Rotating phase surface for improved image quality

Abstract

A (phase-delay) kinoform (2200), arranged for rotation (centred at 2210), divided into a plurality of concentric zones (2220, 2230), each concentric zone sub-divided into a plurality of phase-delay ar

Description

FIELD The present disclosure relates to a holographic projector and method of holographic projection. More specifically, the present disclosure relates a holographic projector and method for processing a holographic reconstruction to remove I reduce the perception of speckle. Even more specifically, the present disclosure relates to removing or reducing speckle caused by the interference between pixels of high resolution images by rapidly applying varying phase-delays to pixels of a holographic reconstruction. The present disclosure also relates to a kinoform or phase surface and a hologram or image pixel structure. Some embodiments relate to a holographic projector, picture generating unit or head-up display. BACKGROUND AND INTRODUCTION Light scattered from an object contains both amplitude and phase information. This amplitude and phase information can be captured on, for example, a photosensitive plate by well-known interference techniques to form a holographic recording, or “hologram”, comprising interference fringes. The hologram may be reconstructed by illumination with suitable light to form a two-dimensional or three-dimensional holographic reconstruction, or replay image, representative of the original object. Computer-generated holography may numerically simulate the interference process. A computer-generated hologram may be calculated by a technique based on a mathematical transformation such as a Fresnel or Fourier transform. These types of holograms may be referred to as Fresnel / Fourier transform holograms or simply Fresnel / Fourier holograms. A Fourier hologram may be considered a Fourier domain / plane representation of the object or a frequency domain / plane representation of the object. A computer-generated hologram may also be calculated by coherent ray tracing or a point cloud technique, for example. A computer-generated hologram may be encoded on a spatial light modulator arranged to modulate the amplitude and / or phase of incident light. Light modulation may be achieved using electrically-addressable liquid crystals, optically-addressable liquid crystals or micromirrors, for example. A spatial light modulator typically comprises a plurality of individually-addressable pixels which may also be referred to as cells or elements. The light modulation scheme may be binary, multilevel or continuous. Alternatively, the device may be continuous (i.e. is not comprised of pixels) and light modulation may therefore be continuous across the device. The spatial light modulator may be reflective meaning that modulated light is output in reflection. The spatial light modulator may equally be transmissive meaning that modulated light is output in transmission. A holographic projector may be provided using the system described herein. Such projectors have found application in head-up displays, “HUD”. SUMMARY Aspects of the present disclosure are defined in the appended independent claims. In general terms, the present disclosure relates to a holographic projector and a method of holographic projection. The hologram is a hologram of a picture. The holographic projector and projection method are arranged to reduce I minimise the perception of speckle (or speckle-like interference patterns) in a holographic reconstruction of the picture. Holographic projectors typically comprise a coherent light source such as a laser. It is well known that the use of coherent light can lead to unwanted interference patterns being formed. That is, the coherent light might interfere with itself. This can create noise (which is not external noise). In the case of holographic projectors, this noise can degrade the image quality of (e.g. virtual) images formed using the holographic projector. One example of this phenomena is referred to as “speckle”. The present disclosure refers to holographic reconstruction by way of example of one method of forming an image using light, such as coherent light, that can exhibit speckle-like effects. A first source of speckle results from diffuse reflections from (imperfect) surfaces. In particular, imperfections of a reflection surface can create fluctuations I subtle phase changes in different portions of an image - such as a holographic reconstruction - reflected by the surface. These phase changes result in spatially varying constructive and I or destructive interference causing light and dark areas in the projected (e.g. virtual) image. A second source of speckle (or, more accurately, speckle-like noise) arises when the images I holographic reconstructions that are formed by the holographic projector have a relatively high resolution - i.e. relatively high density of image pixels or dots per inch or pixels per degree. It is generally desirable for images to have a high density of pixels. For an image not to appear “pixelated”, the points of light I pixels that form that image must be sufficiently close together that they cannot be separated by the eye. However, placing the points of light I image pixels that form the image too close can cause interference between adjacent points I pixels. In particular, cross-talk between light associated with adjacent pixels of the holographic reconstruction causes a speckle-like or grain-like pattern. This interaction I interference may result in a pattern of constructive interference and destructive interference such that there are (unintentional) relatively bright and (unintentional) relatively dark areas in an image of the picture formed by the holographic projector. This pattern may be considered an image pixel crosstalk. However, this pattern is referred to in this disclosure as specklelike noise or speckle for shorthand. For example, fine details of the image, such as text, become harder to view I read. In an image formed of red, green and blue light of specific wavelengths, colour variation from the intended colour of the image may be exhibited because of different speckle patterns produced by the red, green and blue light sources. For example, intended uniform white areas of an image may exhibit colour variation and so may not appear uniformly white. Various effects of this nature occur in a real-world system and are broadly referred to herein as speckle or speckle-like even though it may be debatable by academics and the like whether the effects are truly examples of laser-speckle in accordance with the strict definitions accepted in the field. In other words, the term “speckle” is used broadly herein to refer to undesired optical effects that result in the appearance of a grainy or speckly image. The grain or speckle is considered “noise” in the image. It is known to reduce speckle, or the perception of speckle, by moving the screen or light receiving member - at least in systems that use a screen. The screen is typically diffuse and may be referred to as a “diffuser”. For example, the diffuser may translate back-and-forth in one-dimension or rotate. Rotation is often preferred because it is continuous and does not require a change in direction which, in practice, cannot be instantaneous. By moving, the region of the light receiving member that is illuminated with the image changes overtime, thereby changing the pattern of light of the image that emanates from the light receiving member. This is because statistical imperfections in the light receiving member influence the pattern of speckle so that the pattern changes because a continuously changing set of imperfections is illuminated as the light receiving members moves. As the pattern of speckle changes over time, it is averaged I integrated by the optic system of a human observer, so that the appearance of speckle in the image is reduced. However, there are disadvantages to using a moveable light receiving member (e.g. translating disc diffuser) to reduce speckle. In particular, a holographic projector comprising such a moveable diffuser will be arranged to have good image quality in the holographic reconstruction at a single virtual image distance. But, at other distances, the image quality is significantly degraded. It would be desirable to be able to display virtual images at a plurality of arbitrary depths and to be able to display a plurality of virtual images, each at a different (arbitrary) depth, while also achieving good image quality. This is more complex with the “moveable light receiving member” example. Furthermore, some holographic projectors do not comprise a light receiving member / diffuser. Instead, such holographic projector may be configured such that light that is received by the viewer is spatially modulated in accordance with a hologram (not the picture). For example, a holographic reconstruction of the picture is formed by the lens of the eye performing a hologram-to-image transform. In other words, a holographic wavefront is received by the viewing system (eye) rather than an image wavefront. Such holographic projectors may be referred to as “hologram to eye” projectors. In some embodiments, the holographic projector / projection method comprises an optical system comprising at least a first lens and optionally a second lens. The system is arranged to receive a holographic wavefront at the first lens. An (intermediate) holographic reconstruction of the picture encoded by the hologram is formed downstream of the first lens, optionally between the first and second lens. The wavefront is then received at the second lens (if present) to form a modified holographic wavefront. The holographic projector may be arranged such that the modified holographic wavefront is relayed to an eye-box. A viewing system at the eye-box may perceive a virtual image of the picture from the eye-box. British patent application 2300506.9 filed 13 January 2023 disclosed that rapidly moving a kinoform that is disposed downstream of the first lens (between the first and second lens, if the second lens is present) significantly reduces I minimises the perception of speckle in the holographic reconstruction formed by the viewing system (e.g. eye) even though the introduction of time-varying phase-randomisation on the wavefront might be expected to upset the holographic reconstruction process performed by the viewing system particularly when the hologram is a phase, or even phase-only, hologram. In some embodiments, the kinoform is arranged to apply a different phase-delay to adjacent image pixels I points of light of an intermediate holographic reconstruction between two optical elements such as lenses. By rapidly moving the kinoform, the phase-delay applied to each image pixel I point of light of the intermediate holographic reconstruction is also rapidly changed which, in turn, changes the pattern of speckle. In particular, the pattern of speckle of the holographic reconstruction (perceived from the eye-box) rapidly changes. The kinoform is moved rapidly enough that a plurality of different phase-delays are applied to each image pixel I point of light of the intermediate holographic reconstruction within the integration time of a human eye. So, as the pattern of speckle changes overtime, it is averaged by an optic system (for example, the eye of a human observer). Thus, the appearance I perception of speckle in the holographic reconstruction is reduced. It was found that rapidly changing phase-delay applied to each image pixel I point of light of an intermediate holographic reconstruction, described above, reduces I minimises the perception of speckle (in particular, speckle of the second source, due to high resolution images) without the loss of depth information of the virtual image and at any arbitrary depth I distance of virtual image. In some embodiments (but not all), the processing of the hologram to remove speckle takes place within the optical system at a point that an (intermediate) holographic reconstruction of picture is formed. At this point, it is possible to process the individual pixels I points of light of the holographic reconstruction in order to modify the holographic reconstruction. The holographic reconstruction may be said to reform a holographic wavefront that the second lens relays towards the eye-box. The holographic projector / method of holographic projection may be described as processing the hologram I holographic wavefront (using the moving kinoform) to reduce I minimise speckle. The hologram that is output by the optical system is a processed or modified hologram relative to the hologram that is displayed on the display device I received by the optical system. The processing introduces an apparently (spatially and temporally) random and changing (i.e. time-varying) phase-delay to reduce I minimize the perception of speckle-like noise as described above. The appearance of speckle-like noise may be minimised (i.e. speckle reduction may be optimised) when the kinoform is arranged to apply a different phase-delay to each pixel of the holographic reconstruction in such a way that no two adjacent pixels of the holographic reconstruction have the same phase-delay applied by moving the kinoform. However, to achieve this optimum arrangement, there is a risk that such a kinoform may otherwise adversely affect image quality such as image sharpness. This may be caused by the kinoform (unintentionally) applying different phase delays to different portions of the same pixel or image point of a holographic reconstruction (simultaneously). For example, the kinoform may comprise an array of zones and each zone may be arranged to apply a phasedelay. In the “optimum” arrangement, the kinoform would comprise at least as many zones as there are pixels such that each zone applies a different phase delay to each pixel and each zone would be aligned with its respective pixel such that each pixel interacts with only one zone (when the kinoform is in a particular position with respect to the holographic reconstruction). If a pixel overlaps with I interacts with more than one zone, then more than one phase-delay may be applied to that pixel (with different phase delays being applied to different portions of the pixel). This can reduce image quality I sharpness. The overlap of pixels with multiple zones of the kinoform (and so, pixels having multiple different phase delays applied) may be a particular problem: a) when pixels of the holographic reconstruction are relatively large, b) when the pixels of the holographic reconstruction are close together and I or c) when the pitch of pixels of the holographic reconstruction spatially vary. Each of a) to c) may increase the risk of the pixels interacting with multiple zones of the kinoform, b) and c) may arise when the holographic reconstruction is pre-distorted to compensate fora (complex) optical power of an optical combiner, as described below. The holographic projector may be arranged to relay a holographic wavefront towards an eyebox. In some embodiments, the holographic projector is arranged to relay the holographic wavefront to an eye-box via an optical combiner. For example, if the holographic projector is part of a head-up display system for a vehicle, the optical combiner may be a windscreen or windshield of the vehicle. Such an optical combiner typically has a complex curved shape. The complex curved shape may have a non-uniform optical power. For example, the optical power of the optical combiner may generally be greater in a first direction (e.g. horizontal direction) than in a second direction (e.g. vertical direction). Furthermore, in some embodiments the optical power of the optical combiner may spatially vary along the first and I or second direction. For example, the optical power of the optical combiner may be non-uniform in the first direction. This may mean that light incident on different portions of the optical combiner along the first direction may experience a different optical power. As the skilled reader will appreciate, the non-uniformity of the optical power of the optical combiner may result in distortion of the holographic wavefront as it is relayed from the holographic projector to the eye-box via the optical combiner. Thus, there may be distortion of a (virtual) image that is viewable from the eye-box. In other words, image spots or pixels of the holographic reconstruction may be displaced, translated or otherwise warped relative to their intended position I relative to the picture of the hologram. To compensate for the distortion caused by the optical combiner (and, optionally, distortion caused by other effects I other components of the holographic projector), the holographic reconstruction may be predistorted. More specifically, the hologram that is displayed on the display device may be of a hologram of a picture that has been pre-distorted to compensate for the (non-uniform) optical power of the optical combiner (windscreen I windshield). For example, the picture may be skewed or warped to compensate for the non-uniformities of the optical combiner. This may result in the pixel or image points of the picture being distorted to be closer together (i.e. the pitch of the pixels or image points of the picture may be decreased). Furthermore, to compensate fora changing optical power of the optical combiner, the distorted pixels or image points of the picture may have a non-uniform pitch. Because the optical power of the optical combiner may be non-uniform, and because the variance of the optical power of the optical combiner may be greater along the first direction, the pitch of pixels or image points may be changed I reduced to a greater extent along the first direction than along the second direction. The distortions in the picture may also be present in the pixels of the holographic reconstruction of the picture upstream of the optical combiner. The kinoform may be positioned at this holographic reconstruction and so the kinoform may process a distorted holographic reconstruction (in which pixels of the holographic reconstruction are distorted I repositioned / warped to pre-compensate for the (complex) optical power of the optical combiner). The reduced and varying pitch of the (pre-)distorted holographic reconstruction increases the risk that different phase-delays are applied to a single pixel of the holographic reconstruction simultaneously, particularly if the array of zones of the kinoform are uniform (in terms of spacing and size) but the pixels of the holographic reconstruction are not. One option to resolve this could be to provide a kinoform in which the zones are sized and distributed in a way that corresponds to the distribution of the pixels of the (pre-)distorted hologram. However, this solution is not optimum or practical for a number of reasons. Firstly, it would be prohibitively expensive and complex manufacturing such a non-uniform kinoform. Secondly, a custom kinoform would need to be manufactured for each specific arrangement of holographic projector and optical combiner. For example, in the context of holographic projectors for vehicles, a different kinoform would need to be provided for each different model of vehicle (e.g. having a different windscreen). Thirdly, the holographic projector may be arranged such that amount of pre-distortion of the holographic reconstruction is different depending on, for example, environment conditions such as temperature. So, it may not be realistic or practical to provide a single non-uniform kinoform that has a distribution of zones the avoids overlap of pixels with multiple zones for all environmental conditions. Fourthly, the pixels of the holographic reconstruction may overlap at the kinoform. This may be a result of one or both of: the amount of (pre-)distortion and the pixels being relatively large. This may make it impossible to avoid overlap of pixels with multiple zones if the kinoform is arranged to apply phase delay such that there is a difference in the phase delay applied to all adjacent pixels, in all directions. In some examples, the appearance of speckle is acceptably reduced and a good image quality retained if a kinoform is provided that is arranged to apply a different phase-delay to adjacent lines of pixels of the holographic reconstruction extending in a first direction, and if the kinoform is moved in a second direction that is different to the first direction (optionally, perpendicular to the first direction). As such a kinoform is moved in the second direction, the phase that is applied to each pixel of the holographic reconstruction constantly changes and the difference between the phase delay applied to adjacent pixels in the second direction constantly changes (albeit with adjacent pixels in the lines extending in the first direction always having the same simultaneous phase delay applied). As described previously, this repeatedly and rapidly changes the speckle pattern and so reduces I minimises the perception of speckle. Despite the phase difference between pixels only changing along the second direction as the kinoform is moved in this arrangement, the appearance of speckle reduction is still significantly reduced. This arrangement removes constraints on the design of the kinoform. In particular, there is no longer a need for the kinoform to comprise zones that perfectly align with each individual pixel of the holographic reconstruction. Instead, a single zone can be provided for each line of pixels in the first direction. The same phase delay will be applied to each pixel in the first line and the risk of different phase delays being simultaneously applied to different portions of a pixel is substantially reduced I eliminated (at least in the first direction), even if the pixels overlap (in the first direction). The first direction can be selected as a direction along which (pre-)distortion of the holographic reconstruction is greatest (to compensate for an optical combiner having a relatively larger optical power in the first direction and / or a varying optical power in the first direction). The holographic projector comprises a display device arranged to form a holographic wavefront by spatially modulating light in accordance with a hologram of a picture displayed thereon. The display device may be a pixelated display device. The display device may be a spatial light modulator such as a liquid crystal on silicon spatial light modulator. The spatial light modulator may be a pixelated spatial light modulator. In some embodiments, the holographic projector comprises an optical (e.g. magnification) system, or optical relay, comprising a first lens. The first lens may be arranged to receive spatially modulated light from the display device. The first lens is arranged to form an intermediate holographic reconstruction of the picture by focusing the holographic wavefront towards I substantially at a focal plane of the first lens. Said focal plane of the first lens may be referred to as a first or front plane of the first lens. The intermediate holographic reconstruction of the picture comprises a plurality of pixels or points of light. In some embodiments described in detail herein by way of example only, the holographic projector further comprises a kinoform disposed at the intermediate image plane between a first lens and second lens. This arrangement may be preferred for “hologram-to-eye”. In other embodiments, the kinoform is provided elsewhere in the display system such as at the picture plane of a picture generating unit, “PGU”, of a head-up display. In these arrangements, the picture plane is positioned before the mirrors of a conventional head-up display or before the first waveguide of a compact head-up display. The kinoform of the present disclosure may be positioned anywhere in the display system where an image (or picture) is formed. The kinoform is arranged to apply a phase-delay to each pixel of the image. A movement assembly is arranged to move the kinoform such that a plurality of different phase-delays are applied to each pixel of the image e.g. within the integration time of the human eye. In some embodiments, the kinoform is arranged such that the phase-delay applied to each pixel is different to the phase-delay of the (immediately) adjacent or connecting pixel. In an example disclosed herein, the kinoform is arranged to apply a first phase-delay to each pixel of a first line of pixels of a holographic reconstruction and to apply a second phase delay to each pixel of a second line of pixels of a holographic reconstruction. The first phase delay may be different to the second phase delay. The first and second lines of pixels extend in a first direction. In other words, the first and second lines of pixels are parallel to one another and parallel to the first direction. The first line may be adjacent to the second line. The kinoform may apply the same (first) phase-delay to each pixel of the first line of pixels. In other words, the kinoform may apply a uniform phase-delay to each pixel of the first line. The kinoform may apply the same (second) phase-delay to each pixel of the second line of pixels. In other words, the kinoform may apply a uniform phase-delay to each pixel of the second line. Thus, adjacent pixels of the holographic reconstruction in the first direction may have the same phase-delay applied. However, adjacent pixels of the holographic reconstruction in another direction (for example, the second direction described below) will be different. For example, adjacent pairs of pixels comprising one pixel of the first line and one pixel of the second line will have different phase-delays applied to the different pixels. Thus, it may still be accurate to describe this kinoform as applying different phasedelays to adjacent pixels. For convenience, this example of the kinoform may be referred to as a “striped” kinoform herein. This may be because the kinoform may comprise a plurality of zones, each zone extending in the first direction and each zone being arranged to apply a different phase-delay to adjacent lines of pixels of the holographic reconstruction. The zones may appear as stripes extending in the first direction. As described above, the advantage of providing a movement assembly arranged to move a kinoform (arranged to apply a phase-delay to the pixels of the holographic reconstruction such that a different phase-delay is applied to adjacent to pixels) is to reduce I minimize a speckle effect (or at least the perception of speckle effect). The kinoform (in a first position) is arranged to apply a phase-delay to each image spot or pixel of the (intermediate) holographic reconstruction such that the phase-delay applied to each image pixel is different to the phase-delay of an (immediately) adjacent or connecting. As above, adjacent pixels I points of light interact I interfere with each other to form the speckle pattern. By applying different phase delays to adjacent pixels I points of light of the holographic reconstruction, the interaction I interference is changed and so the speckle pattern is changed. By moving the kinoform (in particular, moving the kinoform with respect to the first lens / the holographic reconstruction), different pixels of the holographic reconstruction interact with different portions of the kinoform and so experience different phase-delays. Importantly, the relative phase-delay between adjacent pixels I points of light of the (intermediate) holographic reconstruction changes as the kinoform is moved. Each time the relative phase-delay changes, the speckle pattern changes. The kinoform is moved rapidly enough that a plurality of different phase-delays are applied to each pixel of the hologram reconstruction within the integration time of the human eye. So a plurality of different speckle patterns are received by a human observer. The (continuously changing) speckle pattern, is averaged by the optic system of a human observer, so that the appearance of speckle in the holographic reconstruction is reduced I minimised. In embodiments in which the kinoform is disposed between two lens, the first lens and second lens may be arranged to form a magnifying or demagnifying telescope. In some embodiments, the front focal plane of the first lens is substantially aligned / coplanar with the back focal plane of the second lens. The first lens may have a first focal length. The second lens may have a second focal length. A separation between the first lens and the second lens may be substantially equal to the sum of the first and second focal lengths. The first lens may comprise a first optical axis. The second lens may comprise a second optical axis. In some embodiments, the first optical axis may be aligned and I or parallel and I or colinear with the second optical axis. In such embodiments, the separation between the pair of lenses may be a separation I distance (measured) parallel to the first and I or second optical axis. In some embodiments, the first optical axis may be angled with respect to the second optical axis. For example, the angle between the first optical axis and the second optical axis may be between 70 and 110 degrees, optionally substantially 90 degrees. In some embodiments, the optical relay may be referred to as a 4f system. The first lens and / or the second lens may both be converging lenses. As used herein, the focal length of the first and / or second lens is the distance from the centre of the respective lens to the principal foci of the lens. Collimated light will be focused at this single point by the first I second lens (as a converging lens). This assumes that the first and I or second lenses (respectively) are thin lenses in air. As used herein, the focal plane of a lens (e.g. the first or second lens) is a plane that is perpendicular to the optical axis of the lens. The distance from the centre of the respective lens to the focal plane is equal to the focal length. In some embodiments, the holographic projector further comprises a waveguide which may be referred to as a waveguide pupil expander or pupil replicator or hologram replicator. The waveguide may comprise a pair of opposing surfaces which may be arranged to waveguide light therebetween. The light that the waveguide is arranged to waveguide may be received directly or indirectly from the second lens. A first surface of the pair of opposing surfaces may be partially-reflective partially-transmissive such that a plurality of replicas of the (modified I despeckled) holographic wavefront are emitted therefrom. In such embodiments, the aperture / pupil of a display device may limit the size of a viewing window (e.g. eye-box). The waveguide may be used to form an expanded (increased size) viewing window by creating and emitting a plurality of replicas of the pupil. Each waveguide generates a plurality of replicas of a holographic wavefront formed by illuminating a hologram displayed on the display device. In some embodiments, the movement assembly is arranged to move the kinoform such that the kinoform has a first position with respect to the (intermediate) holographic reconstruction and a second position with respect to the (intermediate) holographic reconstruction. The first position may be different to the second position. The kinoform may be substantially planar. A first plane may be defined by the substantially planar kinoform. Light of the intermediate holographic reconstruction may propagate towards the kinoform substantially in a first direction. A normal of the first plane may be substantially parallel to the first direction. The movement assembly may be arranged to move the kinoform (rotationally) within the first plane. As used herein, a kinoform is a device that operates only on the phase of an incident wave. The kinoform may comprise a computer-generated pattern to operate on the phase of the incident wave. The computer-generated pattern may be fixed I permanently recorded in the kinoform. For example, in some embodiments, the kinoform may comprise a transmissive material having a refractive index greater than 1 and may operate on the phase of an incident wave based on a changing thickness of that transmissive material. In some embodiments, the kinoform may be (or may be described as) a diffractive optical element. In some embodiments, the kinoform is arranged to apply a phase-delay to each pixel of the holographic reconstruction of between 0 and 2pi radians, optionally 0 and pi radians. In some embodiments, the spatially modulated light comprises light of a plurality of different wavelengths. The kinoform may be arranged to apply a phase-delay to each pixel of the holographic reconstruction of between 0 and 2pi radians, optionally 0 and pi radians, for the longest wavelength. The kinoform may be arranged to apply a phase-delay to one or more of the pixels of the holographic reconstruction that is greater than pi radians fora wavelength of the spatially modulated light that is shorter than the longest wavelength. In some embodiments, the kinoform may comprise an array of (discrete I contiguous) zones I portions I regions, wherein each zone I portion I region is arranged to apply a different phase-delay to incident light to the phase-delay applied by an adjacent zone I portion I region. These zones are generally referred to herein as “phase-delay areas”. The kinoform may be arranged such that the width or the height of an area occupied by each zone I portion I region on the kinoform is substantially equal to the distance between adjacent pixels of the holographic reconstruction. As described above, the movement assembly may be arranged to move the kinoform such that the kinoform has a first position with respect to the holographic reconstruction and a second position with respect to the holographic reconstruction. In the first position, each respective pixel of the holographic reconstruction may be aligned with a respective zone I portion I region of the kinoform. In the second position, each pixel of the holographic reconstruction may be aligned with a different zone I portion I region of the kinoform to zone I portion / region that the respective pixel is aligned with in the first position of the kinoform. In some embodiments, in the first position of the kinoform with respect to the holographic reconstruction, each individual pixel may be associated I aligned with a unique zone I portion I region. In other words, at most one pixel or light point of the intermediate holographic reconstruction may be associated with each zone I portion I region of the kinoform in the first position. Similarly, in the second position of the kinoform with respect to the holographic reconstruction, each individual pixel may be associated I aligned with a unique zone I portion I region of the kinoform that is different the zone I portion / region that the respective pixel is associated I aligned with when the kinoform is in the first position. In some embodiments, there is a one to one ratio between the number of zones / portions / regions (“phase-delay areas”) of the kinoform and the number of pixels / points of light (“image pixels”) of the holographic reconstruction. In other embodiments, there may be more zones I portions I regions of the kinoform than there are pixels I points of light in the holographic reconstruction. This may be to compensate for the fact that the kinoform is moved with respect to the intermediate holographic reconstruction. The kinoform may be arranged such that each pixel I point of light of the intermediate holographic reconstruction is associated with a zone I portion I region of the kinoform in all positions of the kinoform (including the intermediate positions between the first and second position described above). This may require there to be more zone I portions I regions than there are pixels I points of light, particularly if the kinoform is moved translationally with respect to the holographic reconstruction. Each zone I portion / region (“phase-delay area”) of the kinoform may be arranged to apply a phase-delay having one of a plurality of (different) allowable or discrete values. The number of allowable (or discrete) values may be less than the number of pixels I points of light of the intermediate holographic reconstruction such that the same phase-delay is applied to multiple pixels of the holographic reconstruction (although these pixels I zones are not adjacent to one another). For example, the skilled person will appreciate that only two different allowable values for the phase-delay are needed in order for the kinoform to achieve a phase-delay being applied to each pixel with each pixel having a different phasedelay to adjacent pixels (for example, the kinoform could have a chequerboard structure in which zones arranged to apply a first phase-delay are alternated with zone arranged to apply a second phase-delay that is different to the first phase delay). A kinoform having a relatively lower number of allowable values for the phase-delay is simpler and cheaper to manufacture but a kinoform having a relatively higher number of allowable values for the phase-delay may provide an improved despeckling effect. The inventors have found that a good compromise between despeckling ability and manufacturing complexity is if the number of allowable values of phase delay consist of four or more (different) allowable values and I or consists of twelve or fewer (different) allowable values, optionally ten or fewer (different) allowable values. The thickness of the kinoform within each zone I portion I regions (“phase-delay area”) may be constant. In other words, the zones I portions I regions of the kinoform may be defined by areas of constant I consistent thickness. However, the thickness of adjacent zones I portions I regions of the kinoform may be different. The amount of phase delay may be related to the thickness of the kinoform. So, the thickness of the zones may correspond to the phase-delay patterns described above. For example, each zone / portion / region of the kinoform may have a thickness having one of a plurality of allowable values. The number of allowable values of thickness may be equal to the number of allowable values of phase-delay described above. The plurality of allowable values of thickness may consist of four or more allowable values, the plurality of allowable values of thickness may consist of twelve or fewer allowable values, optionally ten or fewer allowable values. In some embodiments, the kinoform may comprise a material having a refractive index that is greater than 1. The refractive index of the kinoform may be greater than the refractive index of air. Thus, light may propagate more slowly through the kinoform than through air. A phase-delay can therefore be achieved by providing a kinoform having different thicknesses in different zones I regions I areas. The phase-delay will be greater in zones I regions I areas of greater thickness. The kinoform may comprise a substantially transparent material. In some embodiments, the kinoform comprises glass or quartz. In some embodiments, the kinoform is arranged transmit light of the holographic reconstruction. In such embodiments, the kinoform may be positioned directly between the first and second lens such that the optical axis of the first and second lenses are parallel. In other embodiments, the kinoform may be arranged to reflect light. For example, the kinoform may comprise a first layer of a transparent material such as glass I quartz (as above) and a second layer of reflective material such as silver or aluminum which may be a coating on the transparent material layer. The transparent material layer may be arranged to apply the phase-delay and the reflective layer may be arranged to reflect light back through the transparent material layer (such that a wavefront is propagated through the transparent material layer twice). In such embodiments, the optical axis of the first lens and I or second lens may be angled with respect to a normal of the first plane defined by the planar kinoform (for example, at 45 degrees). Similarly, the optical axis of the first lens may be angled with respect to the optical axis of the second lens (for example, at 90 degrees). In some embodiments, the kinoform may be displayed on a second spatial light modulator such as a liquid crystal on silicon spatial light modulator (which is not the same display device of the projector arranged to display the hologram, mentioned above). A first aspect of the present disclosure is a kinoform arranged for rotation. The kinoform is (e.g. uniformly) divided into a plurality of concentric zones. Each concentric zone (e.g. uniformly) sub-divided, in the circumferential direction, into a plurality of phase-delay areas. The number of phase-delay areas per concentric zone increases with the radius of the concentric zone. Adjoining phase-delay areas, in the radial and circumferential directions, have different phase-delays. For example, all adjoining phase-delay areas may have different phase-delays. The kinoform of the first aspect is optimized for rotation. In particular, the kinoform of the first aspect can be coordinated with the image pixels of an image such that the image pixels spend the minimum amount of time split between two phase-delay areas during movement of the kinoform. It may be said that the continuous transitions of the image pixels between the different phase-delay areas of the kinoform are optimised for rotational movement. It may also be said that the arrangement of phase-delay areas is synergistic with rotational movement of the kinoform. More broadly, aspects of the present disclosure reduce laser speckle in content displayed via a holographic display, improving image quality and improving resolution by enabling finer features to be resolved by the viewer. Furthermore, the depth information of the hologram is preserved, allowing true 3D content to be displayed. The (phase-delay or light spot) pattern described in this disclosure enables using a rotational motion rather than a linear motion, eliminating any slowing of the despeckling component and thereby achieving a lower and more consistent level of speckle in the image content, leading to a higher perceived image quality. The concentric zones may have substantially uniform size in the radial direction. The phase-delay areas may have substantially uniform size in the circumferential direction. The kinoform may comprise 5 to 15 concentric zones such as 8 to 12 concentric zones. The smallest (in terms of area, radius or circumference) concentric zone may comprise 5 to 10 phase-delay areas such as 8 to 12 phase-delay areas. The phase-delay of each phase-delay area may be 0 to 2tt radians. Each phase-delay area may be defined or delimited by two opposing straight sides, each of which is part of a different radial line. Each phase-delay area may be additionally or alternatively defined or delimited by two opposing curved sides, each of which is part of a boundary of a different concentric zone. The kinoform may comprise a dead zone (e.g. a transparent zone or zone with no phasedelay) at the centre of the concentric zones. The size of each phase-delay area in the radial direction may be substantially equal to its size in the circumferential direction. However, in an exception, one phase-delay area of one concentric zone may be longer than the others (of that concentric zone) in the circumferential direction (e.g. by an amount less than the size of the phase-delay areas in the circumferential direction). The size of each phase-delay area in the radial direction may be substantially equal to its size in the circumferential direction. However, in another exception, phase-delay areas of one concentric zone (e.g. the outermost concentric zone) may be longer than those of the other concentric zones in the circumferential direction (e.g. by an amount much less than the size of the phase-delay areas of the other concentric zones in the circumferential direction). The phase-delays of the kinoform may be pseudo random. A second aspect of the present disclosure is a hologram arranged to form holographic reconstructions (e.g. projected pictures) using a pixel pattern comprising a plurality of pixel positions. The pixel positions are disposed on a plurality of concentric circles (e.g. uniformly separated) around a centre of the pixel pattern. The pixel positions of each concentric circle are (e.g. uniformly) spaced in the circumferential direction. The number of pixel positions per concentric circle increases with the radius of the concentric circle. A third aspect of the present disclosure is a holographic projector arranged to display a hologram and form a holographic reconstruction in accordance with the second aspect. The holographic projector further comprises a kinoform accordingly to the first aspect. Each pixel of the holographic reconstruction is positionally / spatially aligned with a respective phasedelay area of the kinoform. A fourth aspect of the present disclosure is a method of improving the perceived quality of a holographic reconstruction. The method comprises a first step of rotating a kinoform. The kinoform is (e.g. uniformly) divided into a plurality of concentric zones. Each concentric zone is (e.g. uniformly) sub-divided in the circumferential direction into a plurality of phase-delay areas having. The number of phase-delay areas per concentric zone increases with the radius of the concentric zone. Adjoining phase-delay areas, in the radial and circumferential directions of the kinoform, have different phase-delays. There is also disclosed herein a kinoform arranged for rotation, wherein the kinoform is divided into a plurality of hexagonal or dodecagonal phase-delay areas arranged in a hexagonal array. The phase-delay areas are arranged in rows. The phase-delay areas of one row may be at least partially interspersed into gaps between the phase-delay areas of the adjacent rows. That is, the phase-delay areas - or rows thereof - may be “closest-packed” or “close-packed”. Each phase-delay area has a different phase-delay to its immediate neighbours or the adjoining phase-delay areas. Each phase-delay may be pseudo-random. In the present disclosure, the term “replica” is merely used to reflect that spatially modulated light is divided such that a complex light field is directed along a plurality of different optical paths. The word “replica” is used to refer to each occurrence or instance of the complex light field after a replication event - such as a partial reflection-transmission by a pupil expander. Each replica travels along a different optical path. Some embodiments of the present disclosure relate to propagation of light that is encoded with a hologram, not an image - i.e., light that is spatially modulated with a hologram of an image, not the image itself. It may therefore be said that a plurality of replicas of the hologram are formed. The person skilled in the art of holography will appreciate that the complex light field associated with propagation of light encoded with a hologram will change with propagation distance. Use herein of the term “replica” is independent of propagation distance and so the two branches or paths of light associated with a replication event are still referred to as “replicas” of each other even if the branches are a different length, such that the complex light field has evolved differently along each path. That is, two complex light fields are still considered “replicas” in accordance with this disclosure even if they are associated with different propagation distances - providing they have arisen from the same replication event or series of replication events. A “diffracted light field” or “diffractive light field” in accordance with this disclosure is a light field formed by diffraction. A diffracted light field may be formed by illuminating a corresponding diffractive pattern. In accordance with this disclosure, an example of a diffractive pattern is a hologram and an example of a diffracted light field is a holographic light field or a light field forming a holographic reconstruction of an image. The holographic light field forms a (holographic) reconstruction of an image on a replay plane. The holographic light field that propagates from the hologram to the replay plane may be said to comprise light encoded with the hologram or light in the hologram domain. A diffracted light field is characterized by a diffraction angle determined by the smallest feature size of the diffractive structure and the wavelength of the light (of the diffracted light field). In accordance with this disclosure, it may also be said that a “diffracted light field” is a light field that forms a reconstruction on a plane spatially separated from the corresponding diffractive structure. An optical system is disclosed herein for propagating a diffracted light field from a diffractive structure to a viewer. The diffracted light field may form an image. The term “hologram” is used to refer to the recording which contains amplitude information or phase information, or some combination thereof, regarding the object. The term “holographic reconstruction” is used to refer to the optical reconstruction of the object which is formed by illuminating the hologram. The system disclosed herein is described as a “holographic projector” because the holographic reconstruction is a real image and spatially-separated from the hologram. The term “replay field” is used to refer to the 2D area within which the holographic reconstruction is formed and fully focused. If the hologram is displayed on a spatial light modulator comprising pixels, the replay field will be repeated in the form of a plurality diffracted orders wherein each diffracted order is a replica of the zeroth-order replay field. The zeroth-order replay field generally corresponds to the preferred or primary replay field because it is the brightest replay field. Unless explicitly stated otherwise, the term “replay field” should be taken as referring to the zeroth-order replay field. The term “replay plane” is used to refer to the plane in space containing all the replay fields. The terms “image”, “replay image” and “image region” refer to areas of the replay field illuminated by light of the holographic reconstruction. In some embodiments, the “image” may comprise discrete spots which may be referred to as “image spots” or, for convenience only, “image pixels”. The terms “encoding”, “writing” or “addressing” are used to describe the process of providing the plurality of pixels of the SLM with a respective plurality of control values which respectively determine the modulation level of each pixel. It may be said that the pixels of the SLM are configured to “display” a light modulation distribution in response to receiving the plurality of control values. Thus, the SLM may be said to “display” a hologram and the hologram may be considered an array of light modulation values or levels. It has been found that a holographic reconstruction of acceptable quality can be formed from a “hologram” containing only phase information related to the Fourier transform of the original object. Such a holographic recording may be referred to as a phase-only hologram. Embodiments relate to a phase-only hologram but the present disclosure is equally applicable to amplitude-only holography. The present disclosure Is also equally applicable to forming a holographic reconstruction using amplitude and phase information related to the Fourier transform of the original object. In some embodiments, this is achieved by complex modulation using a so-called fully complex hologram which contains both amplitude and phase information related to the original object. Such a hologram may be referred to as a fully-complex hologram because the value (grey level) assigned to each pixel of the hologram has an amplitude and phase component. The value (grey level) assigned to each pixel may be represented as a complex number having both amplitude and phase components. In some embodiments, a fully-complex computer-generated hologram is calculated. Reference may be made to the phase value, phase component, phase information or, simply, phase of pixels of the computer-generated hologram or the spatial light modulator as shorthand for “phase-delay”. That is, any phase value described is, in fact, a number (e.g. in the range 0 to 2tt) which represents the amount of phase retardation provided by that pixel. For example, a pixel of the spatial light modulator described as having a phase value of tt / 2 will retard the phase of received light by tt / 2 radians. In some embodiments, each pixel of the spatial light modulator is operable in one of a plurality of possible modulation values (e.g. phase delay values). The term “grey level” may be used to refer to the plurality of available modulation levels. For example, the term “grey level” may be used for convenience to refer to the plurality of available phase levels in a phase-only modulator even though different phase levels do not provide different shades of grey. The term “grey level” may also be used for convenience to refer to the plurality of available complex modulation levels in a complex modulator. The hologram therefore comprises an array of grey levels - that is, an array of light modulation values such as an array of phase-delay values or complex modulation values. The hologram is also considered a diffractive pattern because it is a pattern that causes diffraction when displayed on a spatial light modulator and illuminated with light having a wavelength comparable to, generally less than, the pixel pitch of the spatial light modulator. Reference is made herein to combining the hologram with other diffractive patterns such as diffractive patterns functioning as a lens or grating. For example, a diffractive pattern functioning as a grating may be combined with a hologram to translate the replay field on the replay plane or a diffractive pattern functioning as a lens may be combined with a hologram to focus the holographic reconstruction on a replay plane in the near field. Although different embodiments and groups of embodiments may be disclosed separately in the detailed description which follows, any feature of any embodiment or group of embodiments may be combined with any other feature or combination of features of any embodiment or group of embodiments. That is, all possible combinations and permutations of features disclosed in the present disclosure are envisaged. BRIEF DESCRIPTION OF THE DRAWINGS Specific embodiments are described by way of example only with reference to the following figures: Figure 1 is a schematic showing a reflective SLM producing a holographic reconstruction on a screen; Figure 2 shows an image for projection comprising eight image areas / components, V1 to V8, and cross-sections of the corresponding hologram channels, H1-H8; Figure 3 shows a hologram displayed on an LCOS that directs light into a plurality of discrete areas; Figure 4 shows a system, including a display device that displays a hologram that has been calculated as illustrated in Figures 2 and 3; Figure 5A shows a perspective view of a first example two-dimensional pupil expander comprising two replicators each comprising pairs of stacked surfaces; Figure 5B shows a perspective view of a first example two-dimensional pupil expander comprising two replicators each in the form of a solid waveguide; Figure 6A shows an idealised view of a portion of image pixels of a first image in the form of a uniform array of pixels, the idealised view being absent of interference effects between pixels; Figure 6B shows an idealised view of a portion of image pixels of a second image in the form of a uniform array of pixels, the spacing between the pixels of the second image being less than the spacing between the pixels of the first image, and the idealised view being absent of interference effects between pixels; Figure 6C shows the portion of the first image of Figure 6A in reality (including interference effects between pixels); Figure 6D shows the portion of the second image of Figure 6B in reality (including interference effects between pixels) and how the relatively smaller pixel spacing in Figure 6D results in significant deviation from the ideal case as a result of speckle; Figure 7A shows representations of the light field associated with individual pixels of the image of Figure 6A16C; Figure 7B shows representations of the light field associated with individual pixels of the image of Figure 6B / 6D; Figure 8 shows a schematic cross-sectional view showing features of a holographic projector according to the present disclosure; Figure 9 shows a front view of a portion of a kinoform of the holographic projector shown in Figure 8 showing pixels of a first holographic reconstruction superimposed thereon; Figure 10 shows a perspective view of the kinoform of Figure 9; Figure 11A shows a front view of a larger portion of the kinoform of Figure 9 and 10 shown in a first position relative to a holographic reconstruction; Figure 11B shows a front view of the larger portion of the kinoform of Figure 11A shown in a second position relative to the holographic reconstruction; Figure 12 shows a flow chart of a method of holographic projector according to the present disclosure; Figure 13 shows a portion of a holographic projector according to the present disclosure; Figure 14 shows a front view of a portion of the kinoform of Figure 9 with pixels of a second (pre-distorted) holographic reconstruction superimposed thereon; Figure 15 shows a front view of a portion of an example of a striped kinoform according to the present disclosure, showing pixel of the second holographic reconstruction superimposed thereon; Figure 16 shows a perspective view of the portion of the striped kinoform of Figure 15; Figure 17A shows a front view of a larger portion of the striped kinoform of Figure 14 and 15 shown in a first position relative to the second holographic reconstruction; Figure 17B shows a front view of the larger portion of the striped kinoform of Figure 17A shown in a second position relative to the second holographic reconstruction; and Figure 18 shows a square array of square phase-delay areas in optimal orientation; Figure 19 shows a square array of square phase-delay areas in a sub-optimal orientation; Figure 20 shows hexagonal array of hexagonal phase-delay areas optimised for rotation; Figure 21 shows a hexagonal array of dodecagonal phase-delay areas optimised for rotation; Figure 22 shows a first embodiment comprising concentric rings of evenly spaced phasedelay areas; and Figure 23 shows a second embodiment comprising concentric rings of evenly spaced phasedelay areas. The same reference numbers will be used throughout the drawings to refer to the same or like parts. DETAILED DESCRIPTION OF EMBODIMENTS The present invention is not restricted to the embodiments described in the following but extends to the full scope of the appended claims. That is, the present invention may be embodied in different forms and should not be construed as limited to the described embodiments, which are set out for the purpose of illustration. Terms of a singular form may include plural forms unless specified otherwise. A structure described as being formed at an upper portion / lower portion of another structure or on / under the other structure should be construed as including a case where the structures contact each other and, moreover, a case where a third structure is disposed there between. In describing a time relationship - for example, when the temporal order of events is described as “after”, “subsequent”, “next”, “before” or suchlike - the present disclosure should be taken to include continuous and non-continuous events unless otherwise specified. For example, the description should be taken to include a case which is not continuous unless wording such as “just”, “immediate” or “direct” is used. Although the terms “first”, “second”, etc. may be used herein to describe various elements, these elements are not to be limited by these terms. These terms are only used to distinguish one element from another. For example, a first element could be termed a second element, and, similarly, a second element could be termed a first element, without departing from the scope of the appended claims. Features of different embodiments may be partially or overall coupled to or combined with each other, and may be variously inter-operated with each other. Some embodiments may be carried out independently from each other, or may be carried out together in codependent relationship. In the present disclosure, the term “substantially” when applied to a structural units of an apparatus may be interpreted as the technical feature of the structural units being produced within the technical tolerance of the method used to manufacture it. Conventional optical configuration for holographic projection Figure 1 shows an embodiment in which a computer-generated hologram is encoded on a single spatial light modulator. The computer-generated hologram is a Fourier transform of the object for reconstruction. It may therefore be said that the hologram is a Fourier domain or freguency domain or spectral domain representation of the object. In this embodiment, the spatial light modulator is a reflective liguid crystal on silicon, “LCOS”, device. The hologram is encoded on the spatial light modulator and a holographic reconstruction is formed at a replay field, for example, a light receiving surface such as a screen or diffuser. A light source 110, for example a laser or laser diode, is disposed to illuminate the SLM 140 via a collimating lens 111. The collimating lens causes a generally planar wavefront of light to be incident on the SLM. In Figure 1, the direction of the wavefront is off-normal (e.g. two or three degrees away from being truly orthogonal to the plane of the transparent layer). However, in other embodiments, the generally planar wavefront is provided at normal incidence and a beam splitter arrangement is used to separate the input and output optical paths. In the embodiment shown in Figure 1, the arrangement is such that light from the light source is reflected off a mirrored rear surface of the SLM and interacts with a lightmodulating layer to form an exit wavefront 112. The exit wavefront 112 is applied to optics including a Fourier transform lens 120, having its focus at a screen 125. More specifically, the Fourier transform lens 120 receives a beam of modulated light from the SLM 140 and performs a frequency-space transformation to produce a holographic reconstruction at the screen 125. Notably, in this type of holography, each pixel of the hologram contributes to the whole reconstruction. There is not a one-to-one correlation between specific points (or image pixels) on the replay field and specific light-modulating elements (or hologram pixels). In other words, modulated light exiting the light-modulating layer is distributed across the replay field. In these embodiments, the position of the holographic reconstruction in space is determined by the dioptric (focusing) power of the Fourier transform lens. In the embodiment shown in Figure 1, the Fourier transform lens is a physical lens. That is, the Fourier transform lens is an optical Fourier transform lens and the Fourier transform is performed optically. Any lens can act as a Fourier transform lens but the performance of the lens will limit the accuracy of the Fourier transform it performs. The skilled person understands how to use a lens to perform an optical Fourier transform In some embodiments of the present disclosure, the lens of the viewer’s eye performs the hologram to image transformation. Hologram calculation In some embodiments, the computer-generated hologram is a Fourier transform hologram, or simply a Fourier hologram or Fourier-based hologram, in which an image is reconstructed in the far field by utilising the Fourier transforming properties of a positive lens. The Fourier hologram is calculated by Fourier transforming the desired light field in the replay plane back to the lens plane. Computer-generated Fourier holograms may be calculated using Fourier transforms. Embodiments relate to Fourier holography and Gerchberg-Saxton type algorithms by way of example only. The present disclosure is equally applicable to Fresnel holography and Fresnel holograms which may be calculated by a similar method. In some embodiments, the hologram is a phase or phase-only hologram. However, the present disclosure is also applicable to holograms calculated by other techniques such as those based on point cloud methods. In some embodiments, the hologram engine is arranged to exclude from the hologram calculation the contribution of light blocked by a limiting aperture of the display system. British patent application 2101666.2, filed 5 February 2021 and incorporated herein by reference, discloses a first hologram calculation method in which eye-tracking and ray tracing are used to identify a sub-area of the display device for calculation of a point cloud hologram which eliminates ghost images. The sub-area of the display device corresponds with the aperture, of the present disclosure, and is used exclude light paths from the hologram calculation. British patent application 2112213.0, filed 26 August 2021 and incorporated herein by reference, discloses a second method based on a modified Gerchberg-Saxton type algorithm which includes steps of light field cropping in accordance with pupils of the optical system during hologram calculation. The cropping of the light field corresponds with the determination of a limiting aperture of the present disclosure. British patent application 2118911.3, filed 23 December 2021 and also incorporated herein by reference, discloses a third method of calculating a hologram which includes a step of determining a region of a so-called extended modulator formed by a hologram replicator. The region of the extended modulator is also an aperture in accordance with this disclosure. In some embodiments, there is provided a real-time engine arranged to receive image data and calculate holograms in real-time using the algorithm. In some embodiments, the image data is a video comprising a sequence of image frames. In other embodiments, the holograms are pre-calculated, stored in computer memory and recalled as needed for display on a SLM. That is, in some embodiments, there is provided a repository of predetermined holograms. Large eye-box using small display device Broadly, the present disclosure relates to image projection. It relates to a method of image projection and an image projector which comprises a display device. The present disclosure also relates to a projection system comprising the image projector and a viewing system, in which the image projector projects or relays light from the display device to the viewing system. The present disclosure is egually applicable to a monocular and binocular viewing system. The viewing system may comprise a viewer’s eye or eyes. The viewing system comprises an optical element having optical power (e.g., lens / es of the human eye) and a viewing plane (e.g., retina of the human eye / s). The projector may be referred to as a ‘light engine’. The display device and the image formed (or perceived) using the display device are spatially separated from one another. The image is formed, or perceived by a viewer, on a display plane. In some embodiments, the image is a virtual image and the display plane may be referred to as a virtual image plane. In other examples, the image is a real image formed by holographic reconstruction and the image is projected or relayed to the viewing plane. In these other examples, spatially modulated light of an intermediate holographic reconstruction formed either in free space or on a screen or other light receiving surface between the display device and the viewer, is propagated to the viewer. In both cases, an image is formed by illuminating a diffractive pattern (e.g., hologram or kinoform) displayed on the display device. The display device comprises pixels. The pixels of the display may display a diffractive pattern or structure that diffracts light. The diffracted light may form an image at a plane spatially separated from the display device. In accordance with well-understood optics, the magnitude of the maximum diffraction angle is determined by the size of the pixels and other factors such as the wavelength of the light. In embodiments, the display device is a spatial light modulator such as liquid crystal on silicon (“LCOS”) spatial light modulator (SLM). Light propagates over a range of diffraction angles (for example, from zero to the maximum diffractive angle) from the LCOS, towards a viewing entity / system such as a camera or an eye. In some embodiments, magnification techniques may be used to increase the range of available diffraction angles beyond the conventional maximum diffraction angle of an LCOS. In some embodiments, the (light of a) hologram itself is propagated to the eyes. For example, spatially modulated light of the hologram (that has not yet been fully transformed to a holographic reconstruction, i.e. image) - that may be informally said to be “encoded” with / by the hologram - is propagated directly to the viewer’s eyes. A real or virtual image may be perceived by the viewer. In these embodiments, there is no intermediate holographic reconstruction I image formed between the display device and the viewer. It is sometimes said that, in these embodiments, the lens of the eye performs a hologram-to-image conversion or transform. The projection system, or light engine, may be configured so that the viewer effectively looks directly at the display device. Reference is made herein to a “light field” which is a “complex light field”. The term “light field” merely indicates a pattern of light having a finite size in at least two orthogonal spatial directions, e.g. x and y. The word “complex” is used herein merely to indicate that the light at each point in the light field may be defined by an amplitude value and a phase value, and may therefore be represented by a complex number or a pair of values. For the purpose of hologram calculation, the complex light field may be a two-dimensional array of complex numbers, wherein the complex numbers define the light intensity and phase at a plurality of discrete locations within the light field. In accordance with the principles of well-understood optics, the range of angles of light propagating from a display device that can be viewed, by an eye or other viewing entity / system, varies with the distance between the display device and the viewing entity. At a 1 metre viewing distance, for example, only a small range of angles from an LCOS can propagate through an eye’s pupil to form an image at the retina for a given eye position. The range of angles of light rays that are propagated from the display device, which can successfully propagate through an eye’s pupil to form an image at the retina for a given eye position, determines the portion of the image that is ‘visible’ to the viewer. In other words, not all parts of the image are visible from any one point on the viewing plane (e.g., any one eye position within a viewing window such as eye-box.) In some embodiments, the image perceived by a viewer is a virtual image that appears upstream of the display device - that is, the viewer perceives the image as being further away from them than the display device. Conceptually, it may therefore be considered that the viewer is looking at a virtual image through an ‘display device-sized window’, which may be very small, for example 1cm in diameter, at a relatively large distance, e.g., 1 metre. And the user will be viewing the display device-sized window via the pupil(s) of their eye(s), which can also be very small. Accordingly, the field of view becomes small and the specific angular range that can be seen depends heavily on the eye position, at any given time. A pupil expander addresses the problem of how to increase the range of angles of light rays That are propagated from the display device that can successfully propagate through an eye’s pupil to form an image. The display device is generally (in relative terms) small and the projection distance is (in relative terms) large. In some embodiments, the projection distance is at least one - such as, at least two - orders of magnitude greater than the diameter, or width, of the entrance pupil and / or aperture of the display device (i.e., size of the array of pixels). Use of a pupil expander increases the viewing area (i.e., user’s eye-box) laterally, thus enabling some movement of the eye / s to occur, whilst still enabling the user to see the image. As the skilled person will appreciate, in an imaging system, the viewing area (user’s eye box) is the area in which a viewer’s eyes can perceive the image. The present disclosure encompasses non-infinite virtual image distances - that is, near-field virtual images. Conventionally, a two-dimensional pupil expander comprises one or more one-dimensional optical waveguides each formed using a pair of opposing reflective surfaces, in which the output light from a surface forms a viewing window or eye-box. Light received from the display device (e.g., spatially modulated light from a LCOS) is replicated by the or each waveguide so as to increase the field of view (or viewing area) in at least one dimension. In particular, the waveguide enlarges the viewing window due to the generation of extra rays or “replicas” by division of amplitude of the incident wavefront. The display device may have an active or display area having a first dimension that may be Less than 10 cms such as less than 5 cms or less than 2 cms. The propagation distance between the display device and viewing system may be greater than 1 m such as greater than 1.5 m or greater than 2 m. The optical propagation distance within the waveguide may be up to 2 m such as up to 1.5 m or up to 1 m. The method may be capable of receiving an image and determining a corresponding hologram of sufficient quality in less than 20 ms such as less than 15 ms or less than 10 ms. In some embodiments - described only by way of example of a diffracted or holographic light field in accordance with this disclosure - a hologram is configured to route light into a plurality of channels, each channel corresponding to a different part (i.e. sub-area) of an image. The channels formed by the diffractive structure are referred to herein as “hologram channels” merely to reflect that they are channels of light encoded by the hologram with image information. It may be said that the light of each channel is in the hologram domain rather than the image or spatial domain. In some embodiments, the hologram is a Fourier or Fourier transform hologram and the hologram domain is therefore the Fourier or frequency domain. The hologram may equally be a Fresnel or Fresnel transform hologram. The hologram may also be a point cloud hologram. The hologram is described herein as routing light into a plurality of hologram channels to reflect that the image that can be reconstructed from the hologram has a finite size and can be arbitrarily divided into a plurality of image sub-areas, wherein each hologram channel would correspond to each image sub-area. Importantly, the hologram of this example is characterised by how it distributes the image content when illuminated. Specifically and uniquely, the hologram divides the image content by angle. That is, each point on the image is associated with a unique light ray angle in the spatially modulated light formed by the hologram when illuminated - at least, a unique pair of angles because the hologram is two-dimensional. For the avoidance of doubt, this hologram behaviour is not conventional. The spatially modulated light formed by this special type of hologram, when illuminated, may be divided into a plurality of hologram channels, wherein each hologram channel is defined by a range of light ray angles (in two-dimensions). It will be understood from the foregoing that any hologram channel (i.e. sub-range of light ray angles) that may be considered in the spatially modulated light will be associated with a respective part or sub-area of the image. That is, all the information needed to reconstruct that part or sub-area of the image is contained within a sub-range of angles of the spatially modulated light formed from the hologram of the image. When the spatially modulated light is observed as a whole, there is not necessarily any evidence of a plurality of discrete light channels. Nevertheless, the hologram may still be identified. For example, if only a continuous part or sub-area of the spatially modulated light formed by the hologram is reconstructed, only a sub-area of the image should be visible. If a different, continuous part or sub-area of the spatially modulated light is reconstructed, a different sub-area of the image should be visible. A further identifying feature of this type of hologram is that the shape of the cross-sectional area of any hologram channel substantially corresponds to (i.e. is substantially the same as) the shape of the entrance pupil although the size may be different - at least, at the correct plane for which the hologram was calculated. Each light I hologram channel propagates from the hologram at a different angle or range of angles. Whilst these are example ways of characterising or identifying this type of hologram, other ways may be used. In summary, the hologram disclosed herein is characterised and identifiable by how the image content is distributed within light encoded by the hologram. Again, for the avoidance of any doubt, reference herein to a hologram configured to direct light or angularly-divide an image into a plurality of hologram channels is made by way of example only and the present disclosure is equally applicable to pupil expansion of any type of holographic light field or even any type of diffractive or diffracted light field. The system can be provided in a compact and streamlined physical form. This enables the system to be suitable for a broad range of real-world applications, including those for which space is limited and real-estate value is high. For example, it may be implemented in a head-up display (HUD) such as a vehicle or automotive HUD. In accordance with the present disclosure, pupil expansion is provided for diffracted or diffractive light, which may comprise diverging ray bundles. The diffracted light field may be defined by a “light cone”. Thus, the size of the diffracted light field (as defined on a two-dimensional plane) increases with propagation distance from the corresponding diffractive structure (i.e. display device). It can be said that the pupil expander / s replicate the hologram or form at least one replica of the hologram, to convey that the light delivered to the viewer is spatially modulated in accordance with a hologram. In some embodiments, two one-dimensional waveguide pupil expanders are provided, each one-dimensional waveguide pupil expander being arranged to effectively increase the size of the exit pupil of the system by forming a plurality of replicas or copies of the exit pupil (or light of the exit pupil) of the spatial light modulator. The exit pupil may be understood to be the physical area from which light is output by the system. It may also be said that each waveguide pupil expander is arranged to expand the size of the exit pupil of the system. It may also be said that each waveguide pupil expander is arranged to expand / increase the size of the eye box within which a viewer’s eye can be located, in order to see / receive light that is output by the system. Light channelling The hologram formed in accordance with some embodiments, angularly-divides the image content to provide a plurality of hologram channels which may have a cross-sectional shape defined by an aperture of the optical system. The hologram is calculated to provide this channelling of the diffracted light field. In some embodiments, this is achieved during hologram calculation by considering an aperture (virtual or real) of the optical system, as described above. Figures 2 and 3 show an example of this type of hologram that may be used in conjunction with a pupil expander as disclosed herein. However, this example should not be regarded as limiting with respect to the present disclosure. Figure 2 shows an image 252 for projection comprising eight image areas / components, V1 to V8. Figure 2 shows eight image components by way of example only and the image 252 may be divided into any number of components. Figure 2 also shows an encoded light pattern 254 (i.e., hologram) that can reconstruct the image 252 - e.g., when transformed by the lens of a suitable viewing system. The encoded light pattern 254 comprises first to eighth sub-holograms or components, H1 to H8, corresponding to the first to eighth image components / areas, V1 to V8. Figure 2 further shows how a hologram may decompose the image content by angle. The hologram may therefore be characterised by the channelling of light that it performs. This is illustrated in Figure 3. Specifically, the hologram in this example directs light into a plurality of discrete areas. The discrete areas are discs in the example shown but other shapes are envisaged. The size and shape of the optimum disc may, after propagation through the waveguide, be related to the size and shape of an aperture of the optical system such as the entrance pupil of the viewing system. Figure 4 shows a system 400, including a display device that displays a hologram that has been calculated as illustrated in Figures 2 and 3. The system 400 comprises a display device, which in this arrangement comprises an LCOS 402. The LCOS 402 is arranged to display a modulation pattern (or ‘diffractive pattern’) comprising the hologram and to project light that has been holographically encoded towards an eye 405 that comprises a pupil that acts as an aperture 404, a lens 409, and a retina (not shown) that acts as a viewing plane. There is a light source (not shown) arranged to illuminate the LCOS 402. The lens 409 of the eye 405 performs a hologram-to-image transformation. The light source may be of any suitable type. For example, it may comprise a laser light source. The viewing system 400 further comprises a waveguide 408 positioned between the LCOS 402 and the eye 405. The presence of the waveguide 408 enables all angular content from the LCOS 402 to be received by the eye, even at the relatively large projection distance shown. This is because the waveguide 508 acts as a pupil expander, in a manner that is well known and so is described only briefly herein. In brief, the waveguide 408 shown in Figure 4 comprises a substantially elongate formation. In this example, the waveguide 408 comprises an optical slab of refractive material, but other types of waveguide are also well known and may be used. The waveguide 408 is located so as to intersect the light cone (i.e., the diffracted light field) that is projected from the LCOS 402, for example at an oblique angle. In this example, the size, location, and position of the waveguide 408 are configured to ensure that light from each of the eight ray bundles, within the light cone, enters the waveguide 408. Light from the light cone enters the waveguide 408 via its first planar surface (located nearest the LCOS 402) and is guided at least partially along the length of the waveguide 408, before being emitted via its second planar surface, substantially opposite the first surface (located nearest the eye). As will be well understood, the second planar surface is partially reflective, partially transmissive. In other words, when each ray of light travels within the waveguide 408 from the first planar surface and hits the second planar surface, some of the light will be transmitted out of the waveguide 408 and some will be reflected by the second planar surface, back towards the first planar surface. The first planar surface is reflective, such that all light that hits it, from within the waveguide 408, will be reflected back towards the second planar surface. Therefore, some of the light may simply be refracted between the two planar surfaces of the waveguide 408 before being transmitted, whilst other light may be reflected, and thus may undergo one or more reflections, (or ‘bounces’) between the planar surfaces of the waveguide 408, before being transmitted. Figure 4 shows a total of nine “bounce” points, B0 to B8, along the length of the waveguide 408. Although light relating to all points of the image (V1-V8) as shown in Figure 2 is transmitted out of the waveguide at each “bounce” from the second planar surface of the waveguide 408, only the light from one angular part of the image (e.g. light of one of V1 to V8) has a trajectory that enables it to reach the eye 405, from each respective “bounce” point, B0 to B8. Moreover, light from a different angular part of the image, V1 to V8, reaches the eye 405 from each respective “bounce” point. Therefore, each angular channel of encoded light reaches the eye only once, from the waveguide 408, in the example of Figure 4. The waveguide 408 forms a plurality of replicas of the hologram, at the respective “bounce” points B1 to B8 along its length, corresponding to the direction of pupil expansion. As shown in Figure 5, the plurality of replicas may be extrapolated back, in a straight line, to a corresponding plurality of replica or virtual display devices 402’. This process corresponds to the step of “unfolding” an optical path within the waveguide, so that a light ray of a replica is extrapolated back to a “virtual surface” without internal reflection within the waveguide. Thus, the light of the expanded exit pupil may be considered to originate from a virtual surface (also called an “extended modulator” herein) comprising the display device 402 and the replica display devices 402’. Although virtual images, which require the eye to transform received modulated light in order to form a perceived image, have generally been discussed herein, the methods and arrangements described herein can be applied to real images. Two-Dimensional Pupil Expansion Whilst the arrangement shown in Figure 4 includes a single waveguide that provides pupil expansion in one dimension, pupil expansion can be provided in more than one dimension, for example in two dimensions. Moreover, whilst the example in Figure 4 uses a hologram that has been calculated to create channels of light, each corresponding to a different portion of an image, the present disclosure and the systems that are described herebelow are not limited to such a hologram type. Figure 5A shows a perspective view of a system 500 comprising two replicators, 504, 506 arranged for expanding a light beam 502 in two dimensions. In the system 500 of Figure 5A, the first replicator 504 comprises afirst pair of surfaces, stacked parallel to one another, and arranged to provide replication - or, pupil expansion -in a similar manner to the waveguide 408 of Figure 4. The first pair of surfaces are similarly (in some cases, identically) sized and shaped to one another and are substantially elongate in one direction. The collimated light beam 502 is directed towards an input on the first replicator 504. Due to a process of internal reflection between the two surfaces, and partial transmission of light from each of a plurality of output points on one of the surfaces (the upper surface, as shown in Figure 5A), which will be familiar to the skilled reader, light of the light beam 502 is replicated in a first direction, along the length of the first replicator 504. Thus, a first plurality of replica light beams 508 is emitted from the first replicator 504, towards the second replicator 506. The second replicator 506 comprises a second pair of surfaces stacked parallel to one another, arranged to receive each of the collimated light beams of the first plurality of light beams 508 and further arranged to provide replication - or, pupil expansion - by expanding each of those light beams in a second direction, substantially orthogonal to the first direction. The first pair of surfaces are similarly (in some cases, identically) sized and shaped to one another and are substantially rectangular. The rectangular shape is implemented for the second replicator in order for it to have length along the first direction, in order to receive the first plurality of light beams 508, and to have length along the second, orthogonal direction, in order to provide replication in that second direction. Due to a process of internal reflection between the two surfaces, and partial transmission of light from each of a plurality of output points on one of the surfaces (the upper surface, as shown in Figure 5A), light of each light beam within the first plurality of light beams 508 is replicated in the second direction. Thus, a second plurality of light beams 510 is emitted from the second replicator 506, wherein the second plurality of light beams 510 comprises replicas of the input light beam 502 along each of the first direction and the second direction. Thus, the second plurality of light beams 510 may be regarded as comprising a two-dimensional grid, or array, of replica light beams. Thus, it can be said that the first and second replicators 504, 505 of Figure 5A combine to provide a two-dimensional replicator (or, “two-dimensional pupil expander”). Thus, the replica light beams 510 may be emitted along an optical path to an expanded eye-box of a display system, such as a head-up display. In the system of Figure 5A, the first replicator 504 is a waveguide comprising a pair of elongate rectilinear reflective surfaces, stacked parallel to one another, and, similarly, the second replicator 504 is a waveguide comprising a pair of rectangular reflective surfaces, stacked parallel to one another. In other systems, the first replicator may be a solid elongate rectilinear waveguide and the second replicator may be a solid planar rectangular shaped waveguide, wherein each waveguide comprises an optically transparent solid material such as glass. In this case, the pair of parallel reflective surfaces are formed by a pair of opposed major sidewalls optionally comprising respective reflective and reflective-transmissive surface coatings, familiar to the skilled reader. Figure 5B shows a perspective view of a system 500 comprising two replicators, 520, 540 arranged for replicating a light beam 522 in two dimensions, in which the first replicator is a solid elongated waveguide 520 and the second replicator is a solid planar waveguide 540. In the system of Figure 5B, the first replicator / waveguide 520 is arranged so that its pair of elongate parallel reflective surfaces 524a, 524b are perpendicular to the plane of the second replicator / waveguide 540. Accordingly, the system comprises an optical coupler arranged to couple light from an output port of first replicator 520 into an input port of the second replicator 540. In the illustrated arrangement, the optical coupler is a planar / fold mirror 530 arranged to fold or turn the optical path of light to achieve the required optical coupling from the first replicator to the second replicator. As shown in Figure 5B, the mirror 530 is arranged to receive light - comprising a one-dimensional array of replicas extending in the first dimension - from the output port I reflective-transmissive surface 524a of the first replicator / waveguide 520. The mirror 530 is tilted so as to redirect the received light onto an optical path to an input port in the (fully) reflective surface of second replicator 540 at an angle to provide waveguiding and replica formation, along its length in the second dimension. It will be appreciated that the mirror 530 is one example of an optical element that can redirect the light in the manner shown, and that one or more other elements may be used instead, to perform this task. In the illustrated arrangement, the (partially) reflective-transmissive surface 524a of the first replicator 520 is adjacent the input port of the first replicator / waveguide 520 that receives input beam 522 at an angle to provide waveguiding and replica formation, along its length in the first dimension. Thus, the input port of first replicator / waveguide 520 is positioned at an input end thereof at the same surface as the reflective-transmissive surface 524a. The skilled reader will understand that the input port of the first replicator / waveguide 520 may be at any other suitable position. Accordingly, the arrangement of Figure 5B enables the first replicator 520 and the mirror 530 to be provided as part of a first relatively thin layer in a plane in the first and third dimensions (illustrated as an x-z plane). In particular, the size or “height” of a first planar layer - in which the first replicator 520 is located - in the second dimension (illustrated as the y dimension) is reduced. The mirror 530 is configured to direct the light away from a first layer / plane, in which the first replicator 520 is located (i.e. the “first planar layer”), and direct it towards a second layer / plane, located above and substantially parallel to the first layer / plane, in which the second replicator 540 is located (i.e. a “second planar layer”). Thus, the overall size or “height” of the system - comprising the first and second replicators 520, 540 and the mirror 530 located in the stacked first and second planar layers in the first and third dimensions (illustrated as an x-z plane) - in the second dimension (illustrated as the y dimension) is compact. The skilled reader will understand that many variations of the arrangement of Figure 5B for implementing the present disclosure are possible and contemplated. The image projector may be arranged to project a diverging or diffracted light field. In some embodiments, the light field is encoded with a hologram. In some embodiments, the diffracted light field comprises diverging ray bundles. In some embodiments, the image formed by the diffracted light field is a virtual image. In some embodiments, the first pair of parallel / complementary surfaces are elongate or elongated surfaces, being relatively long along a first dimension and relatively short along a second dimension, for example being relatively short along each of two other dimensions, with each dimension being substantially orthogonal to each of the respective others. The process of reflection / transmission of the light between / from the first pair of parallel surfaces is arranged to cause the light to propagate within the first waveguide pupil expander, with the general direction of light propagation being in the direction along which the first waveguide pupil expander is relatively long (i.e., in its “elongate” direction). There is disclosed herein a system that forms an image using diffracted light and provides an eye-box size and field of view suitable for real-world application - e.g. in the automotive industry by way of a head-up display. The diffracted light is light forming a holographic reconstruction of the image from a diffractive structure - e.g. hologram such as a Fourier or Fresnel hologram. The use diffraction and a diffractive structure necessitates a display device with a high density of very small pixels (e.g. 1 micrometer) - which, in practice, means a small display device (e.g. 1 cm). The inventors have addressed a problem of how to provide 2D pupil expansion with a diffracted light field e.g. diffracted light comprising diverging (not collimated) ray bundles. In some embodiments, the display system comprises a display device - such as a pixelated display device, for example a spatial light modulator (SLM) or Liquid Crystal on Silicon (LcoS) SLM - which is arranged to provide or form the diffracted or diverging light. In such aspects, the aperture of the spatial light modulator (SLM) is a limiting aperture of the system. That is, the aperture of the spatial light modulator - more specifically, the size of the area delimiting the array of light modulating pixels comprised within the SLM - determines the size (e.g. spatial extent) of the light ray bundle that can exit the system. In accordance with this disclosure, it is stated that the exit pupil of the system is expanded to reflect that the exit pupil of the system (that is limited by the small display device having a pixel size for light diffraction) is made larger or bigger or greater in spatial extend by the use of at least one pupil expander. The converging or diverging light field may be said to have “a light field size”, defined in a direction substantially orthogonal to a propagation direction of the light field. Because the light is diffracted I diverging, the light field size increases with propagation distance. In some embodiments, the diffracted light field is spatially-modulated in accordance with a hologram. In other words, in such aspects, the diffractive light field comprises a “holographic light field”. The hologram may be displayed on a pixelated display device. The hologram may be a computer-generated hologram (CGH). It may be a Fourier hologram or a Fresnel hologram or a point-cloud hologram or any other suitable type of hologram. The hologram may, optionally, be calculated so as to form channels of hologram light, with each channel corresponding to a different respective portion of an image that is intended to be viewed (or perceived, if it is a virtual image) by the viewer. The pixelated display device may be configured to display a plurality of different holograms, in succession or in sequence. Each of the aspects and embodiments disclosed herein may be applied to the display of multiple holograms. The output port of the first waveguide pupil expander may be coupled to an input port of a second waveguide pupil expander. The second waveguide pupil expander may be arranged to guide the diffracted light field - including some of, preferably most of, preferably all of, the replicas of the light field that are output by the first waveguide pupil expander - from its input port to a respective output port by internal reflection between a third pair of parallel surfaces of the second waveguide pupil expander. The first waveguide pupil expander may be arranged to provide pupil expansion, or replication, in a first direction and the second waveguide pupil expander may be arranged to provide pupil expansion, or replication, in a second, different direction. The second direction may be substantially orthogonal to the first direction. The second waveguide pupil expander may be arranged to preserve the pupil expansion that the first waveguide pupil expander has provided in the first direction and to expand (or, replicate) some of, preferably most of, preferably all of, the replicas that it receives from the first waveguide pupil expander in the second, different direction. The second waveguide pupil expander may be arranged to receive the light field directly or indirectly from the first waveguide pupil expander. One or more other elements may be provided along the propagation path of the light field between the first and second waveguide pupil expanders. The first waveguide pupil expander may be substantially elongated and the second waveguide pupil expander may be substantially planar. The elongated shape of the first waveguide pupil expander may be defined by a length along a first dimension. The planar, or rectangular, shape of the second waveguide pupil expander may be defined by a length along a first dimension and a width, or breadth, along a second dimension substantially orthogonal to the first dimension. A size, or length, of the first waveguide pupil expander along its first dimension make correspond to the length or width of the second waveguide pupil expander along its first or second dimension, respectively. A first surface of the pair of parallel surfaces of the second waveguide pupil expander, which comprises its input port, may be shaped, sized, and / or located so as to correspond to an area defined by the output port on the first surface of the pair of parallel surfaces on the first waveguide pupil expander, such that the second waveguide pupil expander is arranged to receive each of the replicas output by the first waveguide pupil expander. The first and second waveguide pupil expander may collectively provide pupil expansion in a first direction and in a second direction perpendicular to the first direction, optionally, wherein a plane containing the first and second directions is substantially parallel to a plane of the second waveguide pupil expander. In other words, the first and second dimensions that respectively define the length and breadth of the second waveguide pupil expander may be parallel to the first and second directions, respectively, (or to the second and first directions, respectively) in which the waveguide pupil expanders provide pupil expansion. The combination of the first waveguide pupil expander and the second waveguide pupil expander may be generally referred to as being a “pupil expander”. It may be said that the expansion / replication provided by the first and second waveguide expanders has the effect of expanding an exit pupil of the display system in each of two directions. An area defined by the expanded exit pupil may, in turn define an expanded eyebox area, from which the viewer can receive light of the input diffracted or diverging light field. The eye-box area may be said to be located on, or to define, a viewing plane. The two directions in which the exit pupil is expanded may be coplanar with, or parallel to, the first and second directions in which the first and second waveguide pupil expanders provide replication / expansion. Alternatively, in arrangements that comprise other elements such as an optical combiner, for example the windscreen (or, windshield) of a vehicle, the exit pupil may be regarded as being an exit pupil from that other element, such as from the windscreen. In such arrangements, the exit pupil may be non-coplanar and non-parallel with the first and second directions in which the first and second waveguide pupil expanders provide replication / expansion. For example, the exit pupil may be substantially perpendicular to the first and second directions in which the first and second waveguide pupil expanders provide replication / expansion. The viewing plane, and / or the eye-box area, may be non-coplanar or non-parallel to the first and second directions in which the first and second waveguide pupil expanders provide replication / expansion. For example, a viewing plane may be substantially perpendicular to the first and second directions in which the first and second waveguide pupil expanders provide replication / expansion. In order to provide suitable launch conditions to achieve internal reflection within the first and second waveguide pupil expanders, an elongate dimension of the first waveguide pupil expander may be tilted relative to the first and second dimensions of the second waveguide pupil expander. Combiner shape compensation An advantage of projecting a hologram to the eye-box is that optical compensation can be encoded in the hologram (see, for example, European patent 2936252 incorporated herein by herein). The present disclosure is compatible with holograms that compensate for the complex curvature of an optical combiner used as part of the projection system. In some embodiments, the optical combiner is the windscreen of a vehicle. Full details of this approach are provided in European patent 2936252 and are not repeated here because the detailed features of those systems and methods are not essential to the new teaching of this disclosure herein and are merely exemplary of configurations that benefit from the teachings of the present disclosure. Control device The present disclosure is also compatible with optical configurations that include a control device (e.g. light shuttering device) to control the delivery of light from a light channelling hologram to the viewer. The holographic projector may further comprise a control device arranged to control the delivery of angular channels to the eye-box position. British patent application 2108456.1, filed 14 June 2021 and incorporated herein by reference, discloses the at least one waveguide pupil expander and control device. The reader will understand from at least this prior disclosure that the optical configuration of the control device is fundamentally based upon the eye-box position of the user and is compatible with any hologram calculation method that achieves the light channeling described herein. It may be said that the control device is a light shuttering or aperturing device. The light shuttering device may comprise a 1D array of apertures or windows, wherein each aperture or window independently switchable between a light transmissive and a light non-transmissive state in order to control the delivery of hologram light channels, and their replicas, to the eye-box. Each aperture or window may comprise a plurality of liquid crystal cells or pixels. Speckle in High-Resolution Images It has been described how a (holographic) projector is arranged to relay light from a display device (SLM) to a viewing system (such as a the eye of a user) at an eye-box. The lens of the eye performs a hologram-to-image transformation such that the viewer receives an image. This image is pixelated. That is, the holographic reconstruction comprises image spots or pixels. It is generally desirable for the image to have a high enough resolution that the user cannot perceive the individual pixels. While such a resolution may be high enough to remove the perception of the individual pixels, another problem is introduced. In particular, light associated with adjacent (close together) pixels will interact (interfere) creating a pattern of dark and light areas which degrades the guality of the image. Throughout this disclosure, this effect is referred to as speckle. This is described in more detail below. Figures 6A to 6D show how the speckle effect increases as the resolution of an image increases. For simplicity, amplitude information is omitted from each of Figures 6Ato 6D. Figures 6A and 6E3 represent how the pixels of a portion of an image (or holographic reconstruction) would ideally appear in the absence of speckle (specifically, in the absence of interference effect between pixels). Figure 6A corresponds to a portion 600 of a first image having a relatively low resolution. Figure 6B show a portion 610 of a second image having a relatively high resolution. Thus, the individual pixels 602 of Figure 6A are further apart than the individual pixels 612 of Figure 6B. In the idealised case of Figures 6A and 6B, the image pixels of both the first and second images form a uniform array. Figures 6C and 6D represent how the pixels of the portions 602,610 of the images (or holographic reconstructions) appear in reality (including the speckle effect). Figure 6C shows the portion 600 of the first image but with the speckle effect. Figure 6D shows the portion 610 of the second image but with the speckle effect. Because the pixels of the first image (of Figures 6A and 6C) have a relatively large spacing, any interference between light of adjacent pixels is minimal. As such, the speckle effect in Figure 6C is substantially negligible and the pixels maintain their appearance of being in a uniform array. However, the image is unacceptably poor given its low resolution (and so pixelated) nature. Because the pixels of the second image (of Figures 6B and 6D) have a relatively small spacing, the interference between light of adjacent image pixels in the second image is much more significant than in the first image. As such, the speckle effect in Figure 6D is not substantially negligible. In particular, interference between light associated with pixels of the second image results in dark areas 614 and light areas 616 being formed (which are not present in the idealised case of Figure 6B). So, as a result of the speckle effect, the image of Figure 6D does not appear as a uniform points of light but rather is granulated and noisy. So, although the resolution in the second image may be acceptably high so as to, in theory, not appear pixelated, the speckle-effect significantly degrades the quality of the image that is actually perceived. So, in summary, without speckle mitigation, image quality may either be poor as a result of being too low resolution and appearing pixelated or (at higher resolutions) noisy and so poor because of the speckle-effect. The causes of the speckle shown in Figure 6D are explained in more detail with reference to Figures 7A and 7B. Figures 7A and 7B each show representations of the light field associated with individual pixels as well as an intensity profile of the image pixel (which has the form of a sine squared function). Figure 7A show representations of the light field associated with individual pixels of the first image (of Figure 6A / 6C). Figure 7B show representations of the light field associated with individual pixels of the second image (of Figure 6B / 6D). Figure 7A represents two adjacent pixels 702, 704 of the first image (of Figure 6A / 6C) and Figure 7B represents two adjacent pixels 706,708 of the second image (of Figure 6B / 6D). The light associated with each pixel 702 to 708 takes the form of a sine squared function, also shown in Figures 7A and 7B. A first sine squared function 712 is associated with pixel 702, a second sine squared function 714 is associated with pixel 704, a third sine squared function 716 is associated with pixel 706 and a fourth sine squared function 718 is associated with pixel 708. Only the first and second so-called side-lobes of each sine squared function (either side of the respective main lobe) are shown in Figures 7A and 7B. As discussed in relation to Figure 6A and 6C, the pixels in the first image have a relatively large spacing compared to the pixels of Figure 6B and 6D. As such, the two adjacent pixels 702,704 of the first image (shown in Figure 7A) are further apart than the two adjacent pixels 706,708 (shown in Figure 7B) of the second image. Each pixel 704 to 708 has an intensity which varies in space substantially in the shape of a sine squared function. The appearance of the pixels (which are ideally circular) may be affected by interference of the sine squared functions. In particular, the perceived intensity pattern at any point in space will depend on the sum of the interactions / sum of the value of the sine squared functions of all the pixels at that respective point in space. For simplicity, two pixels and the respective sine squared functions are shown in Figures 7A and 7E3. The spacing between pixels 702,704 is relatively large, so there is no significant overlap between first and second respective sine squared functions 712,714, associated with these pixels. Of course, some higher-order side-lobes (not shown in Figures 7A and 7B) may interact. But the amplitude of the higher-order side-lobes is so small as to be negligible and so the interference caused by the interaction of the higher-order side-lobes is also negligible. As a result, the main lobes 722,724 of the first and second sine squared functions 712,714 substantially contribute to substantially circular and spatially separated points of light (pixels 702,704). The situation is different when the pixels are closer together. The spacing between pixels 706,708 is relatively small, so there is a significant overlap between the third and fourth sine squared functions 716,718 associated with these pixels. As can be seen in Figure 7B, the lobes of the third and fourth sine squared functions 716,718 are overlapping. In particular, there is some overlap of the second order side-lobes of the third sine squared function 716 with the main lobe of the fourth sine squared function 718 (and vice versa). Similarly, the first and second order side-lobes of the third sine squared function 716 overlap with the first and second order side-lobes of the fourth since function 718. This overlapping resulting in areas of constructive interference and areas of destructive interference. For example, the interaction between the main lobe of each sine squared function with a second order side lobe of the other sine squared function is constructive. This constructive interference results in the pixels 706,708 appearing elongated I oval rather than round. The interaction between the side lobes of each sine squared function (between the main lobes) has two destructive regions 730 and one constructive region 732. The constructive region 732 results in an additional (unintentional) point of light 734 between the pixels 706,708. For simplicity, figure 7B merely shows the interaction of the sine squared function of two pixels (pixels 706,708). In reality, an image is formed by an array of pixels, with the sine squared functions of each pixel interacting with at least several other nearby pixels. For example, each pixel may have four adjacent pixels, the light of which each of which may significantly interact / interfere with light of that respective pixel. Thus, the resulting interference pattern is much more complex than that shown in Figure 7A and 7B (e.g. see the complex pattern formed in Figure 6D as a result of speckle I interfernceO. However, it should be clear from the two pixels shown in Figure 7B how the interference effect of relatively close pixels can degrade image quality. For example, Figure 7E3 shows both the effect of the elongation of pixels 706,708 and the introduction of an unintentional bright region between the pixels substantially reduces image quality. In addition, unintentional dark regions may also be created. The interference described in relation to Figure 7B and the corresponding reduction in the quality of resulting images is referred to in this disclosure as speckle. Hologram processing to reduce speckle Figure 8 is a schematic cross-sectional view showing features of a holographic projector according to the present disclosure. These features are arranged in order to minimise I reduce speckle in the images formed by the holographic projector. In particular, Figure 8 shows a spatial light modulator 800 (which, in some examples, is a liquid crystal on silicon pixelated spatial light modulator) and a magnifying telescope 801 comprising a first lens 802 (which, in some examples, is first convex Fourier lens) and a second lens 804 (which, in some examples, is a second convex Fourier lens). The first lens 802 comprises a front focal plane 812. The front focal plane 812 corresponds to the plane at which incoming collimated light is focussed by the first lens 802 and so the front focal plane 812 depends on, for example, the optical power of the lens. The second lens 804 comprises a back focal plane 814 and a front focal plane 816. The back focal plane 814 and the front focal plane 812 (of the first lens) are aligned so as to form a single surface. In some embodiments, the focal length of the first and second lenses 802, 804 is the same such that the front focal plane 812 and the back focal plane 814 are at the midpoint between the first and second lenses 802,804. However, this is optional. A kinoform 820 (which could also be described as a diffractive optical element) is disposed between the first lens 802 and the second lens 804. The kinoform 820 is disposed substantially at the front focal plane 812 of the first lens 802. The holographic projector further comprises a movement assembly arranged to move the kinoform 820. As will explained in more detail below, the position and arrangement of kinoform are such that rapid movement of the kinoform 820 with respect to the first lens (e.g. with respect to the optical axis of the first lens) substantially reduces speckle. During operation of the holographic projector, a hologram of a picture is displayed on the spatial light modulator 800. Light is emitted from a coherent light source (e.g. a laser) of the holographic projector, this light is incident the spatial light modulator 800. The light is spatially modulated in accordance with the hologram displayed on the spatial light modulator 800 to form a holographic wavefront. The holographic wavefront is received by the first (Fourier) lens 802 such that holographic light is transformed to the image domain. The first lens 802 focusses the light to form a holographic reconstruction 822 of the image (of the hologram). The holographic reconstruction 822 is pixelated and is represented by the dots 824 shown in Figure 8. The holographic reconstruction 822 is formed at the front focal plane 812 of the first lens 802 in the example shown in Figure 8. The holographic reconstruction 822 interacts with the kinoform 820. The kinoform 820 applies a phase-delay to each pixel 824 of the holographic reconstruction 822. In particular, the kinoform 820 is arranged such that the phase delay applied to each pixel 824 of the holographic reconstruction 822 is different to the phase-delay applied to each pixel that is adjacent to the respective pixel. After the phase-delay has been applied to each pixel, the holographic reconstruction 822 may be referred to as a modified holographic reconstruction (the modified holographic reconstruction having phase delays applied to the pixels relative to the holographic reconstruction 822). Light of the modified holographic reconstruction is received at the second lens 804 to be transformed I reformed into a holographic wavefront. Figures 9 and 10 show a portion of the kinoform 820 in more detail. Figure 9 is a front view of the portion of the kinoform 820 with dots I pixels 824 of the holographic reconstruction 822 superimposed thereon. The kinoform 820 comprises an array of discrete zones 826. Each zone 826 in the array of zones is substantially square in shape. In this example, the array is a four by four array of zones (in other words, the array comprises 16 zones). Each zone 826 is aligned with a single pixel 824 of the holographic reconstruction 822. The interpixel distance of pixels of the holographic reconstruction is equal to the width and length of the each zone 826. Each zone 826 of the kinoform 820 is arranged to apply a phase-delay having one of four (different) allowable or discrete values. The different phase delays are represented by the different shading of the zones in Figure 9. In particular, a first phasedelay value is represented by the diagonal shading (see zone 830, for example); a second phase-delay value is represented by the dotted shading (see zone 832, for example); a third phase-delay is represented by the vertical shading (see zone 834, for example; and a fourth phase-delay is represented by the horizontal shading (see zone 836, for example). Each of the first to fourth phase-delay is between 0 and pi radians. In some examples, the amount of phase-delay applied by the kinoform 820 to each pixel is dependent on the thickness of the kinoform in a particular zone. In such embodiments, the kinoform 820 comprises a transparent material such as glass or quartz having a refractive index greater than 1. Thus, light propagates more slowly through the kinoform 820 than through air and so the phase-delay applied to light associated with each pixel will increase with increasing thickness. Figure 10 is a perspective view of the kinoform 820 showing how each zone 830 has a thickness and how zones of the same type (i.e. zones arranged to apply the same phase-delay) have the same thickness. In particular, there are four different thicknesses of zones and these correspond to the same order as the zones shown in Figure 9. It should be clear that the number of discrete phase-delay values may be less than or greater than four (for example eight). Furthermore, Figures 9 and 10 shows only 16 pixels and 16 zones, however it should be clear that this is representative only and not limiting (in fact, generally the number of pixels and zones will be significantly higher than 16). Applying the phase-delay to the pixels 824 of the holographic reconstruction (in a way which results in the adjacent pixels having different phase-delays applied) changes the speckle pattern of the image formed at the viewing system. This is because the relative phase of the light associated with adjacent pixels is changed and so the interference pattern (e.g. shown, for example, in Figure 7E3) is also changed. Figure 9 shows a portion of the kinoform 820 in a first position with respect to the holographic reconstruction 822. However, as described above, in use of the holographic projector, the kinoform 820 is rapidly moved by the movement assembly (not shown in the Figures). The movement assembly is arranged to move the kinoform 820 translationally in a first plane (the normal of the first plane being parallel to an optical axis of the first lens 802). The movement assembly is arranged to move the kinoform 820 from a first position to a second position relative to the holographic reconstruction. The first and second position are represented by Figures 11 and 12 respectively. Figures 11A and 11B show a larger portion of the kinoform 820 than shown in Figures 9 and 10. In particular, Figures 11A and 11B show that the kinoform comprises more zones 826 than the holographic reconstruction 822 comprises pixels 824. This means that there are some zones 900 of the kinoform which are not aligned with a pixel 826 in each position. Figures 11A and 11B show that the first position and second position are separated be a distance of four pixels (but this is merely optional, other distances between the first and second position are possible). Figures 11A and 11B show that, in the second position, each zone 826 of the kinoform 820 is substantially aligned with a different pixel 824 of the holographic reconstruction 822 to the pixel of the first position. It should be clear, that as the kinoform 820 is moved from the first position to the second position (with respect to the holographic reconstruction) each pixel of the holographic reconstruction 822 will be aligned with four different zones and so have four different phase-delays applied. At all times, the phase-delay applied to adjacent pixels is different. So, as the kinoform 820 is moved, four different speckle patterns will be formed at the viewing system. The inventors have advantageously found that rapidly moving the kinoform between the first and second positions shown in Figures 11A and 11B results in a plurality of different speckle patterns being formed at the viewing system. By moving the kinoform at a sufficiently high frequency between the first and second position, this changing speckle pattern is averaged by the optic system of a human observer (because a plurality of different phase-delay values are applied and associated speckle patterns form within the integration time of the human eye). The result is that the appearance of speckle in the holographic reconstruction is reduced. In one embodiment, the inventors have found that a frequency of oscillation between the first and second position of the kinoform of 1 kHz is a suitable value. In the example above, the kinoform is moved translationally between the first position and the second position. In other examples, the kinoform may be moved rotationally. Furthermore, in the example above, the kinoform is transmissive, with light of the holographic reconstruction being transmitted through the kinoform. In other examples, the kinoform may be reflective. For example, the kinoform may comprise a reflective coating on a back surface arranged to reflect light of the holographic reconstruction. Figure 12 shows a flow chart of a method of holographic projector according to the present disclosure and using the holographic projector described above. Step 1202 of the method comprises displaying a hologram of a picture on a display device (in particular the spatial light modulator 800 described above). Step 1202 may optionally comprise calculating the hologram first. The hologram may be a computationally generated hologram. Step 1204 of the method comprises spatially modulating light in accordance with the hologram to form the wavefront. This step comprises illuminating the spatial light modulator 800 with a coherent light source (in this example, one or more lasers). The coherent light will be encoded with the hologram to form a holographic wavefront. Step 1206 of the method comprises forming a holographic reconstruction 822 of the picture by focusing the holographic wavefront towards a front focal plane 812 of a first lens 802. As described in more detail below, in relation to the embodiment of Figure 13, the exact plane at which the holographic reconstruction is in complete focus may vary, but will generally be substantially at or adjacent to the front focal plane 812 of the first lens 802. The holographic reconstruction 822 of the picture comprises pixels 824. Step 1208 of the method comprises applying a phase-delay to each pixel of the holographic reconstruction 822 using the kinoform 820 disposed between the first lens 802 and the second lens 804. The zones 826 of the kinoform 820 are arranged to apply a phase-delay to each pixel that is different to the phase-delay of the adjacent pixels. As described above, in one example, there are a discrete number of (e.g. four) different phase-delay values that are applied to the pixels. Step 1210 of the method comprises moving the kinoform 822 with respect to the holographic reconstruction 822. In this way, different zones 826 of the kinoform 822 are aligned with different pixels of the holographic reconstruction 822 overtime. Thus, each pixel of the holographic reconstruction has a plurality of different phase-delays applied thereto. The kinoform 822 is moved (between a first position and a second position) rapidly (for example, at a frequency of 1 kHz). In this way, the phase-delay for each pixel of the holographic reconstruction is changed a plurality of times within the integration time of the human eye. This changes the speckle pattern formed a plurality of times within the integration time of the human high such that the speckle pattern is averaged by the optic system (e.g. eye of a human observer). Thus, the appearance I perception of speckle in the holographic reconstruction is reduced. Step 1212 of the method comprises receiving light of the modified holographic reconstruction from the kinoform using the second lens. Step 1212 further coupling the modified wavefront into a waveguide having a pair of opposing surfaces arranged to waveguide therebetween, wherein a first surface of the pair of opposing surfaces is partially-reflective partially-transmissive such that a plurality of replicas of the modified holographic wavefront are emitted therefrom. Figure 13 shows a portion of a second embodiment of the holographic projector according to the present disclosure. The second embodiment is very similar to the first embodiment and like features are numbered accordingly. The main difference between the first embodiment and the second embodiment is that holographic projector of the second embodiment is arranged to display a first virtual image viewable from the eye-box at a first virtual image distance and a second virtual image viewable from the eye-box at a second virtual image distance. In this example, the first virtual image distance is 1.5 meters and the second virtual image distance is 25 meters. The holographic reconstruction of the embodiment of Figure 13 comprises a first portion 1302 associated with the first virtual image and a second portion 1304 associated with the second virtual image. The first portion 1302 of the holographic reconstruction is focused at a first plane 1306 and the second portion 1304 of the holographic reconstruction is focused at a second plane 1308. As will be well understood by the skilled person, the exact position where the (intermediate) holographic reconstruction is in focus between the first and second lens depends on the corresponding virtual image distance of the virtual image. In particular, the shorter the virtual image distance, the further the holographic reconstruction will be in focus from the first lens 802 (and the closer it will be to the second lens, not shown in Figure 13). If the virtual image distance is at infinity, the holographic reconstruction will be formed I in complete focus at the front focal plane 812 of the first lens 802. In this example, the first virtual image distance is less than the second virtual image distance and so the first portion 1302 of the holographic reconstruction is closer to the front focal plane 812 of the first lens 802 than the second portion 1304 of the holographic reconstruction; but both virtual image distances are non-infinite and so the first and second portions 1302, 1304 are not exactly at the front focal plane 812. The inventors have found that the despeckling effect of the kinoform is maximized I optimized when the kinoform is disposed exactly at a plane of the holographic reconstruction (in complete focus). However, in the example shown in Figure 13, the holographic reconstruction comprises two portions which are completely focused at two different places. In such cases, the inventors have found that it is advantageous to position the kinoform between the first plane 1306 and the second plane 1308 so that a despeckling effect is achieved at all depths. In particular, and as shown in Figure 13, the inventors have found that it is advantageous to position the kinoform 820 at the midpoint between the first and second planes 1306,1308. The holographic projector of Figure 13 is described as being arranged to form two virtual images at 1.5 meters and 25 meters respectively. This is merely exemplary. The kinoform may be used to reduce the speckle effect at any arbitrary distance of a virtual image and for any number of virtual images at any arbitrary distance. In each case, it is advantageous for the kinoform to be placed at an average position I mid-point between the planes formed by the plurality of portions of holographic reconstructions. Modified speckle reduction As described above, the kinoform 820 is rapidly moved between a first and second position to cause a plurality of different speckle patterns to be moved. Doing so within the integration time of the human reduces the appearance of speckle. Kinoform 820 comprises zones 826. Each pixel 824 of a holographic reconstruction is aligned with one zone 826 (for example, in the first position). Because the pixels 824 (shown in Figures 9 and 11) are in a uniform, regular, square array (e.g. are evenly spaced), the zones 826 of the kinoform 820 also form a uniform, regular, square array. Furthermore the pixels 824 shown in Figures 9 and 11 are relatively small. This means that it is relatively straightforward to have each pixel 824 of the holographic reconstruction aligned with (only) one zone simultaneously (e.g. in the first or second position of the kinoform 820). However, this is not always the case. Figure 14 shows pixels 1424 of a second holographic reconstruction. The pixels 1424 ofthe second holographic reconstruction are relatively much larger than the pixels 824. Furthermore, the pixels 1424 do not form a uniform I regular array. Herein, a horizontal direction (x-direction) will be referred to as a first direction and a vertical direction (y-di recti on), that is perpendicular to the first direction, will be referred to as a second direction. There are four lines 1402 of pixels. The four lines 1402 of pixels extend in the first direction (i.e. are horizontal I are in the x-direction). A pitch ofthe pixels in each of the four lines 1402 (i.e. a distance between the centre of adjacent pixels) in the first direction is less than a corresponding pitch ofthe pixels 824 ofthe first holographic reconstruction. In other words, the pitch in the first direction has generally been reduced in the second holographic reconstruction. Furthermore, the pitch of pixels in the first direction of each of the four lines 1402 varies along the first direction. In particular, the pitch is largest between the two middle pixels 1424 of each line such that the two middle pixels 1424 do not overlap. The two outer pixels is smaller such that the two outer pixels do overlap with their adjacent pixels (on either side ofthe middle pixels). As described previously, the reason for the reduction in, and changing, pixel pitch in the second holographic reconstruction is to compensate for the magnification of an optical component. In particular, the holographic projector is arranged to relay a holographic wavefront to an eye-box via an optical combiner which, in this example is a windscreen or windshield. The eye-box and optical combiner are not shown in the drawings. The optical combiner has a non-uniform optical power. In this example, the optical power ofthe optical combiner is greater in the first direction than in the second direction. Furthermore, the optical power of the optical combiner varies along the first direction. In this example, the optical power of the optical combiner is greater at the edges and less in the middle along the first direction. This is because the curvature of the optical combiner is generally greater in first direction than the second direction and because the optical combiner has a complex curve shape which increases with distance from the centre, along the first direction. The complex curve shape results in non-uniform magnification (e.g. distortion / warping) of the image viewable at the eye box relative to the holographic reconstruction at the plane of the kinoform. By pre-distorting the holographic reconstruction, this distortion / warping can be compensated for such that the image viewable from the eye-box appears as intended. In the example of Figure 14, the pixels 1424 have been pre-distorted I repositioned in such a way that, when the holographic wavefront is distorted by the optical combiner, an image comprising a regular array of pixels (resembling that of Figure 11) is viewable from the eye box. Such a pre-distortion process will be familiar to the skilled reader. Figure 14 shows how the pixels 1424 interact with a portion of kinoform 820. In particular, Figure 14 shows how it is no longer the case that each pixel is aligned with a single zone 826 of the kinoform. For example, outer pixels 1426 are aligned with two zones 826 each, such that a different phase delay is simultaneously applied to different portions of each of these outer pixels 1426. The inventors have found that this overlapping of zones 826 by the pixels 1424 may adversely affect the quality I clarity of a holographic projection and so have provided an alternative (second) example of kinoform 1520 which is described herein. Figure 15 shows a portion of a second kinoform 1520. The kinoform 1520 comprises a one dimensional array of four discrete zones 1526. Each zone 1526 extends in the first direction and, in this example, extends along the full length of the kinoform 1520. The zones 1526 may be described as forming stripes extending in the first direction such that the kinoform 1520 may be referred to as a striped kinoform. Because the zones 1526 extend as stripes along the first direction, the risk of an overlap or misalignment of pixels 1424 with any particular zone in the first direction is mitigated. Each pixel 1424 within a single line 1402 is aligned with the same zone (along the first direction) but adjacent lines are aligned with different (but adjacent) zones. Each zone 1526 of the kinoform 1520 is arranged to apply a phase delay having one of four (different) allowable or discrete values. The different phase delays are represented by the different shading of the zones in Figure 15. In particular, a first phase-delay value is represented by the diagonal shading in a first zone 1530; a second phase-delay value is represented by the vertical shading in a second zone 1532; a third phase-delay is represented by the horizontal shading in a third zone 1534; and a fourth phase-delay is represented by the dotted shading in a fourth zone 1536. Each of the first to fourth phasedelays are between 0 and pi radians. In some examples, the amount of phase-delay applied by the kinoform 1520 to each pixel is dependent on the thickness of the kinoform in a particular zone. In such embodiments, the kinoform 1520 comprises a transparent material such as glass or quartz having a refractive index greater than 1. Thus, light propagates more slowly through the kinoform 1520 than through air and so the phase-delay applied to light associated with each pixel will increase with increasing thickness. Figure 16 is a perspective view of the kinoform 1520 showing how each zone 1526 has a thickness and how zones of the same type (i.e. zones arranged to apply the same phase-delay) have the same thickness. In particular, there are four different thicknesses of zones and these correspond to the same order as the zones shown in Figure 15. By moving the kinoform 1520 in the second direction (i.e. perpendicularly to the first direction), different lines of pixels are brought into aligned with different zones such that the relative phase difference between adjacent pixels of adjacent lines changes. This changes the speckle pattern of the image formed at the viewing system. This is because the relative phase of the light associated with adjacent pixels between adjacent lines is changed and so the interference pattern (e.g. shown, for example, in Figure 7B) is also changed. The inventors have found that, even though relative phase between pixels is only changed in the second direction (rather than changing for all pairs of adjacent pixels, as in Figure 9 etc.), the change in the interference pattern is substantial enough to meaningfully reduce the speckle effect. The inventors have recognised that this can be achieved a risk of degrading image quality I sharpness. Figures 17A and 17B show a larger portion of the kinoform 1520 than shown in Figures 14 and 15. In particular, Figures 17A and 17B show that the kinoform comprises more zones than the holographic reconstruction comprises lines of pixels. This means that there are some zones of the kinoform which are not aligned with a pixel in each position. Figures 17A and 17B show that the first position and second position are separated be a distance of four zones (but this is merely optional, other distances between the first and second position are possible). Figures 17A and 17B show that, in the second position, each zone of the kinoform is substantially aligned with a different line of pixels of the holographic reconstruction to the pixel of the first position. It should be clear, that as the kinoform is moved from the first position to the second position (with respect to the holographic reconstruction) each pixel of the holographic reconstruction will be aligned with one of four different zones and so have four different phase-delays applied. At all times, the phase-delay applied to adjacent pixels of adjacent lines is different - but pixels within a will have the same phase-delay applied. So, as the kinoform is moved, four different speckle patterns will be formed at the viewing system. The example of the second kinoform 1520 and the second holographic reconstruction are schematic and exemplary only. For example, a four by four array of pixels 1424 is shown in the figures. However, it should be clear that, typically, the holographic reconstruction will comprise many more pixels, for example 1 or 2 megapixels. Furthermore, because only a few pixels are shown in the Figures, the reduced and varying pitch of the pixels 1424 in the first direction has been exaggerated Optimised configurations for rotation British patent application 2300506.9 filed 13 January 20023 discloses a component (e.g. despeckling component or kinoform) which can be used for despeckling by placing it at an intermediate image plane and translating it perpendicular to the optical axis if the system. The pattern of this component is designed such that it disrupts the interference between adjacent image pixels while maintaining the phase within a single image pixel, thereby enabling high image quality over a range of content depths. However, moving in a linear manner means that there is a time when the despeckling component must turn around and is therefore moving slowly. This turn around means that fewer different phase states of the despeckler are accessible and the interference between adjacent pixels is not disrupted or is disrupted more weakly, leading to an increase in laser speckle during the turn around time of the despeckling component. Therefore, while laser speckle is reduced there is a limit to the effectiveness of despeckling due to the linear driving of the component. Increasing the frequency of the linear motion leads to undesirable audible noise while increasing the amplitude leads to a larger size and higher power consumption. One method of some embodiments to eliminate the turn around of the despeckling component is to move to a continuous circular drive motion, such as may be applied by mounting the component on the drive shaft of a motor. However, if the pattern previously used when driving in a linear motion is applied to a circular drive motion then the performance is reduced. This is illustrated with reference to Figures 18 and 19. Figure 18 shows a kinoform 1800 arranged for rotation about a centre 1810. An expanded view of a portion 1820 of the kinoform 1800 is shown to illustrate the structure of the kinoform in relation to the image pixels (or, at least, the image pixel positions) of the holographic reconstruction (or “image”). The expanded view shows a 4 x 4 array of phasedelay areas 1830 (also referred to herein as “phase zones”) such as first phase-delay area 1840. Each phase-delay area is square or at least quadrilateral. Figure 18 also shows a regular, square array of 4 x4 image pixels such as first image pixel 1850. Each image pixel is circular or elliptical. Figure 18 shows an image pixel is formed in the centre of each phase-delay area. The array of phase-delay areas and image pixels are coordinated in this orientation of the kinoform. However, this coordination may be disrupted by rotation of the kinoform. Figure 18 shows the component in its optimal orientation, with the orientation of the pattern aligned with the orientation of the image pixels at the intermediate image plane. By design, there is at most one image pixel within a given phase zone. However, at some time later the component (i.e. the kinoform) will have rotated by 45° into the orientation shown in Figure 19. Figure 19 shows a kinoform 1900 arranged for rotation about a centre 1910. An expanded view of a portion 1920 of the kinoform 1900 is shown to illustrate the structure of the kinoform in relation to the image pixels (or, at least, the image pixel positions) of the holographic reconstruction (or “image”). The expanded view shows a 6 x 6 array of phasedelay areas 1830 (“phase zones”) such as first phase-delay area 1940. Each phase-delay area is still square or at least quadrilateral - but the array is globally rotated. The array of phase-delay areas of Figure 19 is rotated 45 degrees counterclockwise compared to the array of phase-delay areas of Figure 18. Figure 19 shows the 4 x 4 array of image pixels, such as first image pixel 1950, of Figure 18. Each image pixel is circular or elliptical. The array of image pixels is not rotated by 45 degrees compared to Figure 18 because only the kinoform is rotating. Figure 19 shows that the 1:1 alignment of image pixels to phase zones is disrupted by the rotation of the array of phase zones. The array of phase-delay areas and image pixels are not coordinated in this orientation of the kinoform. With the pattern misaligned relative to the image pixels, there are positions where two spots share a single phase zone (see, for example, the pair of pixels 1960 of Figure 19) and the interaction between the two spots (or pixel positions) is constant for a period of time. The effect of this is similar to that caused by having a turn around time, where the amount of speckle reduction is not constant with time, and the higher speckle at certain times is perceptible by a viewer and degrades the overall image quality. A regular square array of phase zones and a square array of image pixel positions may therefore be sub-optimal for a rotating kinoform. An improvement on the pattern shown in Figures 18 and 19 is achieved by increasing the rotational order of symmetry of the phase zones. For example, a regular array of hexagons (Figure 20) may be used instead of squares. Figure 20 shows a kinoform 2000 arranged for rotation about a centre 2010. An expanded view of a portion 2020 of the kinoform 2000 is shown to illustrate the structure of the kinoform in relation to the image pixels (or, at least, the image pixel positions) of the holographic reconstruction (or “image”). The expanded view shows a 4 x 4 array of phasedelay areas 2030 (“phase zones”) such as first phase-delay area 2040. Each phase-delay area is hexagonal and the phase-delay elements are closed-packed in a regular hexagonal array that will be familiar to the reader. Alternate rows of phase zones are offset in horizontal direction of the figure in accordance with the hexagonal packing arrangement. For example, the image pixels of the even rows are vertically aligned with the gaps between the image pixels of the adjacent, odd rows. Figure 20 shows a corresponding 4x4 array of image pixels, such as first image pixel 2050. Each image pixel is circular or elliptical. Figure 20 shows a 1:1 alignment of image pixels to phase zones. The array of phase-delay areas and image pixels are coordinated in this configuration. The image pixels are therefore also arranged in a hexagonal array. This may be achieved by a variety of different methods. For example, this may be achieved by known methods of modifying the hologram calculation or known methods of interlacing together two different holographic reconstructions using a phase-ramp offset for one of the holographic reconstructions. In relation to Figure 20, the difference between the optimum orientation and the maximum misalignment, occurring at n x 30° (where n is an integer) is reduced, resulting in improved despeckling performance. As explained above, to use such a pattern requires that the arrangement of (image) pixels be adjusted. The distance between adjacent pixels is not changed, and therefore this does not affect image quality, but may be undesirable for other reasons such as having to resample input image data onto the non-square pixel pattern. There is no single shape with higher rotational symmetry than a hexagon which can tile a plane, but semiregular tiling patterns, such as that composed of dodecagons and triangles shown in Figure 21, may be used. Figure 21 shows a kinoform 2100 arranged for rotation about a centre 2110. An expanded view of a portion 2120 of the kinoform 2100 is shown to illustrate the structure of the kinoform in relation to the image pixels (or, at least, the image pixel positions) of the holographic reconstruction (or “image”). The expanded view shows a 4 x 4 array of phasedelay areas 2130 (“phase zones”) such as first phase-delay area 2140. Each phase-delay area has a dodecagonal shape and the phase-delay elements are closed-packed in a regular array similar to that of Figure 20. Alternate rows of phase zones are offset in horizontal direction of the figure in accordance with the close-packing arrangement. For example, the image pixels of the even rows are vertically aligned with the gaps between the image pixels of the adjacent, odd rows. Figure 21 shows a corresponding 4x4 array of image pixels, such as image pixel 2150. Each image pixel is circular or elliptical. Figure 10 also shows a 1:1 alignment of image pixels to phase zones. The array of phase-delay areas and image pixels are coordinated in this configuration. The image pixels are therefore also arranged in a substantially hexagonal array. Again, this may be achieved by a variety of different methods as explained above. Each dodecagon (e.g. first image pixel 2150) has a random phase (more specifically, phase-delay), while each triangle (e.g. first triangle 2160) may have either a random phase (or phase delay) or all triangles may share the same phase (or phase delay). In an embodiment, the irregular pattern of Figure 22 achieves improved despeckling performance. In overview, the radial zones are each separated by the image pixel pitch of the pixel pattern in the radial direction. Each radial zone is further split up into zones circumferentially, each of which has a minimum edge length equal to the image pixel pitch of the content in the circumferential direction. Because it is not necessarily the case that an integer number of circumferential zones will fit into a given radial zone, either the circumferential length of each phase zone may be increased by a small amount (embodiment of Figure 22) or a single circumferentially larger phase zone may be included within the pattern to accommodate the extra length (embodiment of Figure 23). Figure 22 shows an embodiment comprising a kinoform 2200 arranged for rotation about a centre 2210. The kinoform is divided into a plurality of concentric zones, such as a first concentric zone 2220 and a second concentric zones 2230, that are concentric about the centre 2210 of the kinoform 2200. Each concentric zone is sub-divided into a plurality of circumferential zones. Each circumferential zone has a respective phase-delay that is different to all its neighbours. Each circumferential zone is therefore a phase-delay area as per the examples described above. Figure 22 shows a plurality of phase-delay areas of the first concentric zone 2220 such as first phase-delay area 2240. In an embodiment, all phase-delay areas of all concentric zones have the same first dimension, r, in the radial dimension and the same second dimension, c, in the circumferential direction. In the embodiment of Figure 22, there is an exception which is the second phase-delay area 2250 of the first concentric zone 2220. The second phase-delay area 2250 has the same first dimension, r, of the other phase-delay areas of the first concentric zone 2220. However, the second phase-delay area 2250 has a longer second dimension, c + A, where A is less than c. In an embodiment, there is one larger circumferential zone in each radial zone. Each circumferential zone has a different pseudo-random phase-delay - that is, different to at least all its immediate (e.g. adjoining) neighbours. Figure 22 also shows a subset of the pixel positions 2260 of the corresponding (light) pixel pattern of the holographic reconstruction. The pixel positions of the (light) pixel pattern have a first periodicity (radial image pixel pitch) of r2 in the radial direction and a second periodicity (circumferential image pixel pitch) of c2 in the circumferential direction. In an embodiment, r = r2 and c = c2. In an embodiment, A <c and n x c + A = A, wherein A is an integer and n is the number of circumferential zones of the corresponding radial zone. Figure 23 shows another embodiment comprising a kinoform 2300 arranged for rotation about a centre 2310. The kinoform is divided into a plurality of concentric zones, such as a first concentric zone 2320 and a second concentric zones 2330, that are concentric about the centre 2310 of the kinoform 2300. Each concentric zone is sub-divided into a plurality of circumferential zones. Each circumferential zone has a respective phase-delay that is different to all its neighbours. Each circumferential zone is therefore a phase-delay area as per the examples described above. Figure 23 shows a plurality of phase-delay areas of the first concentric zone 2320 such as first phase-delay area 2350. In an embodiment, all phase-delay areas of all concentric zones have the same first dimension, r, in the radial dimension. In this embodiment, the phase-delay areas of the first concentric zone 2320 have a second dimension, c + 8, in the circumferential direction. The phase-delay areas of the other concentric zones may have a second dimension ofc in the circumferential direction. The second phase-delay areas, such as first phase-delay area 2350, have a longer second dimension, c + 8, where 8 is much less than c. Each circumferential zone has a different pseudo-random phase-delay - that is, different to at least all its immediate (e.g. adjoining) neighbours. Figure 23 also shows a subset of the pixel positions 2360 of the corresponding (light) pixel pattern of the holographic reconstruction. The pixel positions of the (light) pixel pattern have a first periodicity (radial image pixel pitch) of r2 in the radial direction and a second periodicity (circumferential image pixel pitch) of c2 in the circumferential direction. In this embodiment, r = r2 and c = c2. In an embodiment, 8 « c and n x (c + 5) = B, wherein B is an integer and n is the number of circumferential zones of the corresponding radial zone. In this embodiment, all circumferential phase zones in a given radial phase zone are the same size. The kinoform I despeckling component of the present disclosure may be made of glass and manufactured in a direct write photolithography and dry etch process, for example. The kinoform I despeckling component may be made of polymer on glass and the different phase zones manufactured in the polymer via a nanoimprint process, for example. Additional features The methods and processes described herein may be embodied on a computer-readable medium. The term “computer-readable medium” includes a medium arranged to store data temporarily or permanently such as random-access memory (RAM), read-only memory (ROM), buffer memory, flash memory, and cache memory. The term “computer-readable medium” shall also be taken to include any medium, or combination of multiple media, that is capable of storing instructions for execution by a machine such that the instructions, when executed by one or more processors, cause the machine to perform any one or more of the methodologies described herein, in whole or in part. The term "computer-readable medium” also encompasses cloud-based storage systems. The term “computer-readable medium” includes, but is not limited to, one or more tangible and non-transitory data repositories (e.g., data volumes) in the example form of a solid-state memory chip, an optical disc, a magnetic disc, or any suitable combination thereof. In some example embodiments, the instructions for execution may be communicated by a carrier medium. Examples of such a carrier medium include a transient medium (e.g., a propagating signal that communicates instructions). It will be apparent to those skilled in the art that various modifications and variations can be made without departing from the scope of the appended claims. The present disclosure covers all modifications and variations within the scope of the appended claims and their equivalents.

Claims

1. A kinoform arranged for rotation, wherein the kinoform is divided into a plurality of concentric zones, each concentric zone sub-divided into a plurality of phase-delay areas, wherein the number of phase-delay areas per concentric zone increases with the radius of the concentric zone and adjoining phase-delay areas have different phase-delays.

2. A kinoform as claimed in claim 1 wherein the concentric zones have substantially uniform size in the radial direction.

3. A kinoform as claimed in any preceding claim wherein the phase-delay areas have substantially uniform size in the circumferential direction.

4. A kinoform as claimed in any preceding claim comprising 5 to 15 concentric zones such as 8 to 12 concentric zones.

5. A kinform as claimed in any preceding claim wherein the smallest concentric zone comprises 5 to 10 phase-delay areas such as 8 to 12 phase-delay areas.

6. A kinoform as claimed in any preceding claim wherein the phase-delay of each phase-delay area is 0 to 2k.

7. A kinoform as claimed in any preceding claim wherein each phase-delay area is defined by two straight sides, each of which is part of a radial line, and two curved sides, each of which is part of a boundary of a concentric zone.

8. A kinoform as claimed in any preceding claim further comprising a dead zone at the centre of the concentric zones.

9. A kinoform as claimed in any preceding claim wherein the size of each phase-delay area in the radial direction is substantially equal to its size in the circumferential direction.

10. A kinoform as claimed in claim 9 with the exception that one phase-delay area of one concentric zone is longer than the others in the circumferential direction.

11. A kinoform as claimed in claim 9 with the exception that the phase-delay areas of one concentric zone are longer than those of the other concentric zones in the circumferential direction.5 12. A kinoform as claimed in any preceding claim wherein the phase-delays of thekinoform are pseudo random.

13. A method of improving the perceived quality of a holographic reconstruction, the method comprising:10 rotating a kinoform, wherein the kinoform is divided into a plurality of concentriczones, each sub-divided into a plurality of phase-delay areas, wherein the number of phasedelay areas per concentric zone increases with the radius of the concentric zone and adjoining phase-delay areas have different phase-delays; andilluminating a hologram to form a holographic reconstruction on the kinoform, wherein15 each pixel of the holographic reconstruction is aligned with a respective phase-delay area of the kinoform.

Images