WO1996021992A2

WO1996021992A2 - Method for increased colour reproduction

Info

Publication number: WO1996021992A2
Application number: PCT/SE1996/000018
Authority: WO
Inventors: Stig Berglund
Original assignee: Optica Nova Onab Ab
Priority date: 1995-01-14
Filing date: 1996-01-12
Publication date: 1996-07-18
Also published as: WO1996021992A3; WO1996021883A2; WO1996021883A3; WO1996021996A2; WO1996021997A3; WO1996021997A2; WO1996021995A2; WO1996021995A3; WO1996021996A3

Abstract

A method for projection for reproduction of electronically registered pictures from a first gamut (Gc), representing the registered picture, to a second gamut (Gr⊥Gd), which mainly is an intersection of a fourth gamut (Gd) with a third gamut (Gr), where the fourth gamut (Gd) represents a gamut for a projector arrangement and the third gamut (Gr) represents the gamut for realistic sceneries, which is larger or smaller than the first gamut (Gc) for the registered picture, where a colour in the registered picture with the colour coordinates (xc, yc) in the projector arrangement is reproduced with expanded or contracted colour coordinates (xr, yr), which point mainly lies in the second gamut (Gr⊥Gd).

Description

TITLE: Method for increased colour reproduction

TECHNICAL FIELD:

The invention relates primarily to a method of the art described in the preamble of the enclosed Claims 1 and 8.

According to one aspect of the invention is thus aiming at a method for projection with reproduction of electronically registered images from a first gamut, representing the registered image, to a second gamut, mainly representing an intersection between a fourth gamut and a third gamut, where the fourth gamut represents a gamut to a projection arrangement and the third gamut represents realistic sceneries which may be larger or smaller than the first gamut.

BACKGROUND OF THE INVENTION:

There are already several projection arrangements for projection of video pictures on LCDs on the market. These projection arrangements, however, have not been very successful, due to several inherent drawbacks.

One of the drawbacks of the known projectors is that they offer deficient viewing ergonomics if, for instance, the projector is placed in a room at a position corresponding to half the width of a vertical wall onto which the picture is to be projected, and the projector furthermore is situated at a level corresponding to the viewer's eye level. In that case the projected picture's centre will be situated at the corresponding height and width on the vertical projection surface, so that the projector or the viewer will obstruct the picture. It would therefore be desirable to offer a projection construction that provides a parallel shifting of the projected picture, so that the projector can be placed next to the ceiling, the floor and/or on a side wall. The illumination arrangements have a common focal area, which in the orthogonal projection towards the respective image generating device primarily is situated in the centre of the respective image generating device (IGDs), so that said deficient viewing ergonomics arise.

Yet another limitation of today's TV projectors is that they only allow projection of two-dimensional pictures. There are methods known which provide projection by alternately showing a right and a left registered picture and at the same time mechanically switching the polarisation filters between two orthogonal polarisation directions, whereby the viewer is able to see three- dimensional pictures through the use of polarised glasses having orthogonally polarised glass. The use of mechanical devices imply considerable disadvantages, such as disturbing noises, and increased power consumption, volume and weight.

It is, of course, possible for persons who normally wear glasses while watching TV, to have their glasses for the right and left eye, respectively, designed so that they render a, between themselves, orthogonal polarisation. It is then advisable to not choose horizontal, respectively vertical polarisation neither in the projector nor in the glasses, which polarisation, due to reflection, can produce stress for the left and the right eye while watching natural sceneries.

A serious limitation of today's projection arrangements is the imperfect colour reproduction. The area of the colour diagram that is defined by the three primary colours is not sufficient for a good reproduction of colours — especially in nature.

A method for improvement of the imperfect colour rendering in hard copies is described in the US patent 5,317,426 by a method transmitting a gamut for the hard copy to an expanded gamut for a cathode ray tube. The US patent 5,317,426 describes, however, not how this can be applied to moving images, such as cine films, computers and similar media. Neither is there description of how the method could be applied to others than for cathode ray tubes.

By using laser or laser diodes as illumination sources, there is a risk during projection that very disturbing speckles in the picture are generated. Existing solutions have considerable disadvantages, which either cause the necessity to use mechanically movable components — for example a rotating diffuser next to the object plane, or that a spectral broadening of the light takes place, as is the case during modulation of the laser illumination source. In the case of the rotating diffuser it is also difficult to produce enough speckle reduction without impairing the resolution of the picture due to the fact that the diffusor cannot be placed sufficiently close to the object.

Since there are frames around the pixels in the IGD, there is the disadvantage of a visible pixel structure in the picture.

On the market today there are many types of projection screens, which are partly for use with spectrally broad- banded light and partly of the principally, in relation to the screen, centrally illuminated type.

In the case where the projector is strongly shifted to one side in relation to the projection screen, special demands are set up as to how the screen should reflect the light to the viewer's eyes.

In the case where the projection screen is illuminated with light in a number of narrow bands, there are today no narrow-banded illumination sources which are bright enough to obtain sufficient contrast in the picture at normal in¬ door lighting. This problem is especially relevant to a new type of display, i.e. the DMD, Digital Micromirror Device, which is described in the article "The Digital Micromirror Device (DMD) and its transition to HDTV" by the authors J.M. Younce and D.W. Monk in "The 13^th International Display Research Conference", Strasbourg, August 31- September 3, 1993, pp 613-616. It is hard to, for practical reasons, in the DMD create colour projection by implementing room-multiplexing, i.e. to generate picture information in spatially separated parts in the display unit(s). One is then forced to use time-multiplexing, which means that the same parts of the display unit are used for reproduction of the different colours. The problem is now solved by using a rotating colour filter which is transilluminated by a lamp. This is, considering the light efficiency, a highly unfavourable solution. By instead implementing narrow-banded, fast-switching illumination sources, such as laser diodes, one would obtain a less complex and a more efficient arrangement.

Today's projectors with only one projection lens can generally be divided into two main categories. In the first category there are several IGDs which are each submitted to transillumination by one-coloured light with a different colour light for each IGD. The light from the different IGDs is then linked together via a dichroical linking device consisting of dichroical mirrors, for example in the form of dichroical prism.

In the second category there is only one IGD, transilluminated by a broader spectrum. There are then different colour filters in the adjacent image elements which produce colour pictures. This second category therefore has the great disadvantage of considerable loss of light through the filtration of the image elements. This process also produces a heating of the IGDs, which can be substantial because of the often large luminous flux. Later on we intend to show designs of projectors in this category that do not have the mentioned disadvantage of light-loss in the colour filters. We will henceforth mainly limit this discourse to projectors in the second category — even though, as is stated below, it is applicable even on projectors in the first category.

SUMMARY OF THE INVENTION:

A general purpose of the present invention is to achieve an improved colour rendering when projecting moving pictures, recorded with a limited gamut, e.g. by the three primary colours.

The purpose of various aspects of the invention is obtained through the designs of the projection arrangements which are stated in the appended independent claims.

The further purposes with other aspects of the invention are obtained through the designs of the projection arrangement that are stated in the appended dependent claims.

An aspect of the invention is based on a projection arrangement which includes a number of IGDs, for instance LCDs, or more generally SLMs (Spatial Light Modulators), for off-axis projection on a screen with one projection lens, which are illuminated by several, for instance, three differently coloured light sources that have possibly been obtained through colour separation of light from a single light source, a shared projection lens in the ray path after the out-put of the IGD, and a focusing illuminating optical system whereby the projection lens is situated in or at the shared focal area. In light of the above stated facts, one aspect of the invented projection arrangement is mainly characterised in that the illumination arrangements' said shared focal area is situated, so that its optical orthogonal projection against the plane in which the IGD is located is substantially shifted from the centre of the IGD, whereby the projected picture obtains a parallel displacement, thus rendering the desired ergonomical advantage. Laser diodes are the for this purpose best suited illumination sources because of their narrow spectral band width. A particular advantage with such an arrangement is that the projection picture can be made particularly light-intensive and that a good function and viewing ergonomy are provided.

According to a first embodiment according to one aspect of the invention, light is directed at different angles of incidence from the various illumination sources against the IGD. The light first hits a positive microlens matrix with a positive lens for each of the different colour groups which together form an image spot and that focus the various colours towards separate negative tilted microlenses or microprisms, which focus the light through the middle of the respective colour pixel elements. A tilted microlens is defined as a microlens that is deflecting a through the middle of the lens passing light ray, so that the deflection is different from the deflection that a light ray should have if the tangential surface at the middle of the lens were parallel to the plane of the microlens matrix itself. At the exit of the IGD there are matrices of microlenses, microprisms and/or diffractive optical elements (DOEs) to focus the light into the entrance pupil of the projection lens according to one aspect of the invention's principle. The first matrix on the out-put side consists of positive microlenses collimating the from the pixel apertures coming light. The second matrix consists of microprisms deflecting the light into the middle of the entrance pupil of the projection lens. The two microlens matrices on the entrance side can be placed on either side of a sheet of glass. The microprisms are having the same purpose as a very compact colour corrected field lens. The second matrix with tilted microlenses on the in-put side may very well be affixed directly onto the substrate on the in-put side of the SLM. This is especially the case if the microlenses or microprisms are produced for example by etching. Techniques for etching of microlenses on glass exist. The same is true also for the first matrix on the out-put side of the SLM. The latter matrix can moreover be the only one on the out-put side if the lens and prism operation is integrated into one and the same matrix with refractive or diffractive elements.

The above stated technologies of using microlens matrices is also applicable on the projectors in the first category by instead using a positive microlens on the entrance side for each pixel.

According to a second embodiment, the microlens array principle is combined with a field lens, which focuses the light on the out-put side of the IGD according to one of the invention's principles. It can be difficult to achieve a satisfactory colour correction in the field lens as it has to bring about big changes in the field angles. It is therefore advantageous to compensate the colour aberrations in the field lens with the prismatic elements in the microprism array.

One of the fundamental principles in achieving the purposes of one aspect of the invention is using time multiplexing possibly in combination with spatial multiplexing by, in one and the same IGD, using pixels dedicated to different colours - either by using a colour separation technology or by allowing light to be blocked in cells that do not have the colour for which the cell is dedicated. The term time multiplexing refers to the fact that vide information is divided up in time so that one picture i sent to the IGD in multiple in time separated partia pictures, where each of these partial pictures reproduc the picture information for a number of colours.

The ordinary colour systems for television are, fo different, partly historical, reasons not very efficient This is partly due to having to work with the at the tim existing phosphors, which left large parts of the colou diagram uncovered. In particular large parts of the green more saturated colours are poorly reproduced. It i therefore a special purpose of one aspect of the inventio to provide an improved colour reproduction. According t the C.I.E. colour system from 1931 are the current colou co-ordinates according to both the European system EBU an the American NTSC, drawn up in a diagram, evidentl completely insufficient for all saturated colours an especially in green, where the colour tolerance, however is bigger. Even with the transition to monochromati illumination sources, which are known to represent maxima colour saturation and thus are on the borderline of th colour diagram according to C.I.E., can a great improvemen be achieved. This is already possible using today's lase sources, which within the near future will be improved an less expensive. Much of the documentation is, however, already registered with the old basic colours, and on would like to reproduce this documentation wit compatibility. There is the possibility to extrapolate b elongating the vector from the white point in the diagra in order to reproduce colours, which according to the registered values are close to the limiting triangle, wit points that are outside.

The improvement that is possible by using monochromatic colours, can according to one aspect of the invention's principle, be improved further by using more than three colours. According to the principle for metamerics are the sense of seeing and the brain thus constituted that the brain cannot make out the difference between two stimuli, even if they are physically different, but sees the same colour nuance although the spectral division is different, and it is therefore possible to reproduce one single, subjective impression in many different ways. Every imaginable colour impression is theoretically defined to the colour (consequently not considering the intensity) of a point in a two-dimensional colour diagram, because of which two parameters are enough for the colour information. Ordinary television, of course, transfers information for each pixel through one intensity parameter and two colour parameters. From the ordinary colour diagram, however, it is clear that it is impossible to cover all observable colours with three colours, no matter how they are chosen, and it is therefore preferable to use monochromatic colour sources and to work with more than three, preferably four, five or six, different colours.

A disadvantage with the known projectors is then that they have a severely limited colour dynamics (gamut) . This largely depends on the fact that only three colours are used to generate the picture wherefore it is impossible to with good efficiency reproduce colours of down to 400 nm and up to 700 nm. Should one expand the number of colours in a projector in the first category, one would have to, using today's technology, expand the number of IGDs with as many as the number of colours, thus making this kind of projector even more expensive. Should one expand the number of colours in a projector in the second category, one would, using the existing technology, worsen the already poor light efficiency.

In accordance with a first embodiment, according to one aspect of the invention, according to the second category the one IGD is illuminated with different time multiplexed illumination sources of different colours. The illumination sources are directed in correlation with the picture information to the IGD so that the information of a certain colour is generated via the IGD when said colour illuminates the IGD. One hereby obtains the possibility to with very few pixels generate pictures with a very good colour reproduction.

In accordance with a second embodiment, according to one aspect of the invention, according to the second category the one IGD is used for both spatial and time multiplexing.

The time multiplexing is generated in the manner stated above, while the spatial multiplexing is generated by colour separation implementing microlens matrices according to the above stated two principles. It is in this way possible to produce a picture using, for instance, a gamut between six primary colours with, for example, three colour pixels which each are transilluminated by two colours that are close in the spectrum and are sequentially alternated in time.

It is possible to utilise a gamut in a display which is larger than the gamut that is rendered by the camera or the equivalent received video signal. According to one aspect of the invention this is achieved in the following manner.

One first decides a second gamut G_rDG_d, which mainly is the intersection between a fourth gamut Gd and a third gamut Gr, where the fourth gamut Gd represents a gamut for a projection arrangement and the third gamut Gr represents the gamut for realistic sceneries which is larger or smaller than the first gamut Gc for the registered picture of the received signal's gamut. All three gamut areas are supposed to have a mainly shared achromatic point x„ and y_n. The received signal supposedly has the colour co-ordinates x_c and y_c located within G_c. The colour co-ordinates for the received signal's dominant colour is calculated as the point where a line through the points x_n,y_n and x_c,y_c crosses the edge of G_r and is designated x_b and y_b.

A colour in the registered picture with the colour co- ordinates x_c,y_c in the projection arrangement being represented by expanded or contracted colour co-ordinates x_r,y_r of which point mainly lies within the second gamut (G_rnG_d) .

The point x_r,y_r is chosen with a constant nuance and chroma, which is adopted to the relative distance from a white point x_n,y_n to the edge of the second gamut G_rDG_d in relation to a distance from the white point x_n,y„ to the edge of the first gamut Gc along a line through the colour co-ordinates x_c,y_c and the white point x_n,y_n.

The intersection point x_bc,y_bc is determined between the edge of the first gamut Gc and a line passing through the point x_n,y_n which is common for all gamuts G_d,G_r,G_c and the point x_c y_c for the registered picture.

The intersection point x_b,y_b is determined between the edge of the intersection G_rf.G_d, between the fourth gamut Gd and the third gamut Gr, and the line passing through the white point x_n,y_n, which is common for all gamuts G_d,G_r,G_c and the point x_c,y_c for the registered picture.

The colour co-ordinates _bc,y_bc ^{ror tne} point of the edge of G_c which makes up the point of intersection between the edge of G_c and a line through the achromatic point x^y,. in the colour diagram is calculated. The colour's excitation potential p_rβ is determined as:

P_re = l+k_rβ*((x_b-x_n)/(x_bc-x_n)-l) (1) or p_rβ = l+k_r._*((y_b-y_n)/(y_bc-y_n)-l) (2) where k_re is chosen in the interval {0,1}

When k_re which is from now on called the colour expansion factor, equals 0, the colour reproduction is the same in the projector and in the by the camera registered image. When kre equals 1 there is maximum colour expansion. The possibility of choosing the k_re within said interval should be given in the projector's control unit.

A finite number of intersection points Xi,yι is determined between the lines, that are delimiting the first gamut G_c and the line passing through the point x_n,y_n, which is common for all gamuts G_c/ G_r,G_d, and the point x_c/y_c for the registered picture.

New, to the intersection G_rnG_d> between the gamuts G_d and G_r expanded colour co-ordinates x_r, y_r are calculated as :

X_r = P_rβ*( Xc-Xn) +Xn ( 3 )

If the number of primary colours in the display is larger than three, then the relative light intensities are not distinctly determined but can be chosen freely from a metametrical area with combinations of primary colours that render the same colour co-ordinates x_r,y_r. It is therefore possible to divide the by the different colours generated gamut into a number of non-overlapping triangles, which are limited by the points of the different light sources in the colour diagram. Every point in the picture can thereby, at a certain point in time, be represented by light from a maximum of all three light sources which generate the triangle in which the point's colour co-ordinates, or expanded colour co-ordinates, are located. Generally speaking, will the number of triangles equal the number of different-coloured illumination sources minus two.

In general, at least three light sources can be used in order to represent a given point pO with the coordinates (xO,yO) in the colour diagram. In principle, the luminous flux ΦO in lumen is maximised for the point pO (or the inverse of the luminous flux is minimised for said point) .

m ΦO = 683*Σ a_i*S_i*y_i lumen (5) i=l

with the constraints:

n = 0 (6)

0 < a_i < l, i = l,n (8)

a_£ = 1, i = n+l,m (9) where: m is the total number of light sources, ra > n;

n is the number of light sources to be used to represent a certain colour, n > 3;

Si, i = l,n are the powers of the light sources in Watt;

(x_i,y_i,z_i), i = l,n are the tristimulus values for the n laser diodes;

a_t, i = l,m indicates the maximal transmission for the different light sources and for point the point PC in the colour diagram.

The luminous flux in a pixel or group of colour pixels is then determined from a product of the luminance value of the video signal and the vector a_t ,i = l,m.

G_c is a convex polygon where the corners are given by the primary colour co-ordinates, which are determined by the camera. This facilitates the calculation of the point X_bc Y_bc which simply can be calculated as the intersection of two straight lines. To facilitate the calculation of the point x_b,y_b it could be suitable to approximate the edge of G_rDG_d with a number of straight lines, which together form a polygon G_rr. The point x_bc,y_bc can formally be calculated as:

Min(Abs(x_i-x_n))-Sgn(x_i-x_n)+x_n,Min(Abs(y_i-y_n))-Sgn(y_i-y_n)+y_n (10) for all i = l,n where Xi,yι for i = 1 to n, are intersection points between the n lines which define the area G_c and the line which passes through the points x_n,y_n and x_c,y_c- Abs designates the absolute value and Sgn designates the sign (+ or -1).

The point x_b y_b can be calculated analogously.

The above described colour adjustment could briefly be described as the point x_r,y_r chosen so the shade of the colour is maintained whilst the colour saturation can be adapted to the relative distance from the point x_c,y_c to the edge of G_rilG_d in relation to the distance from the point x_c Y_c to the edge of G_c along a line through the point x_c,y₍c and a white point x_n,y_n. In this case C.I.E. colour co¬ ordinates from 1931 is used. Naturally it is possible to apply the same principles on other colour co-ordinate system, such as C.I.E. co-ordinate system C.I.E.L.A.B. (L*,a*,b*) and C.I.E.L.U.V. (L*,u*,v*) from 1976 where the co-ordinates a*,b* and u*,v* respectively is used instead of the co-ordinates x,y as shown above, whilst L* is mainly held constant. L* is normally denoted as lightness index or psychometric lightness.

According to one aspect of the invention the visual impression of speckles in the picture is eliminated by colours. In a preferred design is the screen shaped as a diffuser with a light-absorbing coating on which the thin- film coating is applied. The irregular or diffuse surface is shaped so that its Fourier transform mainly renders a significant spectrum with considerably higher frequencies the frequencies which are obtained at Fourier transformation of the on the picture projected pixel structure. It is obviously of great significance that the light which reaches the screen is reflected as efficiently as possible before reaching the eyes of the viewers. This is possible to achieve by designing the screen with microfacets or DOEs onto which a light-absorbing coating is applied, together with the above mentioned filter which mainly reflects the narrow light bands from the projector. The microfacets or the DOEs are designed so that light is reflected back within an angular area in which the viewer's eyes are likely to be. The screen will, of course, not have uniform qualities, since the angle of incidence of the light varies across the screen, and the viewer's eyes will be directed in different angular positions in relation to different parts of the screen. Within each pixel on the screen, there are a number of facets with both vertical and horizontal randomly tilted angles. The distributions of the tilted angles is calculated in consideration to incidence and the desired angular area for reflected light, and the number of mirror elements is chosen in consideration to the diffraction, so that an even distribution of light is obtained. It would be possible to manufacture the screen with the help of embossing using moulds generated with the help of electron beam lithography or laser scanning. The diffusor can very well be modelled with micro-facets according to the above described principle. It is of course possible that the screen is manufactured with its irregular surface in a light absorbing material, at which the coating with the light absorbing layer can be omitted. When the projector is positioned asymmetrically in relation to the projection screen, then the screen itself will be asymmetric with regard to the light spread distribution. This is for example the case when the projector is placed in connection to a normal to the screen emerging from a corner of the screen. To be able to use the one and the same screen with the projector placed in connection to each one of the corners of the screen, it is possible to model the screen such that it is possible to turn the screen upside down to be able to place the screen in connection to the diametrically opposite corner. A screen structure for the remaining two corners can be applied on the other side of the screen.

The principles for the different aspects of the invention in the above descriptions are in no way limited to only projectors in the second category, but are completely applicable on projectors also in the first category. Thus, for instance, a six-colour projector can be designed with three IGDs, where each IGD is illuminated by two colours, and where colour separation, on the analogy of the one shown in figure 5, is used. The advantage of not having to resort to time multiplexing, makes it possible to chose low power illumination sources and also results in a better colour saturation, since the illumination sources shine without interruption. The latter can be of significance to the life of the illumination sources, since intermittent light in all probability cause a faster wear. In the same way it is also possible to design a four-colour projector with two IGDs and two illumination sources per IGD.

Different aspects of the invention will below be described as example cases with reference to the appended figures.

BRIEF DESCRIPTION OF THE DRAWINGS:

Fig. 1 schematically depicts a projector for 3-D projection. Figs. 2 to 4 schematically depict different alternative positions of the polarisation-twisting window in 3-D projectors.

Fig. 5 shows the principle for a colour separation technology in an LCD-projector.

Fig. 6 schematically illustrates how one expands the of the camera given gamut to a realistic gamut.

Fig. 7 schematically illustrates how one expands the of the camera given gamut to a, with a polygon approximated, realistic gamut.

Fig. 8 schematically depicts the make of a IGD with a despeckelator / depixelator.

Fig. 9 shows a cross section of one part of the projection screen.

Fig. 10 shows in a diagram how the diffuse reflectance, with the help of thin-film technology, is designed in the case of narrow-banded illumination sources.

Fig. 11 schematically shows the design of a reflecting screen with microlenses as the foundation for the in figure 9 illustrated coating.

Fig. 12 schematically depicts a six-colour projector where both spatial and time multiplexing are used.

Fig. 13 illustrates an example of the division of a six- colour gamut in four colour triangles.

Fig. 14 schematically illustrates a six-colour projector with two IGDs with three-colour separation in each of the IGDs. Fig. 15 illustrates a preferred manner of choosin polarisation directions at the use of 3-D projectors.

Fig. 16 schematically illustrates the make of a 3-D thre colour projector, where an off-axis field lens is used t direct the light into the projection lens.

Fig. 17 schematically illustrates a detailed study of a IGD with condensor, despeckelator/depixelator and fiel lens.

Figs. 18 to 21 illustrate different placements of th projector in relation to the projection screen.

PREFERRED EMBODIMENTS:

Fig. 1 schematically depicts a projector for 3-D projection. The laser diode light sources 2R, 2G, 2B can beside a number of laser diodes in the colours red, green and blue also include light distributing optics in the form of for example diffractive optical elements (DOE) so that an even illumination of the IGD IK is obtained via the condensor C. The light in the green colour generally incides with an angle W against the IGD IK. The light in the red and the green colour incide with the angles El and E2 in relation with the to the green colour whereby one in the IGD IK obtains the colour separation described below in connection with figure 5. Via the connections V and H to the unit Ul picture information comes in with right and left stereo pictures. From the unit Ul proceeds alternate left and right pictures via II to the IGD IK. At the same time the polarisation alternates 90 degrees in the window LI with liquid crystals on an impulse from Ul via SI. The light from the IGD Kl passes first a polariser Pi and then the polarisation twisting window LI. Consequently the light that proceeds from the IGD with a left picture at the passage through the projection lens 7 on to the screen 11 will have an orthogonal polarisation against the polarisation for a right picture. When then the viewer of the picture wears in the same way orthogonally polarised glasses he will experience a three-dimensional picture. If the IGD already contains a polariser and an analyser, the polariser PI can be left out. If the provided video signal, via VH, already comes in alternating left and right pictures, the unit Ul is simplified to give only the necessary alternation impulses to the polarisation alternating window LI. The projector has here been depicted in projection that could be a side projection. The projector can however be designed so that similar conditions also apply at vertical projection. The light from the illumination sources have been given the average incident angle W onto the IGD IK in order to simplify the focusing of light to the projection lens 7. The angle W can be chosen within the interval 0 to AW degrees, where the angle AW is an average angle for the light that leaves the IGD IK. To further facilitate the focusing of light in the IGD IK to the projection lens 7, the light from each of the light sources 2R, 2G and 2B can be either collimated after the passage through the condensor C, or convergent with angle of convergence CW that is in the interval 0 to OW degrees, where OW is the angle of convergence that the light has when it leaves the IGD IK. In extreme cases where A=AW and the light after passing through the condensor has the angle of convergence OW, can the deflecting function in the matrix 22 in figure 5 be left out. Practically, it is suitable to chose a value for W in the interval of 0 to AW/2 degrees and a value for the angle of convergence for the light from the condensor C to picture generation element IK that is in the interval 0 to OW/2 degrees.

Figs. 2 to 4 schematically illustrate different alternative ways to place the polarisation-twisting window LI in 3-D projectors. The illumination source is designated with L, the IGD with 1, the projection lens with 7 and the projection screen with 11. The arrangement in figure corresponds principally with the one in figure 1, wit placement of the polarisation-twisting window LI betwee the IGD and the projection lens 7. In figure 3 has th polarisation-twisting window LI been placed immediatel after the projection lens 7. Figure 4 represents a case o the illumination source generating polarised light fo which the direction can be altered by a polarisation twisting window between the illumination source L and th IGD 1.

Fig. 5 very schematically depicts a section of the IG according to one aspect of the invention where colou separation is used, i.e. the different colours R, G and are separated in the dedicated light valves 1R, 1G and I in an SLM 1. Green light is here thought of as incidin orthogonally against the IGD, i.e. the angle W in figure is 0, while red light incides with the angle E2 and blu light with the angle El. The light in the different colours first come in to a microlens matrix 322, wit positive microlenses 94 with an effective focal length that is considerably larger than the distance A between the matrices 322 and 323, and at the same time considerably shorter than the distance B between the matrix 322 and the SLM 1. Light in the three different colours hit a negative microlens each, i.e. the microlenses 19R, 19G and 19B in a second microlens matrix 323 and are then focused to a light valve each, i.e. the valves 1R, 1G and IB in an LCD matrix 1, whereby light in the three different colours hit the light-directing, possibly with diffractive structures colour corrected combinations 7R, 7G and 7B of microlenses and microprisms in the matrices 21 and 22. The microlenses in the matrix 21 have an effective focal length equivalent to the distance between the LCD matrix 1 and the matrix 21. The negative microlens 19G can be an ordinary on-axis lens. It is however, especially considering the diffraction and its influence on the contrast, necessary that the light is focused into the middle of all the light valves 1R,1G,1B in the SLM 1. Therefore the lenses 19R and 19B can be made out of a combination of prismatic parts and negative lenses, i.e. tilted negative lenses. In the general case, when the angle W in figure 1 is not zero, all the negative lenses in matrix 323 are tilted. It is certainly possible to shape the lenses 94 as well as the lenses 19R, 19G and 19B, as diffractive optical elements. The two matrices 21 and 22 can at advantage be combined into one matrix, either as refractive elements, which is especially suitable when the illumination sources have a spectrally narrow band width, or as diffractive optical elements . In the case of the light valves 1R, 1G and IB being dispersive, for instance of the PDLCD-type, there is no need for a polariser or an analyser. The light transmission will hereby be three times as efficient.

In this figure a possible arrangement is shown, where negative microlenses are used in the second matrix. It is of course possible to model this matrix with only prismatic elements, at which the focal length of the positive microlenses in the first matrix mainly will equal to the optical distance to the pixel apertures to which the light from the microlens in question shall be linked with help of the prismatic elements in the second matrix.

Fig. 6 schematically illustrates how one expands the of the camera given gamut G_c into a realistic gamut G_r during reproduction with a 6-colour display with a fourth gamut G_d. CIE designates the CIE colour diagram. x and y designate the co-ordinates in the CIE. A colour c in the first gamut G_c of the camera is expanded to the colour r in the realistic gamut G_r. Naturally, the first gamut G_c can be contracted instead of being expanded where this may be applicable.

Fig. 7 schematically depicts how one expands the of the camera given gamut G_c into a with a polygon approximated fourth realistic gamut G_rr. A colour c in the camera's G_c is expanded to the colour r in the with a polygon approximated realistic gamut G_rr. Naturally, the first gamut G_c can be contracted instead of being expanded where this may be applicable.

Fig. 8 schematically illustrates the design of an IGD with a despeckelator/depixelator. The light valves in the IGD are designated with IV, the from figure 5 integrated matrices 21 and 22 are designated with MM, while the despeckelator window is designated with 88 where a light distribution with the standard deviation of DV is obtained. When the light distribution is varied by putting an alternating voltage device VA of suitable amplitude and sufficient frequency, and which is superimposed on a DC device VD (which is needed for minimal dispersion) on the despeckelator window 88, then will the phase of the light which reaches the viewer's eyes vary in such a way that the visible impression of a speckle pattern is integrated out. The angle DV is chosen so that the standard deviation of the phase changes amounts to at least 90 degrees. The frequency of the phase changes should amount to at least 60 Hz. One can easily see that the despeckelator window also will broaden the visual impression of the pixel aperture and will also therefore function as a depixelator. Because the light from different parts of the IGD incides with different angles on the window 88, it would be suitable to divide it into different segments with between themselves varying values on both the direct current component VD and the alternating current component VA.

Fig. 9 shows a cross section of a part of the projection screen 11, where the filter layer is designated with 41, the diffuser layer, which can be made from microfacets is designated with 42 and the foundation layer is designated with 43. Fig. 10 shows how the diffuse reflection factor for the screen is to be optimally designed, using thin-film technology. The axis r is graduated in a diffuse reflection factor while the axis 1 is graduated in wavelength, with the wavelengths in the illumination wavelengths R, G, B marked. The curve 8 shows the diffuse reflection factor, which is obtained by, for example, coating a diffuse light-absorbing surface with a thin-film filter which transmits between the colours B and G, respectively G and R. The curve 8 could therefore also show the reflection factor of the thin-film filter, where the axis r is graduated in a reflection factor. Because the main part of the surrounding light which hits the screen, and which is located between the colours B and G, respective G and R, will be absorbed, the contrast of the picture will be considerably greater. It is, however, also possible to use this effect to reduce the brightness of the illumination sources in the projector, whereby important energy savings are obtained. Generally speaking, can the transmission curve 8 be designed so that mainly only the projected light wavelengths are reflected, while light with wavelengths outside of these mainly are absorbed by the screen. In order to obtain an obvious effect, should the relation between the reflection factor at the illumination wavelengths R, G and B, and the reflection factor for any wavelength between the illumination wavelengths R, G and B be larger than 1,4. If the spectral band width of the illumination sources is relatively small, it is possible to obtain an obvious improvement of contrast through the absorption of a considerable amount of the surrounding light into the projection screen 11.

Fig. 11 shows schematically the principle for design of a projection screen with microfacets illustrated. The facets 45, which have been covered with an absorbing layer and then with a reflecting layer as illustrated in figure 9, basically cover a pixel on the screen and reflect light in such a way that all viewers are able to see this pixel. Light which is reflected against mirror 45a obtains a different direction from the light which is reflected against mirror 45b. The number of mirrors and the distribution of angles is calculated with the diffraction in consideration, so that an even distribution of light is obtained to all presumptive viewers' eyes.

Fig. 12 schematically illustrates a six-colour projector where both time and spatial multiplexing are being used. The projector is mainly designed as the projector described in figure 1 , with the following additions:

The six illumination sources with the appurtenant light distributing optics are grouped in three groups: 2r, 2R; 2g, 2G and 2b, 2B. In a preferred design are the illumination sources in each group chosen so that they are spectrally close. Light in each group is linked together with the help of dichroical mirrors 5a, 5b and 5c, so that the light in each group will incide towards the condensor C with the same angle relations. The video signal for each picture from the unit Ul is in the unit CU divided into two pictures with information respectively for the colours in the illumination sources 2R,2G,2B and 2r,2g,2b, so that when a picture changes from one colour group to an other are the illumination sources activated or deactivated via the connections 3 or 4. The division of colours in the different illumination sources is for the main part of the, by the six colours defined, gamut not entirely unique, i. e. there are a great number of combinations of the six light colours which are represented by one single point in the colour diagram. It is therefore possible to, as shown in figure 13, divide the, by the six colours defined, gamut into four non-overlapping triangles Gl, G2, G3 and G4, which are limited by the six illumination sources' 2R, 2G, 2B, 2r, 2g and 2b points 3A, 3a, 3B, 3b, 3C and 3c in the C.I.E. colour diagram CIE. x and y are the designations for the co-ordinates in the C.I.E. colour diagram CIE. Every point, for instance CP, in the picture will thereby, at a certain point in time, uniquely be determined by light from a maximum of all three illumination sources which define the triangle, for instance G2, in which the colour co-ordinates or the expanded colour co-ordinates of the point are located. In general will the number of triangles equal the number of different-coloured illumination sources minus two. The luminous flux ΦO for CP maximises or minimises the inverse in the same manner as previously described.

Fig. 14 schematically depicts a six-colour projector with two IGDs and with three-colour separation in each of the IGDs. It is, from the illumination sources 2R, 2G, 2B and 2r, 2g, 2b up to the IGDs IK and 2K respectively, designed in the same way as the device in figure 1, with the difference that the condensor is shown as if its optical axis were coinciding with the normal of the midpoint of the IGD. The light spectrum for the light sources in the groups 2R,2G,2B and 2r,2g,2b are mainly disjoint. The light from the IGD IK is linked via the dichroic prism DS into a device on the whole equivalent to the one described in figure 1. On the output of the unit Ul is a unit CE, where colour expansion is performed in the case of the , via V and H or VH, incoming signal to the unit Ul has been registered with a smaller gamut than the projectors potential gamut. The unit U2, which divides picture information from the unit CE into two partial images in consideration to the colour contents, one via Jl for the illumination sources 2r,2g,2b and one via J2 for the illumination sources 2R,2G,2B. CC designates schematically a connection on the unit CE, by means of which the value of the colour expansion factor k_rβ can be determined. Fig. 15 demonstrates a preferred manner of choosing polarisation directions at the use of 3-D projectors. RE designates the right eye, and LE designates the left eye in a pair of glasses. The polarisation in the two glasses have the directions RP and LP, oriented in the angles RA respectively LA, which are given the values 45 respectively 135 degrees, or 135 respectively 45 degrees. The projected light naturally has the corresponding polarisation directions for right-, respectively left-registered pictures.

Fig. 16 schematically illustrates the design of a 3-D three colour projector, where an off-axis field lens FL is used to direct the light into the projection lens 7. The design is in other respects the same as illustrated in figure 1, with the difference that the condensor is shown as if its optical axis were coinciding with the normal of midpoint of the IGD. The field lens may very well be combined with the prismatic matrix 21 in figure 5, whereby means to colour correction of the colours from the light sources 2R,2G,2B are given.

Fig. 17 schematically illustrates a detailed study of an IGD with condensor C, despeckelator/depixelator 88 and field lens FL. The matrix ML can either consist of only positive microlenses or a combination of positive microlenses and microprisms with little prismatic angle, where the prisms are used to correct the colour aberrations originating in the field lens as well as to, together with the field lens, contribute to the light deflection.

Figs. 18 to 21 illustrate four different positions of the projector 10 in relation to the projection screen 11 and 11'. The screen can be used for projection on booth sides 11 and 11'. When the side 11 of the screen is used, then the projector 10 is positioned at a corner with the symbol HI. When the side 11' of the screen is used, then the projector 10 is positioned at the corner indicated with the symbol H2. The symbols HI and H2 are indicated with HI' and H2' respectively when they are on the back side of the screen.

The term Optically orthogonal' relates to planes or lines, that would be mathematically orthogonal if no plane mirrors were in-between them.

The term Optically parallel' relates to planes or lines, that would be mathematically parallel if no plane mirrors were in-between them.

Although the shown embodiments of the present invention have been described in detail with reference to the appended figures, it should be realized that the invention is not limited to these specific embodiments and that different changes or modifications can be obtained by a person skilled in the art, without departing from the scope of the invention.

Claims

1. A method for projection for reproduction of electronically registered pictures from a first gamut (G_c) representing the registered picture, to a second gamut (G_rnG_d), which mainly is the intersection between a fourt gamut (G_d) and a third gamut (G_r), where the fourth gamut (G_d) represents a gamut for a projector arrangement and the third gamut (G_r) represents the gamut for realistic sceneries, which is larger or smaller than the first gamut (G_c) for the registered picture, characterised in,

(1) that a colour in the registered picture with the colour co-ordinates (x_c,y_c) in the projection arrangement being represented by expanded or contracted colour co-ordinates (x_r,y_r), of which point mainly lies within the second gamut (G_rnG_d),

(2) that the point (x_r,y_r) is chosen with constant hue and chroma, which is adapted to the relative distance from a white point (x_n,y_n) to the edge of the second gamut (G_rnG_d) in relation to a distance from the white point (x_n,y_n) to the edge of the first gamut (G_c) along a line through the colour coordinates (x_c,y_c) and the white point (x^y_^) .

2. Method for projection according to claim 1, characterised in,

(1) that an intersection point (X_bcΥ_bc) J-^s determined between the edge of the first gamut (G_c) and a line passing through the white point (x_n,y„), which is common for all gamuts (G_c, G_r,G_d) and the point (x_c,y_c) for the registered picture, (2) that an intersection point (x_b,Y_b) is determined between the edge of the second gamut (G_rDG_d) and a line passing through the white point (x_n,y_n) and the point (x_c/y_c) for the registered picture,

(3) that the colour has a colour expansion factor k_rβ with values between zero and one and an excitation potential p_rβ, that is calculated as follows:

p_rβ = l+k_rβ*((x_b-x_n)/(x_bc-x_n)-l) (1)

or

p_re = l+k_rβΛ((yb-y_n)/(y_bc-y_n)-l) (2)

3. Method for projection according to claims 1 or 2, characterised in, that the expanded or contracted colour coordinates (x_r,y_r), lying within the intersection (G_rDG_d) of the fourth gamut (G_d) and the third gamut (G_r) can be calculated as:

4. Method for projection according to claims 1,2 or 3, characterised in, that a finite number of intersection points (Xi,Yι) is determined between the lines, that are delimiting the first gamut (G_c) and the line passing through the white point (x_n,y_n), which is common for all gamuts (G_c/ G_r,G_d), and the point (x_c,y_c) for the registered picture.

5. Method for projection according to any of the claims 1 to 4, characterised in, that the intersection point ( bc/Ybc) ^can t"^e achieved according to: Min(Abs(Xi-x_n)

)^,Sgn(y_i-y_n)+y_n

(5) for all in = l,n

6. Method for projection according to any of the preceding claims, characterised in, that the intersection (G_rnG_d) can be approximated with a polygon, which is a fifth gamut

(Gr ) .

7. Method for projection according to any of the claims 1, 2, 3 or 6, characterised in, that a finite number of intersection points (Xi,yι) are achieved between the lines delimiting the fifth gamut (G_rr) and the line passing through the point (x_n,y_n), which is common for all gamuts gamuts (G_c,G_r,G_d), and the point (x_c,y_c) for the registered picture, where a point (X_b,Y_b) on the edge of the fifth gamut (G_rr) can be achieved according to:

Min(Abs(x₁-x )-Sgn(x₁-x +x_n,Min(.Abs(y_i-y )-Sgn<y₁-y_n)+y_n(6) for all in = l,n

8. Method for projection according to any of the preceding claims, characterised in , that the value on the colour expansion factor (k_rθ) can be controlled by a control (CC), preferably via a remote control unit, on a unit (CE) .

9. Method for projection according to any of the preceding claims with reproduction of electronically registered pictures with at least three light sources (2R,2r,2G,2g,2B,2b) , , characterised in, that a sixth gamut (GO) generated by the different light sources (2R,2r,2G,2g,2B,2b) is divided in several disjoint triangles (G1,G2,G3,G4) , where the number of triangles is equal to the number of light sources subtracted by two.

10. Method for projection according to any of the preceding claims, characterised in, that the colours for an arbitrary point (CP) in the sixth gamut (GO) can be represented by three light sources, of which the colour coordinates (3B,3b,3C) generates the triangle (G2), in which the point (CP) lies.

11. Method for projection according to any of the claims 1 to 8, characterised in, that the light flux (ΦO) for an arbitrary point is maximised or that the inverse of the light flux (ΦO) for an arbitrary point is minimised.