US2231669A

US2231669A - Color correction

Info

Publication number: US2231669A
Application number: US241433A
Authority: US
Inventors: Vincent C Hall
Original assignee: Eastman Kodak Co
Current assignee: Eastman Kodak Co
Priority date: 1937-01-16
Filing date: 1938-11-19
Publication date: 1941-02-11
Anticipated expiration: 1958-02-11
Also published as: US2253086A; GB538013A; GB488835A; GB538012A; DE697490C; US2249522A; FR51365E; GB552367A; FR834838A

Description

Feb. 11, 19 41. v, c, HALL 2,231,669

COLOR CORRECTION Filfid NOV. 19, 1938 l NV ENTOR W BY {.4

ATTORNEYS Patented Feb. 11, 1941 UNITED STATES PATENT. OFFICE 2,231,669 COLOR CORRECTION Application November 19, 1938, Serial No. 241,433

4Claim8.

This invention relates to color photography and particularly to the reproduction of color photographs.

It is an object of the invention to provide a 5 method and means for duplicating color transparencies, particularly those made on color sensitive monopack films.

Monopack films having three superimposed color sensitive layers are in common use for maklng color motion pictures and still transparencies such as for lantern slides. Although these photographs reproduce the colors of the original subjects rather faithfully, intolerable departures from perfect color rendition occur when dupli- 15 cates thereof are made by direct printing onto a similar monopack. These discrepancies are believed to be due to the differences between the spectral sensitivities of the three photographic layers and the trlchromatic responses of the eye and to the lack of saturation in the dyes used.

It is an object of the present invention to provide a method and means for overcoming such discrepancies and also for correcting some of the 25 errors in the original photograph whereby the reproduction has improved color rendition.

According to the invention, better color rendition is obtained by controlling in the duplicate, the contrast and relative contrast of the three 3 color images, the rate change of contrast with density and/or the density of one color in accordance with the density of another color at the same point. In the specific form in which this latter feature results in a reduction in density of 3 one color separation image in accordance with the corresponding density of another color, it is similar to the masking method of color correction particularly as described by Murray and Morse in patent application Serial Number 120,964, filed m January 16, 1937. In the present invention, the

degree of masking may even be varied non-linearly with density of one or both of the colors if so desired.

'More specifically, the' invention consists of 45 scanning a color transparency with the three pri-' mary colors, converting the intensities of each of the color beams into electrical energies by means 01' photoelectric cells, amplifying these energies either linearly or in accordance with some power 50 function as described in my copending application Serial Number 234,422 filed Oct, 11, 1938. The energies are then modified, one in accordance with another, to correct for deficiencies in the dyes used and are then used to operate light 55 valves to control the quality of a substantially white light which scans a sensitive monopack film in synchronism with the scanning of the original photograph.

Although the procedure used in the present inventlon is similar to that described by Murray 5 and Morse with theaddition of the specific idea described in my copending application mentioned above, it is fundamentally difierent from either or these since the errors to be corrected are of an entirely different sort. Thisis due to the fact 1 that the present invention is concerned only with direct duplication of a color photograph onto a monopack film, rather than any process involving color separation negatives. The present invention is necessary because direct printing onto a monopack necessarily involves and depends on the scanning of this monopack simultaneously with the three primary colors. The discrepancies due to the lack of saturation of the dyes used are of a much smaller order of magnitude than those (corrected by the masking method of color correction) which are due to deficiencies of inks used in photo-mechanical color reproduction which requires color separation negatives to be made separately. There is still another type of error, namely that due to an over-all excess or deficiency of one of the-colors in the original photograph, which may be due to the sensitivity of the individual layers of the monopack, to the illumination conditions under which the original photograph was made and/or to the processing of the original photograph and which may be corrected by a method such as described by Richard S. Morse in a copending application Serial Number 208,632, filed May 18, 1938. This latter type of correction is concerned either with reproduction by direct duplication or by methods employing individual color separation negatives. It is quite distinct from and may be used along with the present invention by having a gray patch on the original photograph which is also scanned in the three colors and a correction for the departure of the patch from pure gray may be introduced into the amplifier circuits used in the present inven- 5 tion.

Other objects and advantages of the presentinvention will be apparent from the following description when read in connection with the accompanying drawing which shows one embodiment of the invention.

In the drawing a color photograph l0 and a color sensitive monopack are mounted on the same glass cylinder I! which may be rotated and moved along such as by a screw not shown. A 56 source of light i3 connected by leads I i to a power source is placed inside of the cylinder and together with a suitable optical system including a lens i5 and a beam splitter shown as three prisms ISR, ISG, and 16B, and three color filters HR. (red), HG (green), and HB (blue), provides means for scanning the photograph in the three primary colors. Alternatively the original may be scanned by reflected light in any known manner. The intensities of these three scanning beams are converted respectively by photoelectric cells I8R, I 8G, and I813 into electrical energies which by suitable amplifiers i9R, 19G, and I9B provide means for operating light valves UR, MG, and ZIB. The amplifiers ISR, etc., may be linear or non-linear as described in my copending application mentioned above. I have found that it is preferable to use non-linear amplifiers whose response is a power function of .4 or .5 (i. e. approximately a square root function). This latter arrangement reduces the effective contrast by what is apparently about the correct amount.

In the drawing the output of the amplifier [9B is modified by a modifier 2028 in accordance with the output of the amplifier ISG. The output of this amplifier ISG is similarly modified by a modifier 208 in accordance with the output of the amplifier ISR. Although this modification is shown as being in the same direction as that often employed in the masking method of color correction Where correction'for deficiencies of photo-mechanical inks is required, this is not always the type of correction and is practically never the amount of correction which depends on the type of dyes used.

As a specific example I have found that my invention may be used with most monopack films now on the market when the following numerical values are chosen for the correction factors and correction function. The signal in the blue channel is modified by the signal in the red channel so that the density in the yellow layer of film H is reduced by about 30% of the density in the blue-green layer and-the signal in the blue channel is modified by the signal in the green channel so that the density in the yellow layer is further reduced by about 20% of the density in the magenta layer.

These correction factors are dlfieren-t for different dyes, i. e. for different monopacks. In general the correction in the yellow layer should be respectively from 20 to 35% of the density in the blue-green layer and from 15 to 30% of the density of the magenta layer. These correction factors are given in terms of the final densities since they may be produced by various types of modification of one signal by another. For example, if the signals are so amplified that they are proportional to the logarithms of the intensities of the original scanning beams, 50% of one logarithmic signal may be subtracted from another by merely superposing the two potentials in opposite directions across a. resistance (in this case modifiers 20B and 20G would be simple resistances with oppositely orientated superimposed potentials) or by any other suitable method. This gives the desired linear subtraction of densities. As a different example, if the signals are directly proportional to the intensities (or more exactly, proportional to a power of the intensities, e. g. blue current proportional to blue intensity and green current proportional to the cube root of the green intensity) the signals maybe simply divided one by the other to give the desired l near subtraction of densities. One arrangement for producing this cube root amplification is described in my above mentioned copendlng application. One simple way to get division is to provide modifiers 20B and 20G such that each consists of an adjustable resistance in one circuit with the adjustment proportional to the current in the other circuit. These types of correction and many others are also available with the modifiers described in the Murray and Morse application mentioned above.

Considering the signals at some stage where they are linearly proportional to the intensities of original scanning beams (i. e. before any loga rithmic or power function amplification has been applied) and assuming that the printing beams, except for the correction factor, are arranged (by suitable amplifying circuits and corresponding valve responses) to be proportional to these linear" signals, one can easily show that the above defined limits in final density may be written in terms of the linear signal intensities as follows: Density of yellow minus 20% to 35% of the density of blue green means in terms of signal intensities that the intensity of blue is divided by the 0.2 to the 0.35 power of the intensity of red. Density of yellow minus 15 to 30% of the density of magenta means that the intensity of the blue signal is divided by the .15 to the .30 power (roughly the sixth root to the cube root) of the intensity of the green signal.

The above limits in density correction are, of course, not exact. Therefore any error introduced by assuming that the final layers are all developed to approximately the same gamma (contrast) is negligible. If the gammas are slightly different the exact correction factor would have to be multiplied by the ratio of the respective gammas. However, the limits are not exact enough to require consideration of this second order factor in determining them. It is mentioned merely to assure anyone interested in the mathematics of this type of correction that it has not been overlooked.

The amount of correction is, of course, not absolutely critical and I have found that correction factors anywhere in the ranges mentioned give faiirly acceptable results.

The light valves ZIR, etc., control the light beams provided by

lamps

23R, 23G, and 23B, and filters 22R, MG, and 22B are combined by some form of reversed beam splitter shown as a pile of photometer cubes 24 and

lenses

25, 28, 21, and 28. The cubes 24 include semi-transparent mirrors whose edges are shown by the lines 29 and III which combine the three separate color beams into one to pass through the lens 28. The beam splitter 24 includes a third cube 3! which has a semi-transparent mirror (not shown) by which a fourth color may be included if so desired. The diagonal along which the edge of this extra semitransparent mirror will lie in the cube 3| depends on whether it is more convenient to bring the light beam in from the top, the bottom, or from one side.

Having thus described a preferred embodiment of my invention, I wish to point out that it is not limited to the specific construction shown but is of the scope of the appended claims.

What I claim and desire to secure by Letters Patent of the United States is:

1. In an electro-optlcal system for reproducing a color photograph, the combination of means for scanning the photograph with the three primary colors, photoelectric means for transforming the intensities 0;! each of the three scanning beams into electric energies, means for modiiying at least one of the energies in accordance with at least one other of the energies, holding means having an area adapted to accommodate a color sensitive photographic monopack, means for scanning the area with light including the three primary colors in synchronism with the scanning of the photograph and means operated in accordance with the corresponding energies as modified for controlling the relative intensities of the three primary colors in the beam scanning said area.

2. In an electro-optical system for reproducing a color photograph, the combination of means for scanning the photograph with the primary colors, photoelectric means for transforming the intensities of each of the three scanning beams into electric energies, electric circuits connected to the photoelectric means for controlling the proportionality and the relative proportionalities of the energies to the intensities, means for modifying at least one of the energies in accordance with at least one other of the energies, holding means having an area adapted to accommodate a color sensitive photographic monopack, means for scanning the area with light including the three primary colors in synchronism with the scanning of the photograph and means operated in accordance with the corresponding energies as modified a color sensitive monopack with a printing light including the three primary colors and the photograph with the three primary colors, modifying the energy in one of the scanning colors from the photograph in accordance with the energy of at least one other of these colors and controlling the relative intensities of the primary colors in the printing light in accordance with the corresponding energies so modified.

4. The method of reproducing a color photograph which comprises synchronously scanning a color sensitive monopack with a printing light including the three primary colors and the photograph with the three primary colors, dividing the energy of the blue color from the photographs by the .2 to the .35 power of the energy of the corresponding red color times the .15 to the .30 power or the energy of the corresponding green color and controlling the relative intensities of the primary colors in the printing light in accordance with the corresponding energies so modified.

VINCENT C. HALL.