GB2355355A

GB2355355A - Afterglow correction in flying spot scanners

Info

Publication number: GB2355355A
Application number: GB9924031A
Authority: GB
Inventors: Terence William Mead
Original assignee: Cintel International Ltd
Current assignee: Cintel International Ltd
Priority date: 1999-10-11
Filing date: 1999-10-11
Publication date: 2001-04-18
Also published as: GB9924031D0

Abstract

A method of deriving a correction value for an output signal 110 in a flying spot CRT scanner provides correction for afterglow characteristics of a CRT tube 10. The illumination level of a CRT is detected during a blanking interval of the CRT. A contribution towards light output by adjacent portions of a CRT face is calculated by analysing the signal level after a series of delays (for example, 400, 420, Figure 6) to determine the afterglow characteristic and provide compensation.

Description

2355355 Afterglow Correction in FlVinq Spot Scanners This invention

relates to flying spot scanners, and in particular to methods of and apparatus deriving a correction for automatically compensating for deviations from an ideal signal caused by inherent properties of the flying spot scanner such as phosphor afterglow/persistence and or flare.

The invention is particularly applicable to telecine, which operates by imaging a cathode ray tube raster onto a film, collecting the light passing through the film and converting the collected light into a television signal representative of the images on the film. One problem with such telecines is afterglow, which is the persistence of light from the cathode ray tube phosphor after the spot is has moved. When the film changes from clear to dense, the video output should immediately change to a minimum.

However, afterglow effects cause a decay from peak to minimum signals. This decay shows itself as a decaying white streak to the right of any white information on a black picture. This is because an image signal derived from a given location on film will include a contribution from adjacent portions of film which remain illuminated by the decaying light of the phosphor previously scanned. In existing machines, the afterglow characteristic of the phosphor used decays to 10% in about 120 nanoseconds but continues to emit smaller proportions of light which remain significant for about 200 microseconds.

For a infinitely small clear spot on an otherwise dark film, picture streaking will follow the phosphor decay curve. However, if a larger spot of clear film is considered the streaking will correspond to the integral of the phosphor decay curves over the exposed scan time.

To overcome this unwanted streaking it is known to correct for afterglow using a series of individually adjustable differentiating circuits in an analogue corrector circuit.

A method and apparatus for correcting afterglow in a flying spot scanner is disclosed GB-A-2,239,572 and US-A 5,278,653 which describe a means of automatically correcting a digital signal for afterglow. The system disclosed in these documents comprises a corrector circuit which determines the contribution to the image signal at any given location made by afterglow from adjacent scanning locations and which subtracts this contribution from the signal. The system also provides a method of correcting a digital signal produced by a flying spot scanner to compensate for deviations from an ideal output signal caused by the inherent properties of the flying spot scanner, comprising the steps of determining for the signal corresponding to a given location of the flying spot, the contribution to the signal from each of a plurality of adjacent scanning locations or groups of locations, and subtracting each of the determined contributions from the signal to provide a corrected signal. The method described in GB-A-2,239,572 and US-A 5,278,653 requires that all but one pixel, or group of pixels, are blanked at the cathode ray tube (CRT) during the alignment process. Accordingly, to determine the contribution to an image signal from adjacent scan locations due to afterglow, this system is configured when switching on or resetting the machine in an initial alignment operation.

Flare is a phenomenon which is present in all flying spot scanners and may be divided into three broad categories; high frequency flare, low frequency flare and flare ringing at intermediate frequencies. Flare rings are known to be caused by changes of refractive index at the glass/air interface at the face plate of the cathode ray tube.This causes light to be reflected back to the phosphor screen at different positions producing a ring around the spot. Although it is not possible to say exactly what causes high and low frequency flare it is clear that both types are produced from a number of different sources which interact with one another, for example general background light and local scattering at the glass-to-air interface. This latter type varies with the image position and is much greater at the edge of an image and tends to smear the image. Multiple reflections tend to produce higher frequency flare.

The two documents GB-A-2,239,572 and US-A-5,278,653 discussed above also describe a means of automatically correcting a digital signal for flare in a flying spot scanner using the system described. Flare may also be is reduced to an extent by using high quality faceplates and by increasing the thickness of the faceplate as described in GB-A-2,199,443. We have appreciated a problem with the known afterglow and flare correction circuits described, in that the circuit 20 must be configured in a procedure without film in the telecine in an initial alignment operation. We have further appreciated, therefore, that the known system must be stopped from usual operation to perform the alignment. Accordingly, in a broad aspect, the invention provides 25 automatic afterglow correction, and flare correction, dynamically during the normal operation of the telecine, thereby avoiding the need for any preliminary alignment. This has the benefit of correcting for any changes in the afterglow characteristic caused by changes in operating 30 conditions of the CRT or its environment for example as might occur with change of temperature. We have further appreciated that dynamic automatic afterglow correction can be achieved during the normal vertical blanking period of the CRT, since for the preceding period of time the CRT will have been subject to a long period of high illumination.

Accordingly, there is provided a method of deriving a s correction value used to compensate for deviations of an output signal caused by inherent properties of a flying spot CRT scanner, comprising:

operating the flying spot CRT scanner to produce a flying spot raster scan on the CRT face comprising repeated scans with a blanking interval therebetween; detecting the illumination level of the CRT during a portion of the blanking interval; and deriving from the illumination level detected during the portion of the blanking interval a proportion of the output signal, in respect of any given location of the flying spot of the CRT, contributed by at least one proximate location to thereby produce the correction 20 value.

The use of the blanking interval in which to measure the illumination level allows afterglow correction to be performed dynamically, that is during usual operation of the film scanner, rather than stopping the film scanner to perform an initial configuration.

The invention also provides apparatus for deriving a correction value to compensate for deviations of an output signal caused by inherent properties of a flying spot CRT scanner, comprising:

a detector arranged to detect the illumination level during a portion of a blanking interval whilst operating the flying spot CRT scanner to produce a flying spot raster scan on the CRT face comprising repeated scans with a blanking interval therebetween; and means arranged to derive, from the illumination level detected during the portion of the blanking interval, a proportion of the output signal, in respect of any given location of the flying spot of the CRT, contributed by at least one proximate location to thereby produce the correction value.

In the apparatus aspect, a detector is arranged to view the light level of the CRT and from this the correction value is derived.

It is preferred that the detector is configured to detect the illumination level immediately after the start of the blanking interval. This allows detection of the illumination level of the last few pixels of the CRT face as the light output from the CRT decays in the usual manner of afterglow.

In the embodiment the detector is configured to detect the illumination level after each of a series of delays during the portion of the blanking interval and the means for deriving being configured to derive, from the illumination level after each of the series of delays, the proportion of the output signal contributed by a corresponding series of proximate locations to any given location of the flying spot.

In contrast to a known afterglow correction method employing a single short pulse of high illumination on the CRT, this embodiment of the invention thus uses a long period of high illumination, and requires a different form of calculation of the afterglow correction co-efficients.

In the embodiment, the calculation is performed in the means for deriving which is configured such that the proportion of the output signal contributed by a proximate location is derived by subtracting from the illumination level corresponding to that location, the illumination level at the next location in the series.

As an alternative, the detector could be used in the apparatus which is configured to cause the flying spot scanner to scan such that the blanking interval includes illuminating the CRT at a single scan location and blanking the CRT at ail other locations. The derivation is of coefficients of correction in this alternative is similar to the known method and is such that the detector is configured to detect the illumination level after each of a series of delays during the portion of the blanking interval, and means for deriving is so configured that the proportion of the output signal contributed by a proximate location is derived from the illumination level after the corresponding delay.

We have appreciated that the accuracy of a correction value can be improved by averaging, and for that reason, the detector is preferably configured to detect the illumination level during a portion of a subsequent blanking interval, the means for deriving being configured to derive an error in the previously derived correction, and adjust the correction to take account of the error.

This can be performed by configuring the deriving means in two ways. First the means for deriving being configured to derive a subsequent correction during the subsequent blanking interval, calculate the difference between the subsequent correction and the correction and add a percentage of the difference to the correction to take account of the error and produce a new correction.

Second, the means for deriving being configured to derive a subsequent correction during the subsequent blanking interval, add a percentage of the subsequent correction to a percentage of the correction to take account of the error and produce a new correction. Either apparatus will provide continual correction with a reduced (average) error. Particular if the detector is configured to repeatedly detecting the illumination level during subsequent portions of blanking intervals to continually take account of the error and produce a new correction.

The invention also resides in an afterglow corrector and a telecine incorporating the features discussed above.

An embodiment of the invention will now be described, by way of example only, in which:

Figure 1 is a diagrammatic representation of the response of a telecine with afterglow correction; Figure 2 is a schematic diagram of an afterglow corrector circuit; Figure 3 is a schematic diagram of an afterglow corrector circuit embodying the invention; Figure 4 is a schematic diagram of a variation of the embodiment of Figure 3; Figure 5 is a diagrammatic representation of the response of a telecine using an afterglow corrector embodying the invention; Figure 6 shows the delay part of the circuit of Figure 3; Figure 7 is a block diagram of a digital afterglow corrector embodying the invention; and Figure 8 shows the delay part of the circuit of Figure 7.

The embodiment is a flying spot film scanner in the form of a telecine. A general description of flying spot film scanning principles may be found in "TV & Video Engineer's

Reference Book", K.G. Jackson and G.B. Townsend, Butterworth & Heinemann, ISBN 0 7506 1953 8 Chapter 39, to which reference should be made for guidance on the architecture of colour flying spot film scanners and telecine; the principles being well known to those skilled in the art.

A flying spot scanner such as a telecine derives an electrical signal from film by illuminating film with a spot of light which scans the film in a progressive raster. The spot of light is produced by a cathode ray tube (CRT) which produces a spot of light by exciting phosphor with an electron beam. Ideally, the signal representing any one scanning location is formed solely from light attributable to the flying spot at that location. Such an ideal response is shown in Figure la which diagrammatically shows the signal level produced by a given location on the vertical axis against time on the horizontal axis (the time relating to location on the CRT phosphor by the speed of the flying spot). However, in practice, the signal contains additional components from previously scanned locations due to phosphor afterglow.

Thus the actual signal is made up of the component from that location, together with a series of additional components from previous scanned locations, the size of the previous components decreasing with increasing time since scanning. The actual composition is shown in Figure 1b.

To compensate for these additional components, an afterglow correcting circuit reproduces the additional afterglow components - Figure lc - and subtracts these reproduced components from the actual image signal. Thus the corrected signal shown in Figure ld should be equal to the ideal signal of Figure la. For simplicity of description the components of the signal represented in Figure I are shown as a small number of discrete sampled levels, in practice the error would be a continuous curve lasting for many more samples, and the digitally sampled is correction signal would approximate that error signal. The basic functional elements of a known monochrome flying spot film scanner with afterglow correction are shown in Figure 2. Those skilled in the art will know how this application can be extended to colour operation. As 20 previously explained, a known afterglow correction circuit makes one measurement of the correction values as part of an automatic alignment process, and these values are used until alignment is initiated again. The CRT (10) produces a rastered progressive scan light 25 output which is imaged via optics (20) to be focussed on the film (40) passing through a film gate (30). The emerging rastered light is imaged via optics (50) on to a photo electric cell (60) to produce a signal representative of the film image and its electrical output 30 passed to processing (100) which includes correction for the CRT (10) afterglow. The electrical processing in this example being digital and the afterglow corrected output is provided (110). The determination of the afterglow correction (90) is calculated in an auto alignment mode whereby the film (40) is removed from the film gate (30), to leave the photo electric cell (60) with a clear view of the imaged CRT light. The Auto alignment process (70) is initiated which closes the switch (80) to allow the Afterglow Measurement and Correction calculation (90) circuitry to calculate the afterglow correction values for subsequent application to the processing (100) Once the auto alignment process is completed the switch (80) is opened to disconnect the Afterglow Measurement and Correction (90) from the Photo Electric Cell (60). The afterglow correction processing (100) continues to provide correction using the afterglow correction values calculated during the alignment until film is removed from the gate and alignment re-executed.

We have appreciated that the known afterglow corrector cannot provide the ability to provide a film scanner or telecine with dynamic afterglow correction. In the embodiment of the invention, now to be described, the characteristic performance of the CRT is directly seen and measured without the interference of the film between the CRT light and the measurement sensor. This is preferably provided by an additional reference sensing means with equivalent performance to that of the image sensing means, and sensing a sample of the light from the CRT before it passes through the film. If is preferable that the afterglow content of the directly viewing reference sensor should exactly match that of the main image signal being corrected. For this reason the colour response of each reference sensor, in any multi-colour analysis film scanner or telecine, should be the same as that of the equivalent main image sensor channel.

In the embodiment of the invention now described the vertical blanking period of the CRT is used to permit measurement of the afterglow of the CRT following extinguishing of long term high illumination. In an alternative embodiment a singular pulse of illumination is used during the vertical blanking period.

The nature of the afterglow which needs to be corrected for CRT's used in film scanners and telecines may last up to 200 microseconds dependent upon the phosphor used with the CRT. The vertical blanking period (typically lms) of the host film scanner or telecine system must provide adequate time for the measurements to be made. It is for this reason that the shorter horizontal blanking period (typically 6ps) is not chosen.

Any signal system contains noise, so some form of noise integration is preferred. For example, the measured correction value can be added to the current applied correction signal in small quantities each scan interval, or alternatively the afterglow correction can be left in operation whilst the new measurement is made which results in a measurement of the difference between the old applied afterglow correction and the value required to correct the new measurement. This difference, or a proportion of it, is then added to the old correction value to produce the new correction value.

Figure 3 shows the preferred embodiment of the invention. For simplicity, a single monochrome channel is shown but those skilled in the art will understand the well known extension to colour operation. Like components have the same references as before. A CRT (10) produces rastered illumination which is imaged via optics (20) to be focussed on the film (40) held in a film gate (30) The emergent light modulated by the film (40) is imaged via optics (50) on to a photo electric cell (60) which converts the light into an electrical signal representative of the film image which is processed in processing circuitry (100) including afterglow correction, to produce the final output (110). A detector in the form of separate photo electric cell (PEC) (120) may be a dedicated cell for afterglow correction or also used for other purposes such as burn correction. The detector PEC (120) is arranged to have the same performance characteristics as that of the image sensing photo electric cell (60), so that the afterglow sensing is the same, and to directly view the rastered light from the CRT (10). Means arranged to derive a correction value comprises circuitry for afterglow measurement 90, 140 and adder and multipliers 150, 160, 170. During each, or selected, CRT vertical blanking (130) periods the switch (80) is closed to permit the Afterglow Measurement and Correction Calculation in circuitry (90) to be made.

A proportion k% of the newly evaluated afterglow correction is then added (170) to the proportion (1-K)% (160) of the previous afterglow correction value (140) and applied to image processing circuitry (100) to produce the afterglow corrected output (110). This current afterglow correction value is then used to become the new Previous Afterglow Correction (140) A suitable value for k might be 0.1, i.e. 10%.

Figure 4 shows a variation on the embodiment of Figure 3.

In Figure 4 the separate photo electric cell (120) which directly views the CRT (10) raster light feeds an equivalent afterglow correction processing circuit (180) to that contained within the Processing and Afterglow corrector processing circuit (100) During the vertical blanking interval (130) the switch (80) closes and the Afterglow Measurement & Correction Calculation (90) measures the error in afterglow, i.e. the uncorrected afterglow. The error, or a proportion Q% (190) of the measured error is added (170) to the Previous Afterglow Correction (140) and this new afterglow correction is used for the Processing and Afterglow Correction processing circuitry (100) and the Afterglow Corrector processing circuitry (180).

Figure 5 is a diagrammatic representation of the signal level (vertical axis) against time (horizontal axis) for the ideal and actual signal levels at the vertical blanking interval boundary. The ideal signal condition (200) when the CRT (10) is blanked by the vertical blanking signal (130) and the rastered light turns off is a clean step. However, in practice afterglow maintains a residual light output so that the situation is in practice that shown as (210). The description is given, for simplicity, for a three stage afterglow example, namely by considering contributions to light output from the previous 3 scan locations only. The corrected output will not be ideal with just three correction terms, and their associated coefficients, but for the purposes of explanation this is adequate. In practice further processing stages could be included.

Referring again to Figure 5, consider the time at which the spot illuminates a position three pixels just prior to the vertical blanking interval. The light output at this position consists of the true output 1 plus the contribution a, b and c, of afterglow from the previous three pixels, see (210). In each case 1>a>b>c. At the time the vertical blanking operates 1 falls to zero (the beam is interrupted) but a, b and c remain. Thus for the first pixel after the vertical blanking the light output is a + b + c, for the second pixel after the vertical blanking the light output is b + c, and for the third pixel after the vertical blanking operates the output is c - as shown in (210). By the fourth pixel the light by this simplified example, will be turned off. By measuring the residual light for three pixels in time following the operation of the vertical blanking, the co-efficients a, b, and c may be determined using the circuitry shown in Figure 6.

With reference to Figure 6, the input signal is applied to delays of 2T (400), 1T (420) and zero delay.

The delayed and not delayed signals are applied to subtractors (410) (430) whose outputs are latched (440) at sample period 3T after the start of blanking. The latched (440) outputs correspond to values a, b and c. To aid understanding of its operation Figure 6 is annotated with the circuitry values at the 3T period after the start of blanking.

The co-efficient (a) thus represents the proportion of afterglow caused by a pixel immediately before a given scan location. If that pixel was at maximum video level (a clear piece of film) then the afterglow caused by it would have the value (a), similarly if that pixel had value zero (dark film) then it would cause no afterglow. The correction to be subtracted from the signal is the product of the co-efficient and it's associated pixel value (the film image). The co-efficient (b) represents the proportion of afterglow caused by the second pixel before, and co-efficient (c) represents that caused by three pixels before the given scan location.

The processing and afterglow correction circuitry (100) is shown in greater detail in Figure 8. The signal input from the image photo electric cell (60) is applied to subtractor 510 where the afterglow correction signal is subtracted from it.

The input signal which varies with time representing the lines of image is delayed by one pixel period (T) by delay 520, then multiplied by co-efficient (a) by multiplier 530 and applied to adder 540. The input signal is similarly delayed by two pixel periods and multiplied by coefficient (b) using delay 521 and multiplier 531, and also delayed by three pixel periods and multiplied by coefficient (c) using delay 522 and multiplier 532. These three components of the correction signal are added by adder 540 and then subtracted from the input signal in subtractor 510, producing the corrected output signal. Other known apparatus could be used with the measured coefficients to carry out the actual afterglow correction.

The example given of afterglow being just three coefficient terms is a simplification and many more terms are preferred in practice. The known technique of making use of the fact that the overall afterglow decreases with increasing delay, equally applies to the dynamic automatic is afterglow correction. Applying this technique the pixel measurement point will increase by a factor of 2 each measurement time. Thus, the first measurement will be made at 1 pixel time, the second at 2 pixel time, the third at 4 pixel time, the fourth at 8 pixel time etc., in general terms this corresponds to measurements being made at 21 pixels where N is incremental measurement number.

The benefits of dynamic automatic afterglow correction may also be achieved using the additional separate sensor (120) to act as the photo electric cell to detect a single pixel of high illumination during the vertical blanking interval and the afterglow immediately thereafter. The calculation of the coefficients would then be similar to known methods and apparatus for digital correction of afterglow in flying spot telecines. In such an arrangement with reference to Figure 3, the afterglow coefficients would be derived from the photo electric cell (120) which has an uninterrupted view of the CRT (10) and these co-efficients would be applied to the afterglow corrector which forms part of the image processing (100) Using this technique an additional single pixel high illumination of the CRT would be required during the vertical blanking period.

The techniques described herein for dynamic afterglow correction are not directly applicable to flare correction but some benefit may be gained. Dynamic flare correction would require use of a known test pattern on film. A test pattern consisting of a white spot on an otherwise black area of the film would be suitable, but other patterns may be used. Future films might have such a pattern included but this would not be the case for older films.

We have appreciated that it is possible to calculate flare correction from the existing film information by weighting all the correction values according to the sequence of film data values that precede the scanning of the film framing bar. However, this method is complex and prone to errors caused, for example, by de-focussing of the film framing bar. The need for dynamic flare correction is much less than for dynamic afterglow correction, since flare is very much less likely to change over time.

Figure 7 shows the overall block diagram for digital dynamic automatic afterglow correction in an image data system- Black levels cannot be determined until afterglow correction has been carried out. This can be achieved as set out in prior art US-A-5,278,653 and GB-A-2,239,562 by use of a derived black clamp pulse (330) which is fed back to the analogue input (300) to the image signal ADC (310) which feeds the digital afterglow correction (320) driven by the dynamically calculated afterglow co-efficients (340) as described by the various techniques herein.

Whilst the preferred embodiments described use digital processing it is also possible to implement the invention using analogue circuitry. In this case the afterglow correction could be applied in the known fashion of manual analogue correction but using motorised potentiometers, the motors being controlled by the correction coefficients derived as before. Alternatively the potentiometers could be replaced by analogue multipliers. Similarly the circuitry used to derived the correction coefficients could use analogue delay lines and subtractor circuits.

Claims

Claims 1 A method of deriving a correction value used to compensate for

deviations of an output signal caused by inherent properties of a flying spot CRT scanner, 5 comprising:

operating the flying spot CRT scanner to produce a flying spot raster scan on the CRT face comprising repeated scans with a blanking interval therebetween; detecting the illumination level of the CRT during a portion of the blanking interval; and deriving from the illumination level detected during the portion of the blanking interval, a proportion of the output signal, in respect of any given location of the flying spot of the CRT, contributed by at least one proximate location to thereby produce the correction value.
2. A method according to claim 1, wherein the portion of the blanking interval during which the illumination level is detected is immediately after the start of the blanking interval.
3. A method according to claim 1 or 2, wherein the portion of the blanking interval during which the illumination level is detected is one or more pixels in duration.
4. A method according to claim 1, 2 or 3 comprising detecting the illumination level after each of a series of delays during the portion of the blanking interval and deriving, from the illumination level after each of the series of delays, the proportion of the output signal contributed by a corresponding series of proximate locations to any given location of the flying spot.
S. A method according to claim 4, wherein the proportion of the output signal contributed by a proximate location is derived by subtracting from the illumination level corresponding to that location, the illumination level at the next location in the series.
6. A method according to any preceding claim, wherein each location corresponds to a pixel of image.
7. A method according to claim 1, wherein the blanking interval includes illuminating the CRT at a single scan location and blanking the CRT at all other locations.
8. A method according to claim 7, wherein the portion of the blanking interval during which the illumination level of the CRT is detected is immediately after illuminating the CRT at the single scan location.
9. A method according to claim 8, comprising detecting the illumination level after each of a series of delays during the portion of the blanking interval, and wherein the proportion of the output signal contributed by a proximate location is derived from the illumination level after the corresponding delay.
10. A method according to claim 4, 5 or 9, wherein the correction value comprises a series of coefficients, each coefficient giving the proportion of the output signal contributed by the series of proximate locations to any given location of the flying spot.
11. A method according to any preceding claim, further comprising detecting the illumination level during a portion of a subsequent blanking interval, deriving an error in the previously derived correction value, and adjusting the correction value to take account of the error.
12. A method according to claim 11, comprising deriving a subsequent correction value during the subsequent blanking interval, calculating a difference between the subsequent correction value and the correction value and adding a percentage of the difference to the correction value to take account of the error and 15 produce a new correction value.
13. A method according to claim 11, comprising deriving a subsequent correction value during the subsequent blanking interval, adding a percentage of the subsequent correction value to a percentage of the correction value to take account of the error and produce a new correction value.
14. A method according to any of claims 10, 11 or 12, comprising repeatedly detecting the illumination level during subsequent portions of blanking intervals to continually take account of the error and produce new correction values.
15. A method according to any preceding claim, wherein the blanking interval is the vertical blanking interval.
16. A method according to any preceding claim, wherein the deviations of the output signal are caused by afterglow of the CRT.
17. Apparatus for deriving a correction value to compensate for deviations of an output signal caused by inherent properties of a flying spot CRT scanner, comprising:

a detector arranged to detect the illumination level during a portion of a blanking interval whilst operating the flying spot CRT scanner to produce a flying spot raster scan on the CRT face comprising repeated scans with a blanking interval therebetween; and means arranged to derive, from the illumination level detected during the portion of the blanking interval, a proportion of the output signal, in respect of any given location of the flying spot of the CRT, contributed by at least one proximate location to thereby produce the correction value.
Apparatus according to claim 17, wherein the detector is configured to detect the illumination level immediately after the start of the blanking interval.
19. Apparatus according to claim 17 or 18, wherein the detector is configured to detect the illumination level for a duration of one or more pixels.
20. Apparatus according to claim 17, 18 or 19, the detector being configured to detect the illumination level after each of a series of delays during the portion of the blanking interval and the means for deriving being configured to derive, from the illumination level after each of the series of delays, the proportion of the output signal contributed by a corresponding series of proximate locations to any given location of the flying spot.
21. Apparatus according to claim 20, wherein the means for deriving is configured such that the proportion of the output signal contributed by a proximate location is derived by subtracting from the illumination level corresponding to that location, the illumination level at the next location in the series.
22. Apparatus according to any of claims 17 to 21, wherein each location corresponds to a pixel of image.

is
23. Apparatus according to claim 17, wherein apparatus is configured to cause the flying spot scanner to scan such that the blanking interval includes illuminating the CRT at a single scan location and blanking the CRT at all other locations.
24. Apparatus according to claim 22, wherein detector is configured such that the portion of the blanking interval during which the illumination level is detected is immediately after illuminating the CRT at the single scan location.
25. Apparatus according to claim 24, wherein the detector is configured to detect the illumination level after each of a series of delays during the portion of the blanking interval, and means for deriving is so configured that the proportion of the output signal contributed by a proximate location is derived from the illumination level after the corresponding delay.
26. Apparatus according to claims 20, 21 or 25, wherein the correction value comprises a series of coefficients, each coefficient giving the proportion of the output signal contributed by the series of 5 proximate locations to any given location of the flying spot.
27. Apparatus according to any of claims 17 to 26, wherein the detector is configured to detect the illumination level during a portion of a subsequent blanking interval, the means for deriving being configured to derive an error in the previously derived correction, and adjust the correction to take account of the error.
28. Apparatus according to claim 27, the means for deriving being configured to derive a subsequent correction during the subsequent blanking interval, calculate the difference between the subsequent correction and the correction and add a percentage of the difference to the correction to take account of 20 the error and produce a new correction.
29. Apparatus according to claim 27, the means for deriving being configured to derive a subsequent correction during the subsequent blanking interval, add a percentage of the subsequent correction to a percentage of the correction to take account of the error and produce a new correction.
30. Apparatus according to any of claims 27, 28 or 29, the detector being configured to repeatedly detecting the illumination level during subsequent portions of blanking intervals to continually take account of the error and produce a new correction.
31. Apparatus according to any of claims 17 to 30, wherein the blanking interval is the vertical blanking interval.
32. Apparatus according to any of claims 17 to 31, wherein the deviations of the output signal are caused by afterglow of the CRT.
33. Apparatus according to any of claims 17 to 32, wherein the detector comprises a photoelectric cell arranged to view light direct from the CRT.
34. Apparatus according to any of claims 17 to 33, wherein the means arranged to derive the proportion of the output signal comprises one or more delays for delaying a signal from the detector and a subtractor for subtracting the illumination level in respect of is one location from the illumination level in respect of another location.
35. An afterglow corrector for a film scanner comprising apparatus according to any of claims 17 to 34, and means for correcting an output signal in accordance with the correction value.
36. A telecine comprising a cathode ray tube (CRT), a film transport and a detector for producing an output signal representative of an image on film, and an afterglow corrector according to claim 35.