Steinmetz Horn - Replica by Graham Nicholson



Replica Instrument by Graham Nicholson

Article by Russell Gilmour

In 1943, Gerhard Muchow, an esteemed professor of Art and Art History, was walking to work through war-stricken Berlin. Negotiating his way through the rubble and debris in the ruined streets, he noticed a man manoeuvring a wheelbarrow containing miscellaneous brass and copper items. Glancing down into the barrow, Professor Muchow noticed a castellated pattern on a piece of metal. The professor obviously had a keen eye for art and associated artefacts; he realised that this zig-zagged, tabbed seam must have had some kind of significance. Upon enquiring about this curious object, the barrow-man told him that it was a horn. He explained that there was another one in there too, but he couldn't let him take them because the barrow-load had been weighed. It had to go to the foundry at the end of the street to be made into shell casings for war munitions. You can imagine Professor Muchow’s horror! The quick thinking professor asked the barrow-man to wait while he rushed home to collect a Japanese tray. The copper tray weighed more than the two horns. Upon his return, the barrow-man agreed to the professor’s request to exchange the tray for the horns. A momentous day for us, though perhaps a traumatic day for collectors of Japanese trays.
  • Stacks Image 3401

    Meisterzeichen (Maker’s Mark) GFS

  • Stacks Image 3402

    Steinmez - Berlin

  • Stacks Image 3403


At the time of their fortuitous salvage, these were the only known horns made by Georg Friedrich Steinmez. He was born in 1668 and died around 1740. He worked in Nürnburg making trumpets, horns and trombones in undoubtedly the most important centre for brass instrument making. His maker's mark (Meisterzeichen) was an orb and cross with his initials: GF S. The other surviving instruments by Georg Friedrich Steinmez include: a trumpet in F (now in Copenhagen), a trumpet in D (in the Germanisches National Museum, Nürnburg), and an Alto Trombone in E-flat (No.3052) which is in the Berlin Musical Instrument Museum. Georg Friedrich Steinmez had a son who was also a brass instrument maker, presumably his apprentice. Cornelius Steinmez was born in 1702 and died in 1780. His instruments include: a trumpet (now in Brussels), a D trumpet (Braunschweig), and a trumpet in C (Salzburg).

Professor Muchow looked after the G F Steinmez horns until after the war and subsequently donated both to the Berlin Musical Instrument Museum in 1951 (Catalogue Numbers: 4187 & 4188). These horns were described by Willi Wörthmüller, in his doctoral thesis of 1955, as: ‘stable instruments in brass with silver-plated garlands, two and a half times wound in small format, tightly wound up until the bell which is conical. The tubes are cylindrical and made in one piece and the instrument starts with a decorated sleeve'. Wörthmüller goes on to describe a 'thick-walled, small bell' that is ‘well-worked’. 'It has a visible seam' [as we know]. 'The garland is engraved 'MACHT GEORG FRIDERICH STEINMEZ IN NVRNBVRG' [although his surname is now commonly rendered 'Steinmetz'. In this article Steinmez refers to the original instruments and Steinmetz refers to Graham's replica]. 'The instruments are in good condition and are playable. No date is indicated on either instrument and there are no original mouthpieces or crooks.' The description in the Berlin Musical Instrument Museum's 'Catalogue of Brass Instruments' (1976) additionally mentions the ornamental scallop shells on the garlands, the presence of lanyard rings and the fact that the inside of each bell has been [intentionally] blackened. No. 4187 has a creased bell and No. 4188 is in good condition. The instruments remain to be on display today in the Berlin Museum of Musical Instruments.

The story of Gerhard Muchow's rescue of these instruments came with good provenance - it had been told to Graham Nicholson by his friend Rainer Weber, who had studied under Professor Muchow at University. Weber later became one of the best-known musical instrument restorers in Germany. He died in February 2014 so it has become even more important to retell this remarkable story.

Doing just that, Graham relayed the story to me while he was working at the anvil in his workshop, pausing work on the Steinmetz replica for a moment, with a rawhide hammer resting in his hand. Graham had diligently measured the instruments in 1992, and he had developed his first prototype replica by 2000. Graham agreed to make an instrument for me on the understanding that I would come to The Hague and help him to make it, which I was more than willing to do. I made one visit in 2013 and two further visits in 2014. Quite how much help I was, I don't know, but I had an excellent time and it is so inspiring to be around Graham. He has such energy and enthusiasm for all things and he is exceptionally intelligent on a huge array of topics. Watching him make instruments is fascinating. It was an immense privilege to be in the workshop while he crafted a replica Steinmetz horn for me, and it was fantastic to hear the stories about the instruments as they were being made.

“So, that’s how the horns got into the museum in Berlin. Otherwise we wouldn’t have those horns and we wouldn’t have that type of horn at all” Graham surmised. It is true - there are no other horns that are short enough in length to be able to crook them into the highest keys that J. S. Bach required in his music. For example, BWV 65 'Sie werden aus Saba alle kommen' is written for two Corno da Caccia in C alto (but is often played down an octave). The unaltered horn part is within the usual range of a trumpet player, but it could be seen as extreme for a horn player. The 18th Century stadtpfeifer had to be proficient on several instruments; it seems that specialism on only one instrument is a relatively modern idea. So, it seems plausible that these parts would have been played by musicians who played the trumpet.

As well as offering the Georg Friedrich Steinmetz model, Graham has also copied a larger instrument from 1709 by Michael Leichamschneider and another horn by Johann Eichentopf. These makers were contemporaries of Bach. In 1740, at the age of 50, Leichamschneider moved from Vienna to Leipzig. He and Bach died ten years later. It seems likely that they would have, at least, met. Graham has copied these larger instruments for playing in lower keys. The smaller Steinmetz model is intended specifically for performing the higher baroque horn parts, in the keys of G and above.

The Steinmez horns allow us to approach this high repertoire on a historically appropriate instrument. In other words, the Steinmez horns - in their original form - are short enough to be high enough to allow us to crook them down to the keys required for playing the highest baroque horn parts. That is why they are unique. The small bell makes the instruments optimised resonators for playing high horn parts.

Graham listed, more or less on the fingers of one hand, the vaguely similar extant instruments, the other options for high horns - including a pair by Friedrich Ehe that are now in the Carolinium Museum in Salzburg. He explained that those instruments had been copied for use in the Harnoncourt and Leonhardt Bach recordings in the 1970s and he noted that it had been very difficult to play them in tune. The instruments have no surviving original crooks or mouthpieces and Graham attributed the tuning issues of the replicas to the design of the crooks that had been used. “It’s obvious that [with] the conical crooks - no matter how you build the cone - they just don’t work [on this kind of instrument].”

There are no surviving crooks with the Steinmetz horns either, so Graham has had to research and experiment with different ideas.

The earliest evidence we have that horn crooks existed at this time comes from an invoice for "a pair of great new 'Jägerhorn' with four new double crooks and four new tuning bits". This invoice was sent by the Viennese horn-maker Michael Leichamschneider in 1703. This is mentioned in Anthony Baines' book, 'Brass Instruments, their History and Development', (Faber & Faber, 1976), pg.156.

We assume that some instruments at this time were designed to be played in only one key; large-format horns in particular. It seems likely that other instruments, typically small-format (tightly wound) horns, may have been intended for use with a crook or a system of crooks. Using crooks seems especially plausible for horns bearing a trumpet mouthpiece sized receiver, like these by Steinmez.

With Graham's system of crooks, the instrument can be crooked down to C, B-flat, A, G and F, though F requires a wider diameter coupling crook which extends the G crook by a tone. It has to be said that the small-volume bell of the Steinmetz is optimal in the keys of G, A-flat, A, B-flat and C. The key of G is particularly good. Perhaps surprisingly the instrument still responds well when crooked in F using the additional (larger bore) coupler, though its use in F is probably most suited to small ensembles. It is also possible, though not ideal, to play it in low D.

Graham explained his rationale for the crooks to me: "I thought, first of all, let's make cylindrical crooks - so I made cylindrical crooks - and the result was as you saw. I then set about making a set of conical crooks, which is hard work because you have to make conical crooks for each length: so there were then C, B-flat, A and G conical crooks to be made... and they didn't work at all. So, empirically, I think what we have arrived at with [the crooks for] that horn is something which must approach, more or less, what they had... because it works."

An audacious claim, you might think, but Graham avoided making compromises in his replica of the instrument corpus. This meant that he could work on the crooks and the mouthpiece until he found a solution that would work with the instrument to enable us to play the repertoire of the period. He uses a copy of a mouthpiece from a horn in Salzburg, as it has a flat trumpet-like rim, though other mouthpieces could be made to suit the player and the instrument. Graham's result is perhaps just that: Graham's result. By his own admission it is just one interpretation, yet one that enables us to perform on this instrument with the minimum of compromise from the original. In other words, he made changes to what isn't there [the crooks] instead of changing what is [the corpus].

I asked Graham if there are any noticeable differences between the two Steinmez horns in Berlin. “Well, there is one slight variation.”, he explained, “One is 10.6[mm Ø] most of the time, and the other is 10.7[mm Ø] most of the time.”
“Perhaps that was a subtle optimisation for playing the first horn part and the second horn part?”, I enthusiastically offered...
“I think it’s just the kind of variation that you get when you make horns by hand”, he replied. This reminded me of a quote I had seen in a book I had been reading, entitled The Story of ‘A’: A History of Performing Pitch (Scarecrow Press, 2002) by the wonderful Bruce Haynes. I searched this book for the quote and, to my surprise, I discovered that Bruce Haynes was quoting Rainer Weber - the same man who had told Professor Muchow’s story to Graham! A remarkable coincidence.

“If we approach [early instruments] with the numerical mindset of the technologist, looking for absolute answers accurate to the last decimal, we will deceive or disillusion ourselves. We should consider if such a literal, mechanical approach belongs to the methods used in the past. While modern physics works with statistical “probabilities” and relative values, at the same time historical instruments are subjected to measurements to the hundredth of a mm and fractions of Hz or cents. We would do well to look less precisely.”

Unlike most later horns, the Steinmez instruments are only conical in the bell section; the first part of each instrument has cylindrical tubing. That is why cylindrical crooks make good sense. This cylindrical first yard might lead the modern organologist or taxonomist to classify these instruments as trumpets. Re: Weber, they "would do well to look less precisely". There are several good reasons why these are not trumpets. Firstly, trumpets do not have bells as large as this - 191mm Ø - nor would it be advantageous to. Secondly, this instrument was made at a time when the entity of the horn was in its infancy. There were few accepted conventions as to what a horn was. At this point, the horn was an instrument without concrete rules.

"Because the Cor de Chasse was still new during the Baroque... every horn built should be viewed as a "prototype", thereby offering us an understanding of the variance from maker to maker; there was no norm for builders to follow." 8th April 2013 - Lowell Greer

This is especially true of the early part of the 18th century, where we see substantial variations in what a horn can be. The horn was a new instrument and it seems logical that it developed from the established manufacturing processes of other brass instruments. Trumpet makers and players would have had the most relevant skills to develop and play the horn in the beginning. Graham likes to joke that horn players in the 18th Century had not been in the labour exchanges of Europe, waiting for their instrument to be invented!

The Steinmez horns (both in the 18th Century and now) blur the margin between the horn and the trumpet, offering us a different approach to the high baroque horn. Typically, the baroque horn has attracted players who already play the classical horn. These players may have a perception of the baroque horn that is influenced by what came later. Most trumpet players do not have such experience of playing later horns, and so their approach to the baroque horn will inevitably be a different one. Modern experimentation by trumpet players playing the baroque horn could lead to similar results to those players who first played or developed the early baroque horns in the 18th Century. In other words, if the situation is similar, the results may be similar. This is likely to be an esoteric comment, but this scenario may yield an interesting secondary dimension of Historically Informed Performance practice.

Once the basic principles of the instrument had become well-established, horn playing would become an altogether separate discipline to trumpet playing - attracting specialist players and a new pedagogy - and so the two professions went their separate ways. Horn playing went on to reach a zenith with the virtuosic music of the classical period and hand-stopping future-proved the instrument's popularity, and so horns with small bells became less common.

Interdisciplinary experimentation and musical collaboration may help horn players and trumpet players to have greater understanding of the early years of the baroque horn. It's exciting to wonder where this experiment may lead. It is an interesting experiment for me. I personally find that practicing the horn improves my understanding of the harmonic series and how it can be adjusted - and this is hugely beneficial to my trumpet playing. Also, the longer tube of the horn requires a greater volume of air to be moved - which is excellent practice for any brass player. I really enjoy playing the instrument and I get a great deal of pleasure from looking down at the personalised engraving and the generally pleasing shape of the object. Even after only a short time of playing the horn, I have noticed a significant improvement in my trumpet playing.

In 1986, another Steinmez horn came to light. According to Graham, it was handed in to the Shrine to Music Museum (Catalogue Number: NMM 4013) in South Dakota by an ex-tank commander, who had apparently taken it from a castle in Germany during the Second World War. It is very similar to the Berlin Steinmez instruments apart from the fact that it has a different style of ferrule. This is possibly a replacement, but it is still a Nürnburg-style ferrule.

"Current research suggests that it may be one of a group of transitional coiled instruments that led to the development of the horn, while still retaining some features of its trumpet ancestry." - National Music Museum Website

It's incredible to think that all three of the existing Steinmez horns have survived by virtue of what we might call 'opportunistic retrieval'. These horns illustrate an exceptionally interesting chapter in the history of the horn.

Replica Construction:

The first stage of the construction involved cutting parallel strips from a long sheet of brass, using a guided cutting wheel. This long and narrow rectangle of brass sheet was then pressed around a steel mandrel, just pinched around by hand, to make it approximately tube shaped. Graham used a handheld steel roller to apply pressure along the tube on the mandrel, making it a more uniform shape. The mandrel supports the tube on the inside to prevent the tube from becoming crushed when force is applied to the outside. This process brought each side of the sheet neatly together, around the cylindrical mandrel. Graham over-tightened the tube slightly on a smaller diameter mandrel so that the work slightly overlapped, meaning that the tube would have enough sprung tension to hold itself together during the next stage: brazing.

Using a gas torch, Graham applied heat and began to braze along the seam. I had to ask about this fascinating process. How was he sealing the tube without adding silver, lead or another kind of solder? He explained this by dunking his hand into a quenching trough full of water. Withdrawing it, he held his thumb and forefinger up to the light and moved them ever so slightly apart. There was a small bubble of water trapped between his fingers - held there by surface tension. He added a paste of brass powder (schlaglut) along the seam and used the gas torch in his left hand and a piece of steel in his right. He used the steel to drag the molten brass droplets together and he allowed gravity to direct the molten brass into the seam. The droplets are then held by surface tension in the overlap.

Graham is so skilled at brazing; he can braze tubes in the same way that you or I can butter a piece of toast. He spreads the molten brass along the seam, occasionally stopping to inspect it and adds more schlaglut if necessary. He uses borax (and gravity) to control and direct the flow of the molten metal and the borax helps to keep the rest of the tube relatively clean during the whole process. It takes incredible skill - dancing the flame across the current, next and previous sections of the work in order to control the temperature - Graham would warm the next section to be worked, while taking the heat off the current section being brazed - this prevents blasting the molten metal into the wrong place with the force of the modern gas torch, and makes for a more continuous workflow.

This amalgamation between two only slightly different alloys of brass can only happen within a narrow range of temperatures. After annealing and quenching it, the seam of the tube was flattened down and worked on a mandrel. Finally, it was swaged and drawn down to the required diameter using incremental draw bars.

The bell was made in much the same way, but instead of starting with a parallel strip of brass, he used a special template to ensure that there was no overlap once the bell had been formed around the mandrel. He soldered the bell using silver solder as we know that the bell sections on the originals have visible, tabbed seams.

Of course, making this instrument now takes relatively little time compared to how long it must have taken Graham to make the customised tools. Graham made the mandrel himself, after his own careful measurements of the originals in 1992. This must have been a slightly trickier process than copying a trumpet mandrel as he had to work out how the horn mandrel would have looked when it was straight, and reproduce it. It's certainly a labour-intensive process but Graham appreciates that it has always been this way: "The biggest expense [for an 18th Century maker] would have been tooling up to make a mandrel". I asked Graham what had been the most challenging aspect in copying this instrument in particular. "I found the Leichamschneider much more difficult, technically, to make... because the bigger it is, the more different it is from a trumpet, which I was used to making".

While Graham was working on the bell, I had been given the task of decorating the garland. Using a compass, we etched and then cut a disc from a piece of sheet brass. I traced Graham's plan drawing of the garland onto a piece of tracing paper. We then painted the flat brass disc with white acrylic paint and then I copied the tracing onto the painted brass surface using carbon copying paper and an old ball-point pen. Using the sharp nib of a dried-out fountain pen, I then scratched off the thin layer of paint over the guidelines, where the garland should be engraved. This provided a tiny groove, which it was hoped would help to guide the engraving tools. Mounting the disc on a circular wooden block, I began to practice chasing (a form of engraving) on the waste segment. Chasing involved holding a graver in my left hand with the point facing down, with the handle by my thumb. I held a repousse hammer in my right hand and tapped the graver firstly beneath the surface of the metal (and then at a lower angle) along the pre-etched lines. At the end of each cut, I had to push the tool down to lift the cutting edge neatly out of the groove. I didn't start initially on what would become the garland, instead practicing on the waste segment.

After about half an hour of fairly fruitless tapping, I finally felt a 'eureka' moment, when I found the optimum pressure and angle for a smooth cut. The graver guided itself effortlessly where I had intended it to go. Before I could even express my delight at this, Graham enthusiastically shouted 'yeah, that's it!' from the opposite side of the workshop. Without even looking in my direction he could tell - presumably by the sound it made - that I was ready to start engraving on the actual garland.

Graham was at hand to help me with some of the trickier aspects of the design. Engraving the leaves and flowers was not too difficult, as they were organic things it didn't matter if they were slightly off the mark. I needed his help particularly on the parallel lines and with the lettering - which most engravers agree is the hardest thing to master.

I personalised the garland by adding the triskelion symbol of the Isle of Man - 'the Three Legs of Man'. When I finally completed the garland engraving, after around 7 hours of work, Graham cut out the waste segment and shaped the garland into a basic cone. After some calculated hammering, he gently clamped it in place and soldered it, this time using silver solder.

The next task was to carefully shape the garland using the bell mandrel as a guide, so that it would fit tightly and precisely around the bell. This is quite a tricky process as the garland is quite delicate and a lot of time has already been invested into it. The bell fits the mandrel tightly, but the garland has to be slightly and uniformly wider than the mandrel in order to account for the wall-thickness of the bell, so that the garland fits on top of it.

Before fitting the garland to the bell, the scallop shells had to be stamped onto the small end of the garland. Graham had made a special tool for this. The garland was stamped, on the outside, with one swift hammer blow. The work was padded with leather and supported by a rhino tusk-shaped anvil, which showed evidence of repair after it suffered intense temperatures from fire-bombings in Köln during World War II.

The stamped scallop shells are pushed through from behind using a tool with a ball bearing on the end. This shaping through the back of the garland makes the shells stand out and look three-dimensional. Graham has made a hemispherical gauge in a piece of wood which, along with some gentle persuasion from a hammer makes the stamp look like a scallop shell. The top of the scallop shells were then artistically shaped using the small cutting disc of a handheld power tool.

After fitting the garland, the wire around the circumference of the bell had to be pressed with a pattern, then soldered and hammered so that it would fit neatly on the edge of the bell, on top of the garland.

By this point we had almost everything we needed to make a playable instrument. We had a first yard and a bell section, now fitted with the garland. As yet we had no crooks or a mouthpiece, but we borrowed those from another of Graham's horns and tested the instrument in a straight format.

On my next visit to The Hague, the main mission was to bend the horn. This involved filling the whole bell full of lead. This is a fairly dangerous procedure, not only because lead is poisonous but for several other reasons. Graham took time to ventilate the workshop, and he went through the rituals of preparing the lead and removing any slag or imperfections. Next he cleaned the inside of the bell and ensured that it was completely dry. He put a bung in the narrow end of the bell tube and wired the whole bell to a vertical metal support, with the flare at the top. In a way, the bell would act as a funnel. He carefully removed any bubbles from the lead by gently tapping the bell. He allowed the metal to cool and solidify before bending the bell around the fixture.

Having successfully bent the bell, Graham moved on to make the decorated sleeves. He did this by placing a piece of brass on top of an engraved steel plate then he passed this through a machine that resembled a mangle. The engraved pattern on the steel transferred onto the softer brass. This piece was then carefully shaped and soldered to create the decorative sleeves, also known as ferrules.

The next processes involved putting all of the completed components together and soldering them. The parallel coils of the horn were selectively soldered to each other. The instrument was polished and cleaned and we decided to bind the bottom part of the instrument in leather to protect it.

The next visit to The Hague involved making, tuning and fitting the crooks to the instrument. They beautifully fit one inside the other and have the same style of decorated ferrule.


I am so indebted to to Graham Nicholson for his generosity in freely sharing his knowledge with me, and also for his excellent hospitality in The Hague. Thanks also to Daniel Serafini for taking the photograph of me with the horn.
Copyright © Russell Gilmour - All Rights Reserved.

Steinmez Original

  •  Bell


  •  Meisterzeichen


  •  Garland Detail

    Garland Detail

  •  Bell


  •  Garland


  •  Corpus


  •  Garland


  •  Ferrule


Steinmetz Replica

  •  Corpus on Bell Anvil

    Corpus on Bell Anvil

  •  Garland


  •  Graham Nicholson

    Graham Nicholson

  •  Hammer & Drawbars

    Hammer & Drawbars

  •  Rolling Mill - Ferrule

    Rolling Mill - Ferrule

  •  Ferrule


  •  Brazing Tubes

    Brazing Tubes

  •  Engraving Garland

    Engraving Garland

  •  Engraving


  •  Garland


  •  Engraving


  •  Bending


  •  Polishing


  •  Garland Detail

    Garland Detail

  •  Corpus


  •  Montage


For further information, please contact me.

© Russell Gilmour 2014