The photography of coloured objects

Front matter

PHOTOGRAPHY OF COLOURED OBJECTS

BY ■

C E KENNETH MEES DSc.

Preface

This book represents an attempt to pnt clearly the theory underlying the photography of coloured objects, and the application of that theory to' those branches of practice which are of the most immediate importance.

The need for such a book has been impressed upon me by my daily correspondence, and I have hopes that it may serve as a kind of text-book of the subject for many of my friends.

While not employing purely scientific nomenclature and phraseology, I have made no attempt to be “practical,” believing as I do, that the application of an ounce of accurate knowledge is worth a ton of ttnreasoning practice. I have also made no pretence as to being unbiassed, I have frankly discussed the products of my own firm, but I hope that the loss of generalisation caused by this procedure will be compensated by the advantage to be gained from definite information.

It would not be possible, for instance, usefully to discuss the distribution of sensitiveness in “ red-sensitive plates,” whereas it is quite possible to discuss the distribution of sensitiveness in any particular plate. I am also likely to be biassed in argument, but I have not allowed myself to be co7isciously biassed, and have tried to be as fair as possible.

I am very much obliged to many friends for the services chronicled below.

Chapter VI. on “ Photography for Reproduction,” was written by Mr. A. J. Newton.

iii

IV

PREFACE

Chapter VII. on “ Eandscape Photography,’’ was largely drafted by M. Andre Callier.

Chapter V. was criticised by Mr. F. M. Sutcliffe, who made very many suggestions, and also by Mr. H. A. Hutchinson, Mr. Furley Eewis, Mr. E. A. Salt, and Mr. W. H. Smith.

Chapter VII. was criticised by M. Andrd Callier, Mr. J. C. S. Mummery, Mr. W. T. Greatbatch, and Mr. W. R. Bland.

The line illustrations were specially drawn, with the exception of fig. 22, which is taken from A. Konig’s paper on “Sensation Curves” {Zeitschrift fuer Psychologies iv. 241).

For the half-tone illustrations, I am indebted to M. Callier for the frontispiece and the other two mountain views, to Mr. Hutchinson for the landscape, and to Mr. Greatbatch for the meadow view.

The furniture photographs I owe to the kindness of Messrs. Maple, Etd., and of Mr. Erwood ; the rest, with the exception of the portraits by Mr. Barraud, and the photographs of the Maureta7iias which I owe to Mr. Edgar Eee, of Messrs. Thompson & Eee, were made in my laboratory.

’ The cover is from a design by a student of the London County Council School of Photo-Engraving and Lithography.

C. E. Kenneth Mees.

CHAPTER I The Nature of Colour

At the commencement of this book, which is essentially concerned with the analysis and photography of colour, it will be well for us to get a definite idea as to what is meant by ‘ ‘ colour, ’ ’ and with what physical phenomena colour is associated.

The nature of colour is involved in the conception we obtain as to the nature of light. The nature of light has long been a source of speculation, and it was generally held that perception of light depended on the reception by the eye of small discrete particles shot off from the source of light ; just as at one time it was held that the perception of sound depended upon the impact upon the ear drum of small particles shot off from the sources of the sound. This theory of light has the advantage that it immediately explains reflection ; just as an indiarubber ball bounces from a smooth wall, while it will be shot in almost any direction by a heap of stones, so these small particles would rebound from a polished surface, while a rough surface would merely scatter them. This theory of the nature of light appeared adequate until it was found that it was possible, by dividing a beam of light and slightly lengthening the path of one of the halves, and then re-uniting them, to produce periods of darkness similar in nature to the nodes produced in an organpipe, where the interference of waves of sound is taking place. It could not be imagined that a reinforcement of one stream of particles by another stream of particles in the same direction could produce an absence of particles, while the analogy with sound suggested that, just as sound was known to consist of waves in the air, so light also consisted of waves.

Light cannot consist of waves in the air, partly because we know that it travels through interstellar space, where we imagine that there is no air, but also because the velocity of light, nearly 200,000 miles per second, is so great that it is impossible that it could consist of a wave in any material substance with which we are acquainted. It is, however, supposed that there must exist, spread through all space and all matter, a substance which is termed the ether, and that light consists of waves in this ether.

Now, just as in sound we have wave notes of high frequency, that is, wuth many waves per second falling upon the ear, which form the high-pitched or shrill notes, and also notes of low frequency, where only a few waves per second fall upon the ear, forming the bass notes — so with light we may have different frequencies of vibration, some falling upon the eye at very short intervals, while other waves are of only half or even less frequency.

Since the velocity of light is the same for waves of different frequencies, it is clear that the waves of high frequency will be of shorter wave length than those of low frequency, the length of a light wave being the distance from the crest of one wave to the crest of the next.

The wave length of the light, like the velocity, will vary with the medium in which the light is travelling. For instance, when light is travelling through glass, it will only have about two-thirds of the wave length of the light travelling in the air. But it is convenient to consider simply the wave length of light as the length of the wave in free ether, or for practical purposes, in air. White light consists of vibrations of many degrees of frequency, i.e. it consists of waves of various lengths; and a mixture of waves of all lengths in certain proportions, forms what we term white light. If instead of allowing this heterogeneous mixture of waves to fall upon the eye, we

NATURE OF COEOUR omit waves of some frequencies from those entering the eye, then the brain will receive a sensation of colour. Thus colour is associated with wave

length. White light being WAVE LENGTHS

made up of waves of different ^ .

lengths may be regarded as

being made up of light of RED

various colours, and by different devices may be split up into these colours. When this is done there is obtained what is known as the spectrum.

Fig. I shows the average

length of wave corresponding to light of various colours, the diagram being drawn to scale.

Since different lengths of wave correspond to different colours, the spectrum corresponds to a B L U L

scale of different length waves.

The following diagram gives a simple arrangement of the spectrum, the numbers repre- Fig. i

senting the length of the waves

in Angstrom Units (A.U.), which are tenmillionths of millimetres, and the colours being placed against them :

BLUE VIOLET

BLUE

T

ORANGI

RED raiDV

f

Fig. 2

It will be seen that the visible spectrum extends from 7000 to 4000, and is equally divided into regions which may be broadly termed — •

4 PHOTOGRAPHY OF COTOURKD OBJECTS

If we make a filter which only lets through the portion of the spectrum between 6000 and 7000, then we should call that filter a red filter, a filter letting through from 5000 to 6000 would be a green filter, and a filter letting through from 4000 to 5000 would be blue-violet in colour. Thus from the spectrum we already derive the idea that light can be divided into three colours which we may call the primary colours, red, green, and blue-violet.

Remembering this conception of light, let us consider why we term a given filter red. It will appear red because it only lets through red light, but white light consisting of red, green, and blue-violet is falling upon it, so that clearly it is red because it stops or absorbs the green and blue-violet light.

Similarly, a piece of red paper is red because it reflects red light, but it has falling upon it white light consisting of red, green, and blue-violet, so that it must absorb the green and blue-violet light, not reflecting them, and only reflecting the red light. We are therefore justified in saying that anything which absorbs green light and blue-violet light together will be red.

It is this aspect of colour, that objects are coloured because they absorb, which must be clearly and definitely understood if the best results are to be obtained in the photography of coloured objects. Unfortunately, however, the conception of colour as an absorption is not common, though I believe it to be the most usefuFone, and it will be necessary for me to somewhat elaborate this subject in order to prevent misconceptions arising. It seems to me that we should form the habit of considering a red object, not as

one that reflects red, but as one that absorbs green and

blue-violet.

The importance of this definition is that it defines ‘ ‘ red ’ ’ without reference to the colour of the incident

light. Take a scarlet book and examine it by a light

containing no red ; such for instance as the mercury vapour lamp, in which red is almost entirely wanting. The book will no longer reflect red light because there is no red light for it to reflect, but it will still absorb the green and blue-violet light of the lamp, looking black ; it has not changed its nature, and we should still

THE NATURE OF COLOUR be justified in saying that it is red if we define red as we have done above.

In the same way a yellow object is not one which reflects yellow light (there is very little yellow light indeed in the spectrum, and if an object reflected only yellow light it would be so dark as to be almost black), but a yellow colour is due to blue absorption. It reflects the other two components of white light, green and red, so that we should be justified in saying that yellow light consists of green light plus red light, but for our purpose I want to consider yellow simply as a lack of blue ; yellow is minus blue,

BLU L

RED

YEL 1 GREEN

.0 W RED

^BLU^

RED

blue:

RED

Fig. 3 so that if you have a beam of yellow light and add blue light to it, you will get white light.

Now what is green? Well, since white light consists of red light, green light, and blue light, green is clearly

white light minus red and minus blue ; and a green

body is one which absorbs both red and blue. The

difference between a green object and a yellow object

being that the yellow object absorbs blue only, whereas the green object absorbs some of the red light which the yellow object reflects.

We can now make clear what is meant by complementary colours. As is shown in the diagram (fig. 3), white light consists of red light, green light, and blue light. The next section under this shows the blue blotted out, leaving the mixture of red and green — that is, yellow. We should say then, that yellow is complementary to the blue-violet. In the same way, in the next diagram all green and blue are blotted out, leaving only red, so that red is complementary to green-blue. In the bottom diagram all red and blue are blotted out, leaving only green ; green is complementary to this blue-red mixture, which is usually known as magenta.

In general, then, the light absorbed by an object may be said to be complementary to that reflected by it.

So far we have only considered intense colours. We have imagined that a red object absorbs the whole of the green and the blue-violet light, that is to say that its absorption was complete. But most things have only partial absorption — the absorption is incomplete. Partial absorption can be of two forms : it can be gradual, or it can be sharp ; thus, if you take a photograph of a spectrum, and put in front of the spectroscope a solution of erythrosin, then that erythrosin will cut a clean patch of red out of the spectrum between 4800 and 55CX), as is shown in the photograph (fig. 4). But if you put in front of the spectrum a cell containing gentian violet you will get a very gradual diminution of intensity between about 4700 and 6200, with the least light transmitted about 5800 (fig. 5). Thus different dyes and different substances give different classes of absorption, the two kinds being roughly subdivided into (i) sharp absorptions, and (2) gradual absorptions.

Let us examine the effect of a single sharp absorption band in different parts of the spectrum. First, consider a sharp absorption band situated in the red about 6^00 and producing a total absence of red in this part. The remaining colour consists of all the blue-violet and all the green, with some of the red. The actual visual effect of the mixed colour is what one might term a “ sky-blue.” Imagine this band to shift so as to absorb the orange ; absorbing between 5800 and 6200, the colour will be a light violet-blue, because there is a great deal of red being transmitted and less green. If the band shifts into the yellowish green from 5600 and 6000 it will absorb a great deal of the gxeen and none of the red, and the colour will become bluish purple ; as it

Fig. 4

Erythrosin Absorption Spectrum

THE NATURE OF COEOUR shifts lower in the green towards the blue this purple becomes a reddish purple, so that when the band is situated at from 5600 to 5200 we have what is generally known as magenta in colour. As the band shifts towards the blue, the blue fades out of the magenta, green taking its place. When the band is from 4700 to 5200 the colour is a sort of orange, and as the band moves into the blue- violet the orange becomes a yellow, and finally a lemon-yellow. So

'CHART SHOWING RESIDUALS POSITION OF ABSORPTION BANDS

LtM ON — YELLOW

YELLOW

ORANGE

MAGENTA

PURPLE

VIOLET- BLUE

SKY— BLUE

Fig. 6

that if we imagine a single band to pass down the spectrum, we get a change from light sky-blue through purple, magenta, orange, and yellow, to lemon-yellow (fig. 6).

Now it will be seen that there is one class of colour which does not enter at all into this series, namely, the greens. There is really no visual suggestion of green in any colour formed by using a daylight spectrum and absorbing one narrow band only. In order to get a green we must have an absorption both in the red and in the blue. If we absorb the extreme red and also the extreme blue, we shall at once get a green, and as these two bands vary with regard to each other, we shall obtain various shades of greens. Thus, if the red absorption band is very strong and the blue absorption band is weak, we get blue-greens ; if the red absorption weak, and the blue strong, yellow-greens.

Green is almost the only common colour due to two absorption bands, and other colours which on analysis prove to have two absorption bands generally tend to be mere variants in hue of some colours which we have alread}^ discussed under the heading of single absorption bands. A brown colour is fairly common, and the bands of a brown are of a gradual absorption type generally extending through the blue-green with a transmission band in the violet — that is to say, a brown is really a degraded orange, and is a variant on the colour described as orange, resulting from a single absorption band in the blue-green.

Natural colours, however, do not present these sharp absorption bands, the absorptions being always more or less gradual. Thus, in fig. 7, we see a sharp absorption band depicted, which would give rise to a violet-blue colour. An actual violet-blue, however, will have an absorption band of the type shown by the shading in the opposite direction on the diagram. Similarly, fig. 8 shows an ideal green, having sharp red and violet absorption and an actual green with its gradual ending and absorption of the green itself.

The sharpness of absorption bands is of great importance in respect of the luminosity of the colours produced by them. The side of an absorption band, which is toward the red end of the spectrum, generally has a sharp edge, as shown in fig. 7, while that which is toward the blue end has a gradual edge, a considerable amount of absorption remaining even in the transmitted portions of the spectrum. As a result, colours which are bounded by the sharp edges — that is, reds, oranges, and yellows, are bright colours, while colours which are bounded by the gradual edges — blue-greens, blues, and violets, are dark colours. A green will, as a general rule, have a sharp edge at its blue limit and a

THE NATURE OF COLOUR 9

gradual edge at its red limit, and will consequently be of intermediate brightness.

If we divide the spectrum at 5000 and at 6000 so that we get three portions, 4000 to 5000 which we may term blue-violet, 5000 to 6000 which we may term green, and 6000 to 7000 which we may term red, then examination of the luminosity curve, given in Chapter II., fig. 9 will show that about f of the whole light should be ‘ ‘ green, ’ ’ about 3% should be “red,” and about 3^ should be “blue.” But inasmuch as a bright red object will reflect nearly all the incident

Fig. 7

Theoretical and Actual Violet

Fig. 8

Theoretical and Actual Green

“red” light, while a bright green object will only reflect about ^ of the “green” light, and a bright blue object of the “ blue ” light ; a red object will be the brightest, a green object less bright, and a blue object very dark indeed.

A yellow, having only a single sharp absorption edge, is very bright. A yellow object usually reflects even more red light than a red object, and much more green light than a green object.

I have made a number of measurements of the absorption by various filters and colours, of the light

10 PHOTOGRAPHY OF COI.OURED OBJECTS

which they are supposed to transmit or reflect, with the following results :

Pure Dye Filters

Transmitting.

Region for which Transparency was

Trans-

parency,

5900 to red end (tricolour red).

measured.

per cent.

6000 to 4800 (tricolour green) .

35*5

5100 to 4000 (tricolour blue) .

11*5

5100104000 ,,

4700 to 4000 (D. methyl violet)

5600 to red end (E)

5100 to red end (G)

4600 to red end (K2)

72*5

4600 , , . .

K3

K3

K3

K3

Light naphthol green (about 6500 to 4500) .

Dark naphthol green

14 ‘5

Xylen red (5 100 to 4000), purest blue obtainable .

The chief points of interest

are the luminosity of the

(E) ; and red I

(A) ; the

darkening in the greens, and even more in the

blues and

violets.

Printing

Bright scarlet

Inks

83*5

6100 ,,

Bright light blue .

Dyed Wools

These were given me by Dr. E. Konig, of Hochst a/M,

and he stated that he considered them to be very good pure colours.

THE NATURE OF COLOUR ii

Colour.

Region of Measurement.

Percentag

Reflected

Dark purple

4800-4000

9*5

Bright ,,

4800-4000 n n

6500-red end

Dark blue .

4700-4000

Light „ .

5100-4000

Dark blue-green .

5000-5800

II Light „

4900-5400

Dark yellow-green

. 5000-5800

17*5

Bright

6100-4800

26-5

5000-5800

Yellow

5100-red end

Orange

. 5600

■57*5

• 5900

Scarlet

• 5900

6100 , ,

Bright scarlet

. 6100 ,,

Deep red

6100 ,,

. 6500

Chapter II

The Sensitiveness to Cotoured Eight of the Eye AND OF Photographic Prates

WE have seen that the eye distinguishes light of different wave lengths by the production of an appearance of colour, i.e. a ray of light containing waves of a length of 4600 of our units would be called violet, and would be said to be of a violet colour, while if the waves were of the length of 6500 they would be said to be deep red in colour. But the sensitiveness of the eye is not the same for waves of different lengths. The eye cannot per-

Fig. 9 ceive at all waves below 4000 units, i.e. what is known as ultra-violet light ; neither can it perceive rays which are above 7000 units, so that to these we must regard the eye as insensitive. The eye is very little sensitive to the extreme violet rays between 4000 and 4500. The blue affects it

more and appears, as we say, bright. Between 5000 and 6000 the green appears as the brightest part of the spectrum ; above 6000 we have the bright reds, but the intensity rapidly falls off as the waves get longer, until beyond 7000 we see practically nothing. We may then draw a curve showing the sensitiveness of the eye to the spectrum. It will be noted that this curve has a maximum at about wave length 5900, but this only holds for intense light. As the intensity of the light diminishes, not merely does the eye see less, but the relative sensitiveness of the colours changes somewhat, shifting towards the blue. This is what is known as “ Purkinje’s Phenomenon.” The explanation offered for it by Professor Schaum is sufficiently interesting and little known to be worth repetition. It is known that the retina consists of rods and cones, of which the cones are considered to be colour-sensitive, and the rods colour-blind. In the part of the retina exactly opposite the centre of the pupil there is a small depression which contains no rods, but only cones, and here it is found that the Purkinje phenomenon is non-existent, so that the intensity maximum remains constant. So that we may conclude that the colour-sensitive cones alone display no Purkinje phenomenon, and that the phenomenon is due to the association of these cones with the colour-blind rods. It is found that the sensitiveness curve for this region containing only cones is identical with the curve of sensitiveness for great intensities of light, so that this is the curve of the cones. On the other hand, since the rods are much more sensitive to feeble intensities of light than the cones, as is shown by the fact that the sense of light remains after colour can no longer be distinguished, the sensitiveness curve of the rods will correspond to the curve for minimum intensity ; so that for minimum intensity the sensitiveness curve is due to the rods alone, and as the intensity grows the curv^e is more and more influenced by the cones, until with maximum intensity the curve of sensitiveness is almost entirely determined by the cones.

Just as the eye is unequally sensitive to light of different colours, so a photographic plate is unequally sensitive to light of different colours. But if we take an ordinary^ photographic plate and measure its sensitiveness, we shall find that it differs very markedly from that of the eye. The eye can see waves of no shorter length than 4000 units ; a photographic plate can see very much shorter waves, and can detect light which is quite invisible to the eye, this light being usually called ultra-violet light, because it is beyond the violet. Also

SENSITIVENESS OF THE EYE AND PLATE 15 the maximum of sensitiveness of a plate is in the violet^ and all the red-orange and nearly all the green light is invisible to it. The eye may be said to perceive objects mainly by the green and orange light which they reflect. The ordinary plate perceives objects by the blue and violet light which

Allochrome Erythrosin Plate

Plate Sensitiveness Curves to Daylight Fig. lo

they reflect, and this is a great difficulty in the photographic plate when regarded as an instrument for perceiving and recording coloured objects, because the record which a photographic plate makes of coloured objects differs entirely from that which the eye makes.

It was found by Vogel that, by treating plates with dyes.

i6 PHOTOGRAPHY OF COLOURED OBJECTS

they could be given, besides their usual sensitiveness, a secondary sensitiveness in approximately the region of the spectrum which those dyes absorb. Thus if a plate is soaked in a solution of Erythrosin, which absorbs the yellowish green, it will be sensitive to the yellow-green, besides being sensitive to the blue and violet. Plates which have been treated in this way are those which are known as “ orthochromatic, ” the word implying that they can render objects in their true colour values. The ordinary commercial orthochromatic plate, which is made by putting some Eosin or Erythrosin into the emulsion, has a sensitiveness curve of the type shown in fig. lo, and it will be seen at once, on comparing this with the sensitiveness curve of the eye, that, although the plate is certainly better in consequence of this treatment with Erythrosin, it cannot be described as at all comparable in sensitiveness with the eye. It has an enormous excess of sensitiveness in the blue and violet, it has the sensitiveness to the ultra-violet which the eye has not at all, it then has very little sensitiveness indeed to the blue-green, a small maximum of sensitiveness in the yellow-green, and an absence of sensitiveness to the red. It may be assumed that if we take the blue to include the whole spectrum up to 5000, the green to be the spectrum from 5000 to 6000, and the red from 6000 upwards, that the sensitiveness of the ordinary orthochromatic plate is distributed in the ratio of 40 parts in the blue, one part in the green, and none in the red. If we consider that the eye sees the three parts of the spectrum as of equal intensity, then the orthochromatic plate besides the fact that it is not sensitive to the red, has only of the sensitiveness in the green that it would require to be equal in sensitiveness to the eye.

If, however, instead of sensitising a plate in the way we have described, we bathe the finished plate in a solution of certain of the new dyes called Isocyanines, we can prepare a plate which is very much more sensitive both to the red and to the green. Two years ago Messrs. Wratten and Wainwright. Ltd. , succeeded in preparing a plate in this manner, the plate being sensitive to both red and green, and this plate they called the “ Wratten Panchromatic Plate.” The plate is sensitive to the whole visible spectrum ; although it has a considerable excess of sensitiveness in the blue, this

SENSITIVENESS OF THE EYE AND PEATE 17

excess is very much less than in the case of the ordinary orthochromatic plates, and there are no absences of sensitiveness throughout the whole spectrum. The distribution of sensitiveness in this plate is also shown in fig. 10, and it may be said that ^ of its sensitiveness is in the blue, in the green, and in the red ; so that the sensitiveness to blue is seven times too great compared with the rest of the spectrum, while the sensitiveness to green and red together is of that required to have the same sensitiveness as the eye.

In order to attain the same relative sensitiveness as the eye, it is necessary with an ordinary orthochromatic plate or with the panchromatic plate, to use absorbing colour filters which shall diminish the excess of blue light, and it is the consideration of these colour filters and of the effect which they will have on the total sensitiveness of the plate, that is, on the exposure required, which must ^occupy our next chapter.

Section I.— The Multiplying Factor of any Sharp-cut Filter

Suppose that we have a filter which has a perfectly sharp absorption — that is to say, which cuts a clean section out of the spectrum, passing only light between two definite wave lengths, and without any absorption of that light — then, if we wish to find the multiplying factor of this filter, we must consider it in relation to the sensitiveness curve of the plate.

It will be convenient first to consider a filter which does not transmit light below 5000 A.U., i.e. which absorbs the

Sharp-cut Filter on Erythrosin and Panchromatic Plates

Fig. II wdiole of the blue-violet and ultra-violet, but does not absorb any green or any red. This filter will be a bright yellow in colour, yellow being, as we have seen, made up of green light and red light — that is to say, yellow being simply an absorption of blue. Consider the effect of this on an orthochromatic plate which has of its sensitiveness in the blue and ^ in the green. The yellow screen will remove all the blue light, i.e, |-§ of the active light, and it will increase the required exposure 40 times, so that it is what we term a 40 times screen.