Suite of Methods V: Making a Metalcut Matrix

Metalcut – a brief background

The production of a Metalcut is a very interesting process in both the making of the matrix and the printing, for in effect, it combines elements from both Relief and Intaglio. In producing the matrix, similar methods as for engraving is used, with the addition of metal punches. However, for the process of printing, the ink sits on the surface of the matrix, as in relief printing.

Metalcut developed in the late 1450’s, using a copper plate as the matrix. Lines were engraved and incised using an engraver’s burin, and then larger areas cut or scraped away. Heavy use was made of round and shaped punches and patterns for clothing and backgrounds, giving a flattened, decorative effect. Because of the use of punches, and printed evidence of nail holes in the corners of some extant prints, it is believed that at least some metalcuts were mounted on wood to stabilise the matrix during punching, and to keep the surface as flat as possible for printing.

Printing consisted of applying ink to the surface in the relief manner, however, it remains unclear to me as to whether the metalcut print was then produced by relief or intaglio printing methods. It is entirely possible that, given the time they began to appear, that the first metalcuts were hand rubbed, and then presses were used as they were affordable and available. As these were produced by engravers, it is also entirely possibly they were printed on intaglio presses.


Calvary Metalcut on vellum hand coloured image Bavaria 1460-80
Metalcut, hand-coloured, 18cm x 11.9cm
Bavaria 1460-80


Early metalcuts are almost exclusively religious in content. Later metalcuts show a wider range of subjects and became very similar in appearance to engravings, or to woodcuts depending on the style. Over time, the height of the wooden support for the matrix was made uniform to the height of moveable type, and metalcuts were often used as border and illustrations in books. However, given that engraving remained a highly-trained craft compared to woodcut, Metalcut production did not seem sustainable, and the method appears to have been discontinued by around the 1550’s.


Smaller Illustrations for a Hortulus Animae Image by Hans Holbein the Younger, cut by Jacob Faber Metalcut c1519–24
Illustrations for a Hortulus Animae
Image by Hans Holbein the Younger
Metalcut by Jacob Faber, image 5.8cm × 4.2 cm
Germany, c1519–24


Metalcut – producing a matrix by Intaglio methods for Relief printing

Image based on:


Cognicio Cognition from The Knowledge of God and the Seven Cardinal Virtues Sebald Beham Engraving
Cognicio/Cognition from The Knowledge of God and the Seven Cardinal Virtues
Sebald Beham


The process of making a metalcut matrix begins with a copper plate, with lines incised in the same method as producing an engraving. Further marks are made by cutting and scraping away the surface, and the use of punches. However, it must be remembered in planning the image that the ink will be sitting on the surface of the plate – the lines being incised will remain uninked and become the background, as in woodcut production.

As will be described when we come to the Intaglio method of engraving, my attempts at using an engraver’s burin met with almost no success. This is partially because there are years of training involved in engraving, and partly because of recurring tendonitis in my elbow and shoulder. The force needed to push the burin through the copper is too great to sustain without a reasonable chance of injury.

In order to achieve the look of a metalcut for the purposes of this study, I decided on a compromise. Because ink sits on the surface of the matrix, the line below the surface could, in theory, be made by any manner of incising by way of substitution. I therefore decided to replace engraving with etching, and see how the end result compares.

The image was initially prepared by the same method as most of the other methods. I traced a basic line off the Beham engraving using tracing paper, and then filled the image in with more detail. I also removed one of the books from the background to make more space for adding a decorative floor pattern, in keeping with the decorative nature of early metalcuts.

The next step was to prepare the plate for etching, by covering the metal plate with a ground, a thin layer of a substance which resists acid. This is discussed in more depth in the etching process. Once the metal is covered with the ground or resist, the image can be transferred by turning the tracing paper over onto the plate, and the back rubbed with a bone folder. Enough of the drawing is transferred to the ground to show as dull, silvery lines.

Areas which are needed to be bitten away from the surface are then revealed by scratching with a scribe or etching needle. The scribework here is exposing those areas which would, in metalcut, be cut away with a burin.


WM Smaller Cognitio transfer and scribing WM Smaller Cognitio ready for acid 1
Above left: traced lines appear dull silver, while scribed lines are shiny, exposing the metal and right: ready for the acid


The copper plate was then placed into a Nitric Acid (Aqua Fortis) solution and etched, or ‘bitten’, for about 20 minutes. This was done in increments of 5 minutes, with the plate taken out and observed each time. More ground was applied to any area which might have been overbiting, to protect that area whilst allowing other areas to bite more deeply.


WM Smaller Cognitio acid 1WM Smaller Cognitio acid result 1
Above left: plate in nitric acid and right: bitten plate with ground still on surface


Once out of the acid and the ground cleaned off, the edges were filed, and the plate was ready for proofing, that is, making/pulling a print to be able to see the development of the plate. As there was someone else in the studio using black ink with a brayer (printing roller), I decided to use this method of inking, although to the best of my knowledge, something like an inking ball would have been used.

Once inked, the plate was placed on the bed of a relief press and dry intaglio paper applied, and the print produced by the relief method. The image produced was not very satisfactory, and I felt that the dry paper was not making good contact with the matrix. I soaked the next piece of intaglio paper in water for a few minutes, blotted, and repeated the process, with a much better result.


WM Smaller Cognitio first proof
Above: first state, ie, first working proof of the metalcut, and the plate


The main issue with the plate at this point, though, was that I had been too conservative with how much ground I had left on the body. I had been hoping that there would be more light areas on the limbs and face. These areas would have to be scraped away with a scraper and burnisher.


WM Intaglio scraper
Above: a scraping tool with three edges, for use on metal


WM Intaglio burnisher
Above: a burnishing tool, with smooth surfaces for rubbing


After about four hours of scraping and burnishing, I realised that there would be many more hours to get those areas low and smooth enough to not pick up ink. I decided to think upon this problem while moving onto the next step.


WM Smaller Cognitio scraping and burnishing
Above: shiny areas showing the process of scraping and burnishing


One of the very obvious (and to my mind, delightful) elements of metalcut is the decorative punch work. In the absence of any metal punches, I asked a friend to grind off the tips of some nails to various diameters, to create punches. To begin the process, I decided to mark a dotted outline around the edge of the area to be punched, and then proceeded with punch and hammer, using a block of wood underneath.


WM Smaller Cognitio marked for punching
Above: showing the guiding marks to begin punching


My aim was not to make a very deep mark, for as long as the punched area sits below the surface, it will not print. However, to get a reasonable impression, I had to use more force than I initially thought, and this force buckled the plate after a while. This would have been the advantage to having the copper plate nailed to a block of wood, because although I was punching on top of some wood to soften the impact, the plate was not attached, and was able to buckle. I continued, and occasionally turned the plate over and tapped it out as flat as possible. I used a smaller diameter punch for the background, and a slightly larger one for the wall.


WM Smaller Cognitio punching and tools WM Smaller Cognitio punched
Above left: punching in progress, and right: completed


However, I did not have anything fine enough for the draped clothing. On many extant metalcut prints, clothing is given a beautiful treatment of fine punching, with some achieving depth and flowing movement, while others achieving a strange flatness – very lovely either way. I was also worried that the many fine punches I would have to apply would buckle the plate too much.

I decided to compromise again, and using etching to achieve the fine, dotted punch pattern. It then occurred to me that if I had to apply another ground for etching, I may as well use etching to deepen the areas I could not deepen by scraping and burnishing. The plate was covered in ground in preparation for biting, and a scribe used to make many little dots all over the drapery. I tried to follow the folds of the clothes to create a sense of movement, rather than random coverage. The areas of the skin which needed to be further flattened were also exposed.

This time, the plate was bitten for around 15 minutes, again in 5 minute increments to ensure that the ground was not breaking down in the acid, or that some lines were biting too deeply and others, not enough.


WM Smaller Cognitio acid 2 WM Smaller Cognitio acid result 2
Above: the second time in the acid, and the bitten plate with ground still on the surface


After the etching process was finished, the plate was cleaned up and was ready for printing. As I was also doing some intaglio printing at the time, I decided to pull the second (and hopefully final) proof by using the intaglio press.

Part of me was curious to know if there would be much of a difference between relief and intaglio for printing a metalcut, and part of me was concerned that relief printing would not be successful, due to the slight unevenness of the plate as a result of the hammering. Intaglio plates can be slightly uneven and still be successfully printed due to the buffering of the blankets in an intaglio press and the intense point of contact through the roller. This is much harder to achieve in relief, due to the even, flat, downwards pressure of the platen.


WM Smaller Cognitio inked for printing    WM Smaller Cognitio revealed
Above left: the plate inked and ready to print and right: revealing the final state of the plate


Having previously used a brayer, and having forgotten to bring the inking ball I was experimenting with, I decided to use the brayer again. This time, because the plate was not perfectly flat, a narrow-width brayer was more suited to being able to ink up specific areas which would not be covered by a pass with a wider one due to slight unevenness of the plate.


WM Smaller Cognitio print and plate
Above: the final state and plate


The result is something I am very satisfied with, especially for a first attempt, and given that modifications of the process were necessary. There is the crisp, graphic quality apparent in extant metalcuts. While there is not as much intricate work, it still gives a sense of the strange, decorative, positive-negative tension of the originals.


WM Smaller Cognitio final print