Of the many varieties of small portraiture, the vignette is, perhaps, the most popular, and, when well done, is certainly the most refined and delicate. Two things are to be especially avoided in vignetting. The form of the vignette should not follow the form of the figure closely, as it is too often made to do, and dark backgrounds should not be employed. The qualities to endeavour to attain are softness of gradation, and an arrangement of the forms of the vignette that shall throw out the head and figure, and the resulting print should somewhat resemble a sketch, finished if you like, but sketchy in effect. Although the background should be light, it ought not to be white, but of a tint that would just throw up the white of a lady's head-dress. If the background screen could be painted so that a little shade should appear over the shoulders of a sitter for a head, or rather darker behind the lower part of a three-quarter figure, so much the better would be the effect. It would be difficult to find a case where gradation could not be of advantage in a background; however slight, it conduces especially to relief.

Having stated what should be aimed at in vignettes, we now come to the technical methods of producing them.

In many cases vignetting is considered to be a merely mechanical operation, and very often looks like it. Perhaps the trade have more to answer for than the printer, since the qualities of the wares advertised for the use of the vignetter are often exaggerated to such a degree that they are supposed to be suitable to any pictures. Vignette glasses are not so common as they used to be, but they certainly are useful in some instances; we almost think that the methods of producing vignetting apparatus which will be described shortly, superior to them. In case the printer should wish, however, to use these glasses, here is a method by which he may produce them. Have a piece of orange glass, flashed on one side only, rather larger than the size of the picture to be vignetted. Take a rough print, and trace round, in the proper position on the glass with an ink line, the point to which the picture should extend. This should be marked on the unflashed surface of the glass—that is, the surface on which the glass is uncoloured. Place the plate so marked on a white surface, flashed side uppermost, and make a solution of hydrofluoric acid and water, 1 part of the former to 3 of the latter, in a gutta-percha dish or bottle.[22] Make a pad of flannel and cotton wool at the end of a stick, about the size of a large nut, and drop this into the solution. Dab this on the coloured surface of the glass in the central portions where the print is to be completely printed in, gradually working out to the inked line. Always work from the centre to the edges, and dilute the acid with a little water as it approaches the margins. By degrees the flashing will be dissolved away in the centre, and, if properly performed, the colour will gradually be eaten away, till the glass is colourless in the centre, and keeping its full shade of orange at the ink lines. The glass is then washed, and is ready for use. The most popular plan of vignetting is with cotton-wool. We believe that the greater part of the vignetting done in England is by this clumsy, costly, and difficult method. It requires more time and attention than any other way of producing the same results. Its advantages are, that it is more "elastic," and allows the operator more scope for attention than other methods. In the hands of a person who has very great skill, taste, and patience, it is undoubtedly most useful; but when used by anything lower than the highest skill, the results are almost always hard and inartistic. The operation is thus performed. A hole is cut in a piece of cardboard, which is placed over the negative. Under the edges of the cardboard is placed cotton-wool, which is lightly pulled out, so as to slightly shade the vignette, and produce the vignetting gradation.

The next methods of vignetting are dependent on simple laws of optics. Suppose you cut a round hole in a card, say, half-inch in diameter, and so arrange it that all the light getting to a sensitive paper comes through this hole, and that the card is for our experiment placed half-an-inch from the paper. Now place the hole so as to face the sky, but so as the sun has no direct rays falling through the hole. It will be found that the greatest darkening will not occupy a space exactly opposite the hole, but be away from the side on which the light is brightest. The dark round patch will be shaded gradually off till a line is reached where, practically, the light has no effect—that is, if the surface of the card next the paper be blackened. It will be noted, however, that the shading is not equal on both sides, but that the gradation is most extended away from the side on which the light is brightest. A good example of what is meant will be to try the experiment of placing the paper and card flat on the ground in the angle between two walls, both of which are in shadow. It will be seen that the brightest gradation takes place in the direction exactly away from the angle of the walls. Next repeat the experiment, making the hole point to the sky, which is equally illuminated and pointing well away from the sun. It will be found that the gradations are equal, and the greatest darkening exactly opposite the hole. Raise the card next to the height of one inch, and the gradations will be found to be more extended and softer. The reason of this can be well understood by a glance at the figures. In both, suppose A B to represent the section of the card, and C D the hole in it, and the dotted circle the sky, and E F the paper. Take the points a, b, and c on the paper, and let us in the three instances see what relative illumination they will receive. a is opposite the hole, and receives the light from a circle of sky of which d e is diameter, and b from an ellipse of which h k is one diameter, and e from an ellipse of which f g is one diameter. In the first case, where the card is ¹/₂ inch from the paper, h k is about one-fifth of d e, and g f about one-third of d f, and since the ellipses vary as their two diameters multiplied together, the point b would receive only one-twenty-fifth the light that a received, and c about one-ninth.

In fig. 21 the card is raised one inch from the paper, and here f g is about three-quarters of d e, and h k about two-fifths; therefore, in this case, the light on B would be only four-twenty-fifths, or about one-sixth of that acting on a, and about nine-sixteenths or one-half nearly on c. It is thus evident that the further away the card is, the more extended will be the gradations. Again, suppose, in the last figure, the bit of sky at g f was twice as bright as at d e, then the amount of light acting on c would be the same as that acting on a. It will thus be seen how important it is for proper gradation that the hole in the card should be exposed to an equally illuminated sky, or that some artifice should be employed to render the illumination equal. If we paste a bit of tissue paper over C D, this is accomplished, for then it becomes the source of illumination, and it is illuminated equally all over, since on every part it receives the light of the whole sky; but this is not the case if it is transparent to diffused light, and is never the case if it is exposed to direct sunlight, since a shadow of the hole is always cast on the paper beneath. If you choose to put another piece of tissue paper, (say) one inch above the hole, and extending over the whole length of the card, this difficulty is got rid of, and this last piece of tissue paper illuminates that pasted over the hole C D, and the gradations will then be nearly perfect.

Now to apply the above to forming a vignetting block.

Suppose we have a one-inch head to vignette and to show the shoulders and chest, to be of the size of a carte-de-visite, that the background is about a half-tone between black and white, and that but a trace of it shall appear above the head. To make a good vignette, the graduation from black to perfect white should lie within a limit of half an inch for a carte size portrait. The question then arises at what distance from the plate should a vignetting card be cut to help this object, and what shape should be made the hole in the card. We take it that one-fifth of the light necessary to produce a full black tone would hardly produce any effect on the sensitised paper; knowing this and the size of the aperture, we can calculate exactly what height the card could be raised. Take the breadth between the shoulders that is to be fully printed as 1¹/₂ inches, then by constructing a figure similar to figures 18 or 19 we shall find that the necessary height is about one-third of an inch.[23]

By judiciously cutting out an aperture in the card and vignetting, defects in a background may often be entirely eliminated from the print. Proceed in this way: Take a print of the portrait, and cut out the figure in such a way as to get rid of the defective background, and then place this on a piece of thick card (we prefer a thick card, since it will not sag easily, and thus alter the gradation), and cut out an aperture corresponding to it. The outsides of most carte-de-visite frames are raised from the glass about one-third of an inch; place the card on the front so that the aperture corresponds to the figure on the negative, and tack it on to the frame. The dotted lines (fig. 23) show the card fastened on to the frame, and the opening left. This latter may be covered with tissue paper, and the frame placed in diffused light from the sky. In some cases it may be necessary to use a larger printing frame than the ordinary carte frame, in which case the operator should be able to make a vignetting apparatus raised at a proper height from the glass. Suppose it is required to raise the opening half an inch above the glass, and that the card is 4¹/₄ by 3¹/₂.

Take the card and rule rectangles as shown (fig. 24), the inner one being 4¹/₄ by 3¹/₂ inches, the next one ¹/₂ an inch outside that, and the third ¹/₂ an inch outside that again. Cut out the outer rectangle entirely, so that we have a piece of card of the size ABCD. Take a needle point, and prick through the card at the points EFGH and join these points at the back of the card by lines. Now take a sharp penknife, and, having laid a flat edge along, cut the card half through its thickness along KL, LM, MN, and NK. Turn the card over, and cut along the lines corresponding to EH, HG, GE, and FE, also half way through the thickness of the card. Turn the card over once more, and cut out the shaded pieces at the corners. Now bend the card along the cuts, and a raised block will result of this shape. The corners are held together by pieces of gummed or albumenized paper, and the block is ready for an aperture to be cut in it according to the portrait to be printed. Wooden grooves may be glued along the top of the vignetting frame, into which cards containing other apertures can be slipped.[24]

The most practical method of vignetting, a modification of the above, and the one we always prefer in our own practice, is as follows:—

Take a piece of soft wood, half an inch thick for a cabinet size—a thinner piece should be selected for a smaller picture—of a larger dimension than the negative; in the centre of this cut a hole of the shape of, but much smaller than, the desired vignette. One side of the hole should be very much bevelled away, as represented in this section (fig. 26). Place this block on the glass of the printing-frame, bevelled side under, the hole being exactly over the part of the negative from which the vignette is to be printed. The hole must now be covered with tissue paper or ground glass, and the frame placed flat on a table to print. The size of the hole in relation to the size of the vignette will be easily ascertained by a little experience without the labour of elaborate calculations. On dull days the tissue paper or ground glass may be omitted. This method is very simple and effective. A quantity of vignetting blocks of various sizes and shapes could be made by a carpenter, or by the printer, and should be always at hand.

A vignetting block should never be less than a quarter of an inch away from the glass, otherwise the gradations will be too abrupt.