H41: Reproduction of Images: Colour Displays

James H Nobbs


The two major methods of reproduction of images are visual displays and the four-colour printing process.  Both of these methods are based on the trichromatic nature of colour vision.  The wide range of colours seen in the reproduced image is created by controlling the amounts of light in the red, the green and the blue wavelength bands received by the eye.

The most common method of generating colour by the additive system is visual display units, such as computer display screens and television screens.  Visual display screens are composed of individual dots, which, at normal viewing distances, are too small to be seen and the picture appears continuous.  In an image on a video display screen the dots are of constant size but of various degrees of brightness from black to full intensity.

Creating the displayed image

The image on a display screen appears to be present all of the time, but in fact the brightness of each part of the screen is flashing on and off many times in each second.  The flashing of the screen is more rapid than can be followed by the eye and, as a result of “persistence of vision”, the image appears to be continuous.

In video displays the image is built up from a sequence of pulses of light that are move across the screen in a series of lines across the screen and down the screen. 

A typical interlaced scan sequence that is used in television screens is shown in Figure 1.  From “start “ the spot of light moves left to right tracing out the “odd” lines making up Frame 1 (heavy lines) of the image.  The spot then flies upwards and fills in the “even” lines (light lines) of Frame 2.  The scanning sequence is also known as a raster.  If the complete image is scanned within 1/25 second and the sequence repeated many times per second, then your eye will not distinguish the separate pulses and the perception is of a flicker free image.

Figure 1: Interlaced scan sequence to create an image on a  PAL (European) display screen

Image capture

A video camera or a scanner captures an image, by cutting the field of view into a number of thin slices or lines.  Each line is “scanned” in succession and the variation in brightness along the line forms a sequential stream of information.  The sequential stream of information is encoded into digital format for example, or into the composite video signal of broadcast TV.

Image reproduction

To reproduce the image, the stream of information is decoded and used to control the intensity emitted by each little spot scanned along the length of a line on the video screen.  The display builds up the slices again, reproducing each line in the same sequence as the image capture.

Persistence of vision

Video display screens make use of the property of the eye known as “persistence of vision.  If you view a picture for a fraction of a second, and then take the picture away, the eye will continue to “see” it for a short time after.  In humans, a short pulse of light persists in visual impression for about a 1/15 of a second. 

Early television

Image capture, analysis

John Logie Baird made the first successful television transmissions between 1928 and 1935.  Baird used the BBCs medium wave transmitters that were normally used for radio broadcasting.

In this system the pictures were composed of only 30 lines and small details could not be reproduced, this was a low-definition television system.

The first cameras and receivers were almost entirely mechanical in operation, and looked very different from modern equipment.  Figure 2 is an illustration of one of Bairds experimental cameras.  Baird used a system of apertures cut into spinning disks to slice the image into 30 lines at a rate of ten images per second.

Figure 2: Bairds experimental camera capable of producing 10 images per second each composed of 30 lines

Image display, synthesis

Bairds receivers, like the cameras, were semi-mechanical in operation and looked very different from a modern television, as shown in Figure 3 and Figure 4.

A disk within the receiver was spinning at exactly the same rate as the disk in the camera.  The disk had a number of slots cut into the surface.  The slots passed in front of a neon lamp, generating a view of the lamp as a sequence of lines.  The intensity of the light emitted by the lamp was pulsed on and off in the same sequence as the variation in brightness detected by the camera.  In this way the image was recreated line by line so that a continuous image was clearly seen by the viewer.

Figure 3: Baird receiver, rear view showing the
spinning disk A and the neon lamp B

Figure 4: The Baird Televisor

To the viewing public, it seemed more like magic than science.  As a newspaper report stated “Actions taking place far away are reproduced by the combined magic of light, electricity, ether waves and delicate mechanical apparatus”

Image reproduction, visual displays

The most important type of visual display makes use of the cathode ray tube (CRT).  Historically this was the first type of colour display for television and computers and the colour standards for visual display systems are mainly based on the properties of CRT systems.

Figure 5 illustrates how light is generated in a CRT type of display.  The screen of a CRT computer display or television receiver is coated with tiny dots of rare earth phosphors that are electroluminescent.  Within the picture tube, an electron beam is scanned across the screen and when it hits one of the phosphor dots, the kinetic energy of the electrons is absorbed by the phosphor and transformed into light and heat energy.  The light emitted is a characteristic colour and the amount of light depends on the kinetic energy of the electrons in the beam and the intensity of the beam.

Figure 5: Process of light generation in a CRT

This type of display creates the range of colours by means of the additive colour mixing of light from three types of electro-luminescent phosphors.

Additive colour mixing

The additive primary lights are red, green and blue in colour, are chosen to be able to reproduce a wide gamut of colours.

The individual red, green and blue light emitting phosphor dots are close together and are not resolved by the eye so that the eye perceives the combined (additive) colour.

Additive colour mixing leads to some surprises, as shown in the colour-mixing diagram, Figure 6.  For example, the eye blending together light from the red light emitting dots and the green light emitting dots creates the impression of yellow in an image.  Yellow is one of the subtractive colour primaries, the other two subtractive primaries are obtained when green light and blue light are blended (cyan) and when red light and blue light are blended together (magenta).

Figure 6: Additive colour mixing

Shadow mask display system

Close examination through a magnifying glass of a conventional “shadow mask” computer display screen or television screen reveals a sequence of dots of red, green and blue emitting phosphors.

The phosphor dots are grouped as “triads”, that is a red, a green and a blue dot arranged in a triangle as shown in Figure 7. Each triad on the screen is known as a pixel, this is the smallest area (a picture element) of the image that can be resolved.

If the space between the phosphors is filled with black, then the impression of a brighter image of higher purity is obtained.

Figure 7: Phosphor dot triad

There are three separate electron beams present in a colour cathode ray tube display.  The beams scan across the phosphors on the screen, one beam corresponding to each colour channel.

The electron guns that produce the beams are in a triangular arrangement and each of the three beams of electrons arrives at the screen at a slightly different angle, as shown in Figure 8.

A metal mask, the  “shadow mask”, is between the guns and the screen to ensure that each gun activates only one phosphor within each triad.  The success of the shadow mask in keeping the wrong dots from activating determines the “purity” of colour reproduction.

Good colour and spatial resolution is achieved by having a large number of phosphor dots, each group as close together as possible.

Figure 8: Action of the shadow mask in directing the electron beam onto the correct phosphor dot

Sony Trinitron display system

The principle of the Trinitron display system is the same as that described above however the phosphors are laid down as light emitting strips instead of dots. 

The arrangement of is illustrated in Figure 9.  Three electron guns lying in the same horizontal plane emit three beams of electrons.  A metal aperture plate with vertical slots is used instead of a shadow mask to ensure that each beam hits the correct strip.

The horizontal arrangement of the electron guns makes the Trinitron display tube cheaper to manufacture than the shadow mask tube.

Figure 9: The Trinitron, aperture grill system

The efficiency of conversion of electron energy to light energy is higher, because in the triad system the shadow mask stops a lot of the electron beams energy.  However, the spatial resolution of the Trinitron system is, potentially, not as good as with the triad system.

LCD displays

In a LCD, the screen consists of a liquid crystal cell illuminated by an internal back-light.  Transparent electrodes are evaporated onto the inside surfaces of the two glass plates that hold the liquid crystal, as shown in Figure 10

In the liquid-crystal phase, the materials molecules exhibit partial order.  The liquid crystal molecules will allow polarised light to be transmitted through the display until a voltage is applied to the electrodes.  When subject to the electric field, the liquid crystal molecules take another orientation and the second polariser now blocks the light.  By applying an electrical pulse, the liquid crystal acts like a shutter switching the light transmission of the cell on and off.

Each LCD pixel consists of three cells, with a red, green or blue colour filter placed over each portion.  In this way colour is produced in a similar way to the shadow-mask system.

Figure 10: A cross-section of an LCD display

To increase sensitivity transistors can also be placed at every cell to act as a fast switch.  The colour gamut is in quite good agreement with the conventional display standards, but brightness is not as good as the conventional display.  The advantages to using LCDs are that they are far more efficient in terms of power than CRTs and that they can be made very compact in size.

R G B values

Three numbers, R G and B in a graphic image file on a computer describe the colour of each pixel in the image.  These represent the relative amounts of red, green and blue light respectively, that are to be emitted by the phosphors at that point on the display screen.  The units of R G and B are tristimulus units or “T units”.  This type of unit is defined so that a mixture containing equal T unit amounts of the standard red, standard green and standard blue light sources will match the appearance of a defined white light source, known as the white point of the display.  For example, if an area of the graphic image is set to R=255, G=255 and B=255 then it will have the colour characteristics of the display white point.

White point of the display

Three different white points are in common use and the software for modern computers often allows you to adjust the white-point of your monitor, which will change appearance of the images on the screen.

Broadcast TV (D65 average daylight, colour temperature 6500K)

The white point standards for broadcast television (terrestrial EBU PAL, satellite and HDTV) is average daylight (D65) under the CIE 1931 observer.  The white point for the US TV system is D75, daylight with a colour temperature of 7500 K.

PC screens for office use (D93, colour temperature 9300K)

The white point of computer monitors are set to a high colour temperature (9300K) in order to get more light out of the blue phosphor and make the display more visible in well-lit offices.

PC Screens for the graphic arts (D50, indoor daylight colour temperature 5000K)

In monitors designed for graphics-arts applications and desktop publishing, the white point is D50, .

Gamma factor of a display

You might expect that the light output from an area of the screen is linearly dependent on the RGB values of the image.  In other words, if the R G B values are doubled then twice the intensity of light will be emitted from the screen.  For CRT displays, this is not the case; if the RGB values are doubled then about four times the intensity will be emitted from the screen.

The phosphors in a CRT display produce light by converting the kinetic energy of the electrons in the beam to light energy.  The kinetic energy is proportional to the square of the speed of the electrons.  As a result the intensity of the light emitted by the phosphor is proportional to the R G B value raised to a power of approximately 2.0.  The power law coefficient is known as the gamma factor, and the standard value for the PAL system is 1.80.

When the gamma factor for a display is correctly set, an area of mid grey with R =50%, G = 50% and B= 50% (Figure 11, right) will give the same grey perception as the "merged" impression from a fine pattern of dots with 50% of the area set black and the remaining area set as white (Figure 11 left).

Figure 11: Gamma factor test image.  A 50% grey(right) and an area with 50% black and 50% white(left).

When the gamma factor for the display is set correctly, the two rectangles will have the same lightness.  When the gamma factor is not correctly set, the right rectangle will appear to be either lighter (display gamma too high) or darker (display gamma set too low) than the left rectangle.  The software supplied with many types of modern PC graphics driver cards allow the user to set the gamma factor used by the display.

Quality of reproduction

Colour gamut

It is not possible to reproduce all the colours that the human visual system can perceive by blending together light from three coloured sources.  The CIE x y chromaticity chart shown in Figure 12 provides a way of displaying the colour gamut of the visual system.  The full colour gamut, bordered by the spectral colours, lies within the horseshoe shape.  The display gamut is contained within the triangle joining the points labelled R, G and B.  Figure 12 correctly shows the area of the screen gamut relative to that of the full gamut.

Figure 12: Colour gamut of PAL (EBU) phosphors

Viewing conditions

In order to give good colour reproduction, the display screen should be viewed in subdued lighting, and should have a neutral grey surround around the screen.  The user should wear dark clothing, since the screens surface can be highly reflective typically 4% to 18% - and surface reflections distract the user from a proper assessment of the image.

Colour resolution

The most common level of colour resolution used with personal computer screens at present is 8-bits in each of the R, G and B channels, 24 bits per pixel.  This gives 256 levels for each channel and is sufficient for 16.8 million different R G B codes.

       256 x 256 x 256 = 16.8 million

At first sight, this seems like a huge number of colours, easily able to satisfy the human eye.  However, the fact is that the 16.8 million, although equally spaced throughout RGB space, are not equally spaced visually.  RGB space is highly non-uniform, neighbouring colours in some areas of RGB colour space are indistinguishable from each other, whilst in other areas the colour difference is more than ten times greater than the resolution of the eye.  Colours spaced equally throughout 24 bit RGB space appear to the eye to have colour difference steps varying in size by as much as 30:1 through the different regions.

H41: Reproduction of Images: Colour Displays

James H Nobbs