«The Trombone of Anton Schnitzer the Elder in Verona: A Survey of Its Properties and Their Acoustical Significance Hannes W. Vereecke The growing ...»
The input impedance curve in Figure 6 indicates that the main resonance of the mouthpiece is strongly altered by the backbore and thus alters the playing behavior significantly. To demonstrate more clearly the wide divergence between the belly-shapedand conical-backbore mouthpieces, we also determined the resonances of a Dennis Wick 9 BS mouthpiece. Surprisingly, the main resonance of the Wick mouthpiece is closer to the original belly-shaped-backbore mouthpiece than is the main resonance of the modified copy with modern-style (conical) backbore. If one uses both mouthpieces on the same instrument, the varying mouthpiece resonances will result in a different impedance behavior of the overall instrument. This will consequently produce differences in the intonation and playability of the instrument and ultimately the sound as well. Thus the use of a historical mouthpiece with belly-shaped backbore is a more logical choice if one aims to recreate the playing behavior and sound of the original.
Figure 6: Logarithmic depiction of the resonances of a copy of the Verona mouthpiece with belly-shaped backbore and one with modern backbore—i.e., one that is conical with largest diameter at the distal end. In order to demonstrate the wide divergence between them, the impedance curves are also compared with that of a Dennis Wick 9BS mouthpiece.
32 HISTORIC BRASS SOCIETY JOURNAL
The garland is engraved “MACHT. ANTONI. SCHNICZER. ZV. NVRNBERG [crown] M.D.LXXVIIII.” Spinning marks are visible, implying that the bell was turned on a mandrel attached to a lathe, as is still done by modern makers. However, the marks could also be the result of a later restoration. At some point the bell was broken into two parts and subsequently restored with an additional conical tube (see Figure 3). Underneath this tube a point of fracture is still clearly visible. Presumably this irregularity in the bore would cause distortion in the playing behavior and most likely would also degrade the playing quality. The bell is 526 mm long and at approximately the halfway point a gilded ball weighing 56 grams is attached. There are still many open questions concerning the acoustical function of this ball. Observation in a numerical, finite-element model reveals that it alters the vibrational behavior of the instrument considerably. Analysis of these influences on sound and playability is beyond the scope of this paper and requires further research.
The scaling of the bell has the opposite acoustical function from the mouthpiece in that it has the effect of raising the low-pitched resonances. Furthermore, its geometry defines which waves will be radiated and which will be reflected back into the instrument to Figure 7: Comparison of the bell design of three trombones made by Anton Schnitzer the Elder/Younger.
VEREECKE 33 constitute a standing wave. The pressure of a sound wave travelling through an instrument falls as the cross-section increases, so there is a direct relationship between the dimensions of the bore and the acoustic characteristics.21 This relationship can be represented by the bell’s horn function,22 a value that defines which waves will be reflected and which will pass through. The most important effect of the horn function is the so-called “cutoff frequency.”23 Like the brassiness potential parameter, the cutoff frequency is an important parameter that governs the timbre of the instrument to a large extent.24 Arnold Myers investigated several bell profiles and their associated horn function, including a trombone made by Anton Schnitzer the Younger in 1594, currently in the Edinburgh collection.
Comparing the horn function of modern-shaped bells with greater terminal flare with that of the 1594 instrument, Myers concluded that the potential barrier reflects the high-pitch components less effectively, “thus giving the sackbut a mellower sound.”25 Figure 7 indicates that the bore shape of 13.301 is similar to two other similarly scaled instruments made by the Schnitzer family, preserved in Edinburgh and in Nice.26 It is possible that the differences shown are due to inconsistencies in measurement, since the author himself measured only the Verona instrument.
The slide section of 13.301 consists of three slides—two outer and one inner, the latter consisting of a descending slide tube and an ascending tube. The external slide fits well and clamps onto the second outer slide. The external slide thus consists of a doublewalled construction, illustrated in detail in Figure 9. The acoustically effective bore at the entrance measures 10 mm in diameter, which is similar to the instrument made by Anton the Elder preserved in Palais Lascaris in Nice. The smaller the diameter of the bore, the more pronounced the instrument’s resonances will be. As well, the speed of propagation will decrease and thus the wave will need more time to reach the end, seemingly making the instrument longer and thereby lowering the pitch. In general, it can be said that an instrument with a small bore will allow one to play easily in the higher ranges because the resonances are more pronounced and the playing accuracy will generally be better than in wide-bore instruments. However, instruments with small bores are difficult to play in the low register because of the high frictional resistance. Thus the relatively small bore of
13.301 indicates that this instrument should be easier to play in the higher registers.
Crooks and tuning bits
Three straight tuning bits and three detachable crooks are preserved with the instrument.
The tuning bits have respective lengths of 80.4, 104.3, and 134.3 mm (see Figure 3a).
The crooks intrude 20 mm into their ferrules and weigh approximately 52 grams. Their cumulative length is approximately 193 mm, which gives them an effective acoustical length of 173 mm. There is much confusion as to how these crooks and tuning bits relate to each other and to the instrument, and also as to their acoustical significance. The
34 HISTORIC BRASS SOCIETY JOURNALferrules are all provided with marks consisting of a combination of the symbols X, V, //, and /. Diagonal marks, either single or double (/ or //) are engraved on various parts of the slide. All parts of the descending outer slide tube are marked with // and all parts of the ascending slide tube, with /. These marks leave no doubt as to the proper assembly of the various slide parts and the bell-bow, but unfortunately they do not yield sufficient information on the arrangement of the tuning crooks and the various other parts belonging to the bell. In addition to the marks on the slide and the ferrules, each crook has a V or IIII engraved on the side.
Interestingly, the painting by Ludovico Carracci discussed by Markus Raquet and Klaus Marius may yield additional insights into the meaning of these crooks.27 The instrument depicted on this painting shows several morphological similarities to 13.301, and it is depicted with two tuning crooks between the slide and bell part sections. Marin Mersenne’s illustration in Harmonie universelle (1636) offers another iconographical source that bears on the discussion here,28 for it shows an instrument with four crooks inserted between the bell and slide sections. It raises the question: Is 13.301 missing one crook, or is the third crook an extra one? One way or another, an even number of crooks—either two or four—is needed to establish the proper connection. In order to determine the significance of these crooks, one could construct an extra crook and measure the input impedance differences. Another approach, however, is to construct a computational model based on the scaling. This technique has been described previously in this Journal.29 An unknown input impedance can be calculated from a known scaling because there is an unequivocal relationship between the two values. One crook has an effective acoustical length of 173mm, two crooks will add 346 mm to the acoustical length; four would add an additional 692 mm. Table 1 indicates that the instrument is lowered one tone using two crooks, and two tones using four crooks.
diametrically opposed on this issue.30 Apart from its effect on weight, the thickness of the wall greatly influences the vibrational behavior of the instrument. Richard Smith states that vibration amplitude is inversely proportional to the fourth power of wall thickness.31 How—and especially, how much—these vibrations effect the radiated sound is still an open question. Wilfried Kausel and Thomas Moore demonstrated the effect of wall vibrations on sound under extreme conditions.32 They concluded that the wall vibrations of brasswind instruments do affect the radiated sound.33 The significance of these structural vibrations for both player and audience in a performance situation is still under debate.
Figure 8:Wall thickness measuring points. Photo by Maurizio Brenzoni.
These results indicate that raw material of approximately 0.35 mm has been used. The thickness varies due to the technique of manual forming on the mandrel, and reaches its minimum at the end of the bell flare. The crooks and bell bow were made of 0.75 mm. material. Thicker material enables a builder to bend the crooks more easily and still achieve a circular cross-section. The straight tube between the tuning coil and the slide (Figure 8, point 10) has a wall thickness of 0.5 mm. Some makers contend that a thicker wall at this point affects the playing behavior. Although these assertions have not been proved, they raise the question as to whether this was intentional on Schnitzer’s part or merely coincidence.
36 HISTORIC BRASS SOCIETY JOURNALFigure 9: Wall thickness measuring points on the three slides. Photo by Michele Magnabosco.
11 0.35 0. 45 12 0.47 0. 5 13 0. 38 0. 39 14 0. 30 0. 29 15 0. 49 0. 50 16 0. 40 0. 39 17 0. 30 0. 30 Table 3: Wall-thickness measuring points on the slides, measured on axes of 0° and 90°.
The vibrational behavior of brasswind instruments is a function of wall thickness, processing techniques, and the raw material used. The proportions of the various components impact certain physical properties, such as the elasticity module. It will also influence the recrystallization process, which is a key factor in the manufacturing process. XRF analysis revealed that all parts, without exception, are made of an alloy of approximately 20% zinc and between 1 and 1.5% lead. Only the cup of the mouthpiece contains a higher lead component of about 3%. Today, CuZn37 (composed of 63% copper and 37% zinc) and CuZn28 (72% copper and 28% zinc) are the typical alloys used for reconstructions of early brasses. The difference between an alloy of 37% zinc content and one of 20% VEREECKE 37 (as found in 13.301) is considerable, and one assumes that this difference will affect the structural vibrations. However, this requires further research. The XRF spectrum of the gilded ferrules indicates the presence of two elements in addition to copper, zinc, lead, and gold: silver and mercury. Silver was first applied to the brass ferrule, as is still done today by some modern makers. The presence of the element mercury favors the hypothesis that the ferrules were fire-gilded. In this case, mercury was needed to form a gold amalgam, which was applied to the object. Heating causes the mercury to volatize, leaving behind a thin layer of gold.
A benchmark for a good instrument design is one in which all fundamental frequencies of the pitch centers of playable notes lie close to a harmonic series. A straight tube has impedances which are found at uniform distances from each other. The resonances of a straight tube do not lie in a harmonic series, therefore a straight tube is not useable from a musical point of view. Interaction between bell, mouthpiece, and body design are required in order to bring these resonances together in a harmonic series and make a musically useful sounding body. A rule of thumb used by makers in judging a good instrument is that the deviation within the harmonic series should not be more than 15 cents. Analyzing the intonation of trombones is a particular challenge, since the closed position is the only one that can be determined precisely. Despite this difficulty, closed position still gives us a good insight into the acoustical qualities of the instrument, based on its geometrical design. Using the analysis software BIAS, it is possible to calculate
the intonation, using equal temperament as a reference. The instrument was measured without its crooks (see Figure 10).
The intonation analysis indicates that bf, d1, and bf1 are at the farthest limits accepted for use today, using equal temperament as a reference.
For research purposes, the author built a prototype of the instrument in cooperation with Rainer Egger, based on the findings mentioned above. The analysis of this instrument and extensive playing tests confirmed our hypotheses about it. All tubes were soldered, using an argentiferious soldering alloy. In the construction of the bell, the soldering seam was hammered by hand. As in the original, the parts were clamped rather than soldered, except for the slide-bow, which was soldered with low-melt solder. For the acoustical planning and development of the reconstruction, the Brass Instrument Optimization Software (BIOS©) was used.34 Both the original and the copy were measured in the same room at the same time with the same measuring gear (see Figure 11).
Figure 11: Comparison of the impedance curves of original instrument and copy.
VEREECKE 39 As visualized in Figure 11, the acoustical footprint of the copy is very similar to that of the original. The curves match well. The differences in amplitude could be attributable to leakage in the original, to the energy losses due to thinner walls at certain points, or to general frictional loss. Furthermore, the difference could also be attributable to the fact that modern alloys have been used for the copy. In the course of further research we intend to construct another copy using raw materials closer to the original specifications.