here that utilize a spatial light modulator along the light path 906.

An advantage of the invention is to provide a digitized video display at an acceptable cost and with high reliability.

Another advantage of the invention is that it provides a digitized video display that can be easily modified and/or manufactured for various different, incompatible television and video systems.

An advantage of the invention is the provision of a digitized video display with reduced speckle.

An advantage of the invention is the provision of a digitized video display that provides improved picture quality.

While the invention has been described in connection with certain specific embodiments, it is to be recognized that other modifications will be apparent to those skilled in the art, and it is intended that all such modifications be covered by the appended claims.

Claims (14)

1. A projection light image display system with reduced spotting, comprising: a coherent light source (61) for generating a light beam, a screen (51), a spatial light modulator (56) responsive to signals representing an image to be displayed for directing light from the source to the screen to display the image on the screen, characterized in that a transducer (90) is coupled to the screen, which serves to generate acoustic surface waves in the screen where the image is displayed. 2. Projection system according to claim 1, wherein the acoustic surface waves have an amplitude that is greater than a wavelength of the light beam. 3. Projection system according to claim 1 or 2, wherein the surface acoustic waves are substantially parallel. 4. Projection system according to one of the preceding claims, in which the spatial light modulator (56) generates several light beams which impinge on the screen simultaneously at different angles and with different path lengths. 5. Projection system according to one of the preceding claims, in which the screen (51) has a diffusing surface to increase the viewing angle of the image displayed thereon. 6. Projection system according to one of the preceding claims, in which the light directed by the spatial light modulator (36) onto the screen (51) strikes the front surface of this screen. 7. Projection system according to one of claims 1 to 6, in which the light directed by the spatial light modulator (36) onto the screen (51) strikes the rear surface of this screen. 8. A projection system according to any preceding claim, wherein the spatial light modulator (56) includes a two-dimensional array of mirrors selectively offset to reflect a pattern of light rays onto the screen (51) to cause the image to be displayed thereon. 9. A method of reducing speckle in a projection light image display system in which light from a coherent light source is directed through a spatial light modulator onto a screen to display an image on the screen, characterized in that speckle in the displayed image is reduced by generating surface acoustic waves on the screen. 10. The method of claim 9, wherein the surface acoustic waves are substantially parallel. 11. Method according to claim 9 or 10, wherein the acoustic surface waves have an amplitude that is greater than a wavelength of the light beam. 12. Method according to one of claims 9 to 11, in which the light directed onto the screen by the spatial light modulator is in the form of several beams which impinge on the screen simultaneously at different angles and with different path lengths. 13. A method according to any one of claims 9 to 12, wherein the light from the spatial light modulator is directed onto the front surface of the screen. 14. A method according to any one of claims 9 to 12, wherein the light from the spatial light modulator is directed onto the rear surface of the screen.