Chapter 5
Photomechanical Reproduction

This chapter discusses the mechanics of printing, etched Daguerreotypes, photomechanically-reproduced pictures including collotype and its derivatives, photogravure, Woodburytypes, and some recognition factors.


There were several mechanical printing processes that produced pictures difficult to distinguish from photographs, the latter as defined in the Preface. Omitted from this chapter are pictures that obviously are not photographic in origin, such as Currier and Ives lithographs, wood cuts, and line engravings. This is usually evident from angles different from camera perspective, hand-executed shading and other artificialities. However, descriptions of the photographic processes that have played some part in their reproduction are included to clarify identification.

The oldest known photograph in existence was made in 1826 by the Frenchman Nicephore Niepce, who was searching for a way to reduce the labor in engraving lithographic stones and plates. This photograph consisted of a pewter plate coated with bitumen of Judea (see Glossary under asphaltum). Niepce discovered that a thin layer of bitumen or asphaltum became insoluble in certain oils after exposure to light. The approximate modern equivalents of his materials are tar and turpentine, but Niepce was lucky in finding materials from the right natural sources that worked. These materials are mixtures of complex organic compounds whose composition varies with their origin, and Niepce's formula depended on bitumen from Judea and oil of lavender. Niepce's nephew recalled that his uncle first used Dippel's oil, a distillate from animal bone.

An all day exposure to light rendered the bitumen insoluble in the highlights; the unexposed tracts could be washed away, uncovering the base metal for acid etching. When the etched plate was inked and then wiped, the etched pits remained filled with ink transferred to paper on contact. This process, called intaglio printing, enabled Niepce to make the first known permanent image from nature, and many historians (not all, of course) have recognized Niepce as the inventor of photography. The act of pointing the way by showing that a thing is possible is the mark of historical greatness in many fields. Technology may be subsequently altered almost beyond recognition, but only after the original insight.

The oldest surviving specimen of Niepce's work is an etched plate from which many inked prints have been made. In 1827 he made a direct positive image by darkening the exposed metal in iodine fumes. This image has been copied and widely published; the original is in the Gernsheim Collection in the Harry Ransom Humanities Research Center at the University of Texas in Austin.

The Mechanics of Printing

Before the advent of photography there were three types of plates used to print illustrations:

1. Relief Plates.
Relief plates had their inked surfaces raised above the white level, like raised movable type. These plates were compatible with type: they could be clamped in a matrix with type and printed on the same page as text. Examples were blackline wood and metal engraving. Good halftones could not be produced; making the plates almost required the skills of a sculptor, since each line to be printed black had to be cut on both sides.

2. Intaglio Plates.
Intaglio plates had their inked surfaces cut below the white level; they were inked with rollers and then wiped clean on the top surfaces. Examples were steel and copper line engraving, enhanced by aquatint. They were not compatible with type, so illustrations had to be bound on separate pages from text.

Photoengraved intaglio plates are widely used in modern times; even our currency was at one time printed with them. One of the early problems was that wiping the excess ink tended to remove ink from large shadow areas. The problem was solved by dusting or owing a solution of resin on the plate and baking to melt the particles of resin. It was called a ground; the grains of resin provided tooth or roughness to hold the ink during wiping. This was the basis of aquatint, a somewhat misleading name since it had nothing to do with color.

Aquatint was sufficiently fundamental to be carried over into photomechanical processes. Resin in solution produced a "dried mud" pattern of connected lines similar to reticulated gelatin. For a more random pattern, resin was applied as a dust in dusting boxes, and fused to the plates by heat. Aquatint predated Talbot's gauze screen and was used to enhance printing quality in many intaglio variations.

Mezzotints, invented in the 1600's, were a variation of intaglio plates with good halftones. The blank metal plate was first roughened in a random pattern by a metal rocker with a serrated surface. Metal in the shadows was then removed by a skilled graver to varying depths.

3. Planar Plates.
Planar plates were the basis of lithography, which used flat porous stones that retained greasy inks and repelled water. It was not directly compatible with type, but type impressions in greasy ink could be used along with a picture impression by using transfer paper. Lithography, which dated from 1796, was in a sense a chemical means of transferring ink from either raised or intaglio plates, or from crayon sketches.

Compatibility with type was more important to book publishers than to print makers. Halftone reproduction in printing was a major problem. Raised type, inked on the flat top surfaces, produces printed characters with sharp edges. White is the absence of ink, while gray needs just a little ink, and raised type does not modulate the ink density. Individual picture elements either did or did not transfer ink, so density variations had to be achieved through spatial distribution and depth control. Hand engraving was limited by the minimum dimensions that picture elements could be cut. Wood blocks were easy to carve but wood grain limited them to coarse line drawings (coarse by modern standards; some wood cuts were quite pleasing). Steel or copper engravings could have very fine lines or dots if the graver was skilled and patient. Acid etching could save time in removal of metal, but the resist coating still had to be scribed with skill. All of the available processes produced illustrations that were obviously hand drawn artistic representations. The most skillful attempts at realism could not be mistaken for the photographic accuracy to which we are accustomed.

Composition of Printers' Inks
There are many formulations of printers' inks, but they fall into two categories: water based and oil based. Water based inks are somewhat like modern India drawing ink, made of colloidal carbon in water. They have a thin consistency, dry rapidly, and soak into porous paper. Greasy inks are thick, dry slowly, and can be retained in intaglio plates.

Carbon inks are blacker than most photographic images, which tended to be gray or brownish-black. The colors of silver images depend on the particle size of the reduced silver as well as changes in the binder and base. Carbon (the chemical element) does oxidize, but the rate at room temperature is negligibly small; it is much faster in a fire. At normal temperatures carbon is extremely unreactive with other materials. It does not change color, but it can flake off the paper. Dried India ink on smooth paper may show microscopic dried-mud patterns, depending on the degree of penetration, distinguishing it from most photographic emulsions.

It has been reported that some of Gutenberg’s Bibles in the 1400’s were printed with inks containing compounds of copper and lead. The characters are clear and glossy after five hundred years, while others of the same period that were printed with carbon ink are dull and crumbled. The observed differences in aging can probably be attributed to the properties of the binders and the degree of penetration in the paper.

Woodburytypes were printed with “ink” consisting of a water solution of pigmented gelatin. The pigments could be finely divided solid particles, but the gelatin was a colloid. Apparently this was the only printing process that used a medium unlike conventional printers' inks.

The Contributions of Photography

At this point the story gets more complicated. Many workers entered the field because photographically enhanced printing techniques had immediate commercial applications, even though Niepce had to struggle for recognition.

Niepce's invention was a labor saver, and his process was used commercially until the early 1850's, when bichromated gelatin was found to have superior sensitivity and ease of use. It was not type compatible, nor did it have good halftones, but it served to initiate efforts by a large number of workers.

Following are descriptions of the principal processes based on photography. The reprinted 1895 book by Denison [44] contains contemporary details of photogravure and other 19th century processes.

Etched Daguerreotypes
A fatal weakness of the Daguerreotype, besides cost, was the lack of a negative. A Daguerreotype could be rephotographed on another Daguerreotype, but this was expensive. Niepce's process produced multiple copies but was not widely used at this time. Daguerreotypes can be acid etched in their normal form, producing a weak intaglio plate (weak meaning that the etching was shallow and the resulting prints were low in contrast). Nitric acid etches the silver shadows, leaving the raised amalgam dots (see Appendix I for photomicrographs of a Daguerreotype surface). The surfaces were not durable, but could be reinforced by copper or gold plating. Only a few hundred prints could be pulled from the average etched Daguerreotype.

Prints were precisely the size of the parent plate (see Chapter 7), and the left-to-right reversal of the Daguerreotype was corrected in the prints.

Historians do not agree on assigning dates and priorities. Inventors often made announcements of a process and then delayed disclosing the details, either hoping to find financial backing or waiting for patent protection. Without details, other workers could not confirm the announced results. Alfred Donne of Paris and Josef Berres of Vienna were the first to show prints from etched Daguerreotypes in 1839 and 1840 respectively. Their results from simple etching were not of high quality.

In 1841 Hippolyte Fizeau of Paris produced good results by a more complicated process. After lightly etching a Daguerreotype he coated it with linseed oil and wiped it like an intaglio plate. Next he electroplated gold onto the plate, which adhered only to the elevated regions, since oil in the depressions prevented gold adherence. After cleaning, the plate was given a deep etch, the gold now acting as an etch resist. He was able to reinforce the halftones with aquatint resin, and obtained quite creditable quality. Some of his prints appeared in a travel book published in 1841, which was a rapid adoption of the new Daguerreotype process. M. Fizeau did not choose to name his pictures "Fizeautypes", although he might have been so justified.

Some references are: Crawford [38, 237-240]; Eder [48, 577- 580]; Gernsheim [61, 539-540]; Jussim [85, 49]; Newhall [105, 249]; Taft [140, 412].

Photo Relief Plates
The first successful photographic relief halftone process was patented in 1881 by Frederick Ives. This complex process produced good halftones with type-compatible plates. The images have a readily detected dot pattern that distinguishes them from photographs, but it was a landmark process made possible by photography. The Meisenbach process from about the same era also used a grating to produce relief plates.

Photography even aided one of the oldest relief printing processes, that of wood cuts. Photographic images were printed on wood blocks by the collodion process; the images served to guide the wood carver's perspective, but of course the resulting prints were still line prints.

Photoengraving, which produces type-compatible relief plates, should not be confused with photogravure intaglio plates. The later have superior halftones at the expense of incompatibility with type.

Photogravure, also called 'photo-aquatint', produces intaglio metal plates by acid etching through a photographically exposed etch resist. It is analogous to the hand scribed steel or copper engraved plates, hence the name "photo - grav - ure". It produces excellent halftones by substituting the greater detail and continuous tonal range of photography for the fine lines of hand engraving. Photogravures, dating from about 1879, more closely resemble photographic prints than any other photomechanical reproductions with the possible exception of Woodburytypes. The process is well described in Jussim [85].

Fox Talbot patented the first process in England in 1852, using potassium bichromate sensitized glue on steel plates (bichromate sensitizing is described in Chapter 4). Talbot at first used platinum chloride as the etchant; in 1858 he patented ferric chloride etching, still used today.

To solve the problem of ink removal from the shadows during wiping, Talbot used a black gauze screen between the positive and the bichromated resist to create an etched dot pattern. This was the first use of a halftone screen, the results of which can be seen in any newspaper picture today. Later he used powdered aquatint resin for the same purpose (however, this was not original; aquatint, as previously mentioned, dates back at least to the beginning of the 19th century.) Talbot thus laid the complete foundation for modern photo-gravure; he called it "photoglyph", having already used "Talbotype" for positive/negative photography. Figure 4 shows two views of a newspaper picture halftone pattern.

Newspaper picture halftone pattern Newspaper picture halftone pattern close-up
Figure 4


There were many variations. Paul Pretsch of Vienna patented in 1855 a bichromated gelatin-on-glass process based on swelling and reticulation in the shadows. The gelatin was molded in gutta percha and then copper electroplated. Campbell Duncan Dallas adapted this process for his Dallastypes, possibly infringing on Pretsch's patent, yet the name Dallastype has survived instead of Pretsch's. In 1879 Karl Klic of Vienna made copper photogravure plates both with aquatint grain and with a screened grain.

Improvements in photolithography continue to be made, particularly in random dot processes. There are now processes in which the number of dots in a given area depends on the amount of picture detail in that area, resulting in improved resolution.

References: Crawford [38, 243-268]; Dennison [44]; Eder [48, 593- 608]; Gernsheim [61, 544]; Jussim [85, 83; 303]; Newhall [105, 142]; Thomas [142, 94-95]; Welling [150, 85].

Collotypes are inked images on paper and are printed directly from light-exposed bichromated gelatin (a colloid, hence the name "collo-"). The halftones are good to excellent, and some of them can easily be mistaken for original photographs.

Alphonse L. Poetevin (France) patented in 1855 the first collotype process using bichromated gelatin. In zones exposed to light the gelatin hardens and no longer absorbs water; it will absorb a coating of greasy ink for transfer to paper. In non-exposed areas water absorption repels greasy ink. In this respect it is related to lithography, and it used lithographers' ink.

Josef Albert (Germany) improved the process in 1868 and renamed it Albertype, using a glass base. When Albertypes were printed on glazed paper they resembled glossy albumen prints; microscopic examination will show the collotype reticulation pattern. Albertypes were widely used for book illustrations and postcards.

The heliotype, invented in 1869 by Ernest Edwards in England, transferred the gelatin to a more durable metal plate.

The water content of the gelatin in collotypes was an important process variable, resulting in a curious historical sidelight. The process was said to work better in the European climate than in the United States. At that time the American industrial establishment was mostly east of the Mississippi River where the humidity is greater than some European locations. If humidity is detrimental to the process, it probably would have worked well in the arid American southwest, but industrial facilities were lacking there. Whatever the reason, the collotype process is still used in Europe but is practically unknown in the United States. On the other hand, solar enlarging on albumen paper was reported to work better in the United States because of more reliable sunshine. The early literature is filled with advice concerning conditions that appeared to influence the working of processes; sometimes the advice was correct for the wrong reasons. There were many variations of the basic collotype, some of which have been briefly described in Chapter 14, Section 3.

References: Crawford [38, 269-280]; Eder [48, 553; 594; 617-621]; Gernsheim [61, 540; 547-549]; Jussim [85, 72]; Newhall [105, 251]; Thomas [142, 96]; Welling [150, 85].

Walter Woodbury (England) patented his process in 1864; it was in worldwide commercial production until the 1890's (according to Jussim the French name was photoglyptie.) It was the only continuous tone photomechanical process, and prints were available in brown, red, green, blue, and other combinations; brown was commonest. The prints have a superlative halftone, excellent sharpness, no screen pattern, and a beautiful liquid depth. Of all photomechanical processes, Woodburytypes are the most likely to be mistaken for high quality original photographs. Their fatal disadvantage was that they were not type-compatible.

The process used bichromated gelatin on a reinforcing layer of collodion on glass; it was exposed through the collodion after being stripped from the glass. Hot water washed away the unexposed sections in proportion to the degree of exposure, which gave a relief pattern to the gelatin.

To this point the treatment was similar to that of other workers in bichromated gelatin, but his printing process was unique. The gelatin relief was pressed against a lead plate in a hydraulic press. Gelatin behaves like an incompressible fluid, and the soft lead received an accurate intaglio impression. For ink, Woodbury used a heated water solution of pigmented gelatin in the lead mold and transferred it to paper in a smaller printing press. Excess gelatin was squeezed out at the sides, and Woodburytypes had to be edge trimmed. Conventional intaglio ink printing removes the excess ink by wiping before the paper is applied, as described previously, making possible clean margins. Woodburytypes closely resemble toned silver bromide prints and especially carbon prints. All three types, if bound in a book, will be alone on their pages. There is a convenient identification clue: carbon prints were usually labeled "Permanent", while Woodburytypes were labeled "Woodburytype". Such straight-forwardness is salutary, but there were exceptions. Sometimes the legend was on another page, which is lost if the print is no longer in the original binding.

According to Crawford, the largest Woodburytype was 10 x 14 inches, though 7 x 9 inches or smaller was more common; the size of the hydraulic press was the limitation. Woodburytypes were always edge trimmed, and are more likely than carbon prints to show visible raised edges at light/dark boundaries under grazing illumination. A characteristic flaw in Woodburytypes is the presence of tiny dark specks in the highlights, caused by particles of dried gelatin carried over. Woodburytypes were usually a rich brown color, but the gelatin could receive any common pigment. Woodbury later introduced the stannotype, which was made with tinfoil instead of lead plates to eliminate the need for expensive hydraulic presses. The stannotype process was not a commercial success because of competing photographic processes, and the prints are not distinctive unless labelled.

References: Crawford [38, 270; 285-289]; Eder [48, 587-589]; Gernsheim [61, 340-342]; Jussim [85, 57]; Newhall [105, 251]; Thomas [142, 96]; Welling [150, 85].

Distinction Between Pattern and Grain
Prints made by photogravure and collotype can be recognized by their fine-structure. There are three basic patterns visible under low power magnification:

1. Geometrical dot structure characteristic of halftones. It may be cross-hatched, diamond, square, or round dots.

2. Collotypes show random connected lines in a worm-like or wrinkled pattern caused by reticulated gelatin.

3. The random particle pattern of aquatint and photogravure, a process that dates to about 1800 and was used to enhance many printing processes.

Woodburytypes and carbon/gum bichromate prints have no patterns, but there is confusion in some historical literature regarding grain. At least two sources refer to Daguerreotypes and Woodburytypes as "grainless", which is in error. It should have been said that they are lacking in visible texture
We live in a grainy world. A television receiver tuned to a distant transmitter displays what we call snow; in radio it is static. Applied generally to electronic communications, it is more accurately described by the signal-to-noise ratio. There is always grain present in photographs, comprised of image elements at discrete nucleation sites. Daguerreotypes were regarded as grainless in comparison with calotypes that had a paper fiber texture. The true nature of Daguerreotype grain is shown in the scanning electron micrographs in Appendix I.

The patterns in gravure and collotypes are reliable descriptors. When the pattern is geometrical it is unambiguous. Collotype reticulation pattern is also distinctive, but it may require microscopic examination to identify. Aquatint resembles photographic grain; both are random but aquatint particles are larger. Photographic grain is visible to the unaided eye in magnification or enlargements. Matte surfaces were often produced on silver bromide gelatin paper by mechanical stippling. It can be identified microscopically by the regular pattern of sharp-pointed indentations in the emulsion that do not cut through the emulsion.