GUITAR STRINGS FROM THE NINETEENTH CENTURY TO THE ADVENT OF NYLON

                                                                                                  by Mimmo Peruffo

By Mimmo Peruffo (from 'La Chitarra di Liuteria', by Stefano Grondona e Luca Waldner, Ed. L'officina del Libro, Sondrio, November 2001, pp.168-176; see www.guitarclassic.com) 'Las cualidades sonoras del mejor instrumento desmerecen si está provisto de cuerdas mediocres.' (Emilio Pujol, Escuela razonada de la guitarra, Buenos Aires, 1934).

 GUT

A material that has been in use for centuries (strings made of gut have been found in ancient Egyptian plucked string instruments dating from the third dynasty), gut has always been the principal source of strings for musical instruments in the West. Although the process necessary for the production of a gut string had been defined some centuries earlier (in Catalunia, for example, there were detailed written regulations governing the production of vihuela strings as early as the middle of the sixteenth century), it was not until the second half of the seventeenth century that overspun bass strings were invented, consisting of a gut-core (nowadays of nylon multifilament) completely covered by a fine metal wire. Although the earliest manuscript reference to such strings dates from 1659 (E. Hartlib:'Ephemerides'), the diffusion of these efficient bass strings took longer than might be expected: the viola da gamba player Sainte-Colombe, for example, introduced them to France only around 1675. The discovery was of considerable importance, both in terms of construction and musically, such that it is certainly possible to speak of a real dividing line between what came before and what came afterwards. It seems reasonable to suppose that as soon as musicians had much more brilliant bass strings at their disposal, the first thing that came into their minds was to reduce the vibrating length of cumbersome bass instruments, rendering them much more agile. This opened the way to new musical forms, and was also the real driving force behind the addition of a low sixth string to the guitar towards the end of the eighteenth century, with a simultaneous reduction in its vibrating length by comparison with that of a typical five-course instrument (i.e. 68-73 cm). This led directly to the gradual abandoning of courses in favour of simple strings. Thus, far from being mere accessories of the guitar (as they are often regarded today), strings actually conditioned its evolution - to a not inconsiderable degree.

A BRIEF EXPLANATION OF THE PHYSICS OF STRINGS

When a string made from any material is progressively stretched between two fixed points (which determine its vibrating length), at a certain moment it reaches a frequency at which it breaks. This point corresponds to the breaking load of the string, which in the case of gut is about 32 kg/mm². The value of this limit frequency, known as the 'breaking frequency', is completely independent - strange as it may seem - of diameter, as may be easily demonstrated either mathematically (applying the general formula for the strings) or experimentally. This limit frequency is in direct proportion to the vibrating length of the string. In other words, the product of the vibrating length - in metres - and the breaking frequency - in Hz - is a constant defined as the 'breaking index'. The average breaking index of a modern gut string in experimental conditions is 240 Hz/m, obviously corresponding to a breaking load of 32 kg/mm². This means that at a vibrating length of a metre the string will break, theoretically, at 240 Hz. If one divides the breaking index by the tuning frequency chosen for the first string, this will produce the vibrating length at which the string will break. For the first string of a baroque guitar in E at the supposed seventeenth-century tuning standard of A = 415 Hz (according to which E = 315 Hz), the theoretical length at which the first string will break is 75 cm; the choice of a 'working' vibrating length will therefore have to consider a prudential shortening of this limit length. But by how much? To answer this question we must return briefly to the period preceding the advent of overspun bass strings. Musicians had always known that a string works best when, subjected to what seems to be the right degree of tension - that is to say, neither excessively taut nor excessively slack to the touch - it has the smallest possible diameter. Once this degree of tension was established, it then had to be distributed evenly across all the strings of the instrument. It was known, moreover, that as the section of a string increased - its tension and vibrating length being equal - it reached progressively lower frequencies, but that at the same time its total acoustic output (in terms of dynamics, the richness of overtones and the duration of the sound) diminished, to the point of becoming - above certain diameters - practically unacceptable. The only solution possible at that time - they were basically limited to gut - was to increase the vibrating length up to the physical limit determined by the first string, as seen above. Only in this way could one hope to reduce the diameter of all the strings as much as possible (particularly the bass strings, which were the thickest and therefore the most critical), thereby drawing from them their best sonority. Vibrating length and diameter are in fact inversely proportional. On the basis of the vibrating length in surviving plucked string instruments and the mechanical properties of gut strings, researchers have speculated that the working length probably entailed a prudential shortening of the hypothetical 'breaking' vibrating length by about 2-4 semitones. Thus, the above-mentioned vibrating-length limit for the guitar - 75 cm - corresponds to a 'working' length of about 69 cm, which is in fact very typical of surviving five-course instruments. With the appearance of overspun strings, the rule of increasing the vibrating length as much as possible no longer applied, in that the acoustic exuberance of the new bass strings was such as to recommend to eighteenth-century instrument makers a salutary shortening of the vibrating length (about two frets less) so as to increase agility of performance. It goes without saying that because the vibrating length was shortened and the tensions remained the same, the diameters of the first three strings, made purely of gut, had, by the laws of physics, to be increased, inevitably resulting in a certain loss of brilliance and a 'sweeter' sound: more viola than violin, as it were.

THE PRODUCTION PROCESS

At first sight, the production process followed in the nineteenth century and beyond seems surprisingly similar to that of today. There are, however, a number of important differences, which suggest that the strings of that time - above all those produced in Italy (Rome, Padua, Salle and especially the famous Neapolitan strings, which were favoured by Paganini and by late nineteenth-century London) and in France (Paris and Lyon) - were superior to ours in a number of respects. The process entailed the use of lamb gut, which, having been carefully emptied and rinsed in running water for several days, was treated in various ways so as to eliminate all not-muscular membranes and fatty substances. This was done by leaving the guts to soak for several days in alkaline solutions (prepared by dissolving ashes in water), the strength of which was gradually increased to the point where the unwanted membranes and the fat that always accompany catgut could be easily removed by scraping gently with a piece of ditch reed. Several cleaned guts were then placed together (the number determining the diameter of the final string) and twisted repeatedly using a suitable winder, after fixing one end of the proto-string to a peg on one side of the drying frame. When the string had been properly twisted, its free end was fixed to the other side of the frame, putting it in traction. When the frame was full of twisted guts, it was placed inside an airtight chamber, where the guts were bleached with sulphur dioxide prepared by burning sulphur in a bowl. At the end of this process the strings were again twisted and left to dry in the air. The final stage consisted of the sanding of their surfaces by rubbing with a herb with abrasive properties, or with pumice powder. The perfectly sanded strings were then oiled with olive oil, cut off the frame and wound into circles. The tendency today is rather to twist the guts less, thereby producing strings that are too stiff, and to rectify this mechanically: although this guarantees a specific string calibre, the fibres of the string often suffer from over-correction, with the risk of reduced durability.

CRITERIA FOR ASSESSING THE QUALITY OF STRINGS

What were the criteria by which a good string was distinguished from a bad one in the nineteenth century? The first point to emphasise is that musicians of the time - including guitarists - seem to have been able to distinguish good-quality material simply by touch and by sight; Aguado in his Method wrote that 'the guitarist must be master of the strings'. The provenance of strings was in itself regarded as a good indication of quality; any musicians was able to tell an inferior string from a good one: this knowledge had probably been passed down through the centuries from master to pupil within an oral tradition that perhaps began to disintegrate around the beginning of the twentieth century, from which point it became increasingly common to rely blindly on the big string-producing firms that began to establish themselves, particularly in Germany and France, at the end of the nineteenth and the beginning of the twentieth century. The age-old custom of oral transmission may well explain why so little was written about the criteria for choosing strings in the guitar manuals of the time, and why the little we know derives principally from manuals for violin - the instrument around which everything rotated - or from manuals relating to the construction of bowed string instruments in general. To summarise what is written in the bowed string instrument manuals of the time, a good string should be transparent, of a yellowish or gold colour, smooth, well twisted and elastic; in other words, not stiff to the touch.

STRING TYPES AND DIAMETERS

Before considering diameters and working tensions, we should give a moment's thought to the following question: why is it that so few manuals of the nineteenth and early twentieth centuries give dimensions for guitar strings, as they do for strings for the violin and other bowed string instruments (with the single exception of Pujol, which occurs well into the twentieth century)? The mystery is solved with the help of a number of documents of the time, in which we read that the first strings of the nineteenth-century guitar were identical to the first three strings of the contemporary violin, an instrument about which we know a great deal in relation to strings. The guitarists of the time probably regarded this as common knowledge, and therefore felt that it was unnecessary to discuss it in their manuals. To answer our questions fully, at least with regard to the first three gut strings, we must turn our attention to the violin, taking into consideration not only the manuals but also the information that has come down to us from the string makers of the time, principally in relation to the number of guts required for each string. It should be pointed out, however, that this number in fact determines not a precise final diameter but rather a fluctuation around an average diameter, in that guts, being a natural material, will always differ slightly in thickness (at the time, moreover, there was no mechanical means of correcting strings, the only way of ensuring a precise calibre). It is known that the first string of the violin was made from three lamb guts, which produced a diameter of between 0.65 and 0.73 mm. For the second and third strings five and nine guts respectively were used, producing a diameter range of 0.80-0.90 mm for the A string and of 1.04-1.20 for the D string. These were also the E, B and G strings of the guitar of the time of Sor, Giuliani and Coste. There were two main types of guitar string available from the beginning of the nineteenth century to the middle of the twentieth: oiled natural gut for the upper three strings and overspun silk for the three bass strings. It should be emphasised that the acoustic output of strings made of overspun silk is generally greater than that of the strings with a gut core that were used in the eighteenth century for five-course guitars and for bowed string instruments in general. After the addition of the sixth string and the elimination of courses in favour of single strings, and up to the 'enlarged' guitars of Torres (the second half of the nineteenth century), the vibrating length was stabilised at about 62-63 cm, as shown by the manual of Aguado: 27 pulgadas, i.e. around 62 cm (.9132 of an inch, see J.H. Alexandre 'Universal Dictionary of Weights and Measures Ancient and Modern', New York 1867 p. 90) and as seen in the many surviving instruments of the period, whether made in Italy or abroad.

  STANDARD PITCHES

An important element in determining the working tensions of the guitar of that time relates to the frequency of the standard A that was in use in the nineteenth century, which varied considerably, and not only from place to place, but also in the same place from one period to another. In 1834 the Congress of Stuttgart approved a tuning standard of A = 440 Hz, but this recommendation was not followed. In 1858 the French government reported that the tuning standard of the Paris Opéra and the Opéra Italienne was A = 448 Hz, but a year later a French commission for the standardisation of tuning (composed of illustrious figures such as Halévy, Auber, Berlioz, Meyerbeer, Rossini and Thomas) - the first in Europe - established A as 435 Hz through an imperial decree. In England, orchestral pitch was A = 424 Hz in 1813, but this was raised to 452 Hz in 1859. The supposed nineteenth-century tuning standard of A = 435 Hz seems to have been an illusion rather than reality, and this is certainly true up to the second half of the nineteenth century. With the Congress of Vienna of 1885 the standard A was officially established at 870 simple vibrations, or 435 double Hz, a recommendation that was also adopted by the Italian government in 1887, but in fact the tuning standard continued to fluctuate. Only with the meeting called in 1939 by the International Organisation for Standardisation was the situation presented by the jungle of different tuning standards clarified, proposing a standard A of 440 Hz. The rest is recent history. On the basis of diameter range, vibrating length and tuning standard (for the sake convenience, A = 435 Hz), the range of working tensions of the first three strings of the nineteenth-century guitar may be calculated as follows:

1st,   E (325.9 Hz):     7.4-9.0 kg (average 8.2 kg)

2nd,   B (244.0 Hz):     6.9-7.9 kg (average 7.4 kg)

3rd,   G (193.8 Hz):     7.4-9.3 kg (average 8.2 kg)


Extending these figures to all six strings, one arrives at an overall tension of 46.8 kg, which corresponds to the figures given by Aguado, who specifies an overall load of 80 or 90 pounds, i.e. 39-44 kg (1 pounds =489,5 grms; see Horace Doursher 'Dictionnaire universel des poids et measures…', Antwerp 1840, facsimile ed. Amsterdam 1965). As may be observed, these approximate working tensions are certainly higher than we would expect, and if anything rather similar to those we use ourselves.

 OVERSPUN BASS STRINGS

Since overspun strings are made from joining together two different kinds of material, such as metal and silk, it has become customary to describe them in terms of equivalent gut strings. In other words, we refer in calculations to the diameter of a theoretical equivalent solid gut string of the same weight as the overspun string per unit of length: at the same tuning and vibrating length it will therefore have the same working tension. It should be noted, however, that for any given equivalent solid gut string, the ratio between the metal and the silk may be endlessly varied. An increase in one material will obviously entail a reduction in the other, if the total weight of the string is to remain constant (that is, its equivalent gut string, and therefore the working tension of the tuned string). It goes without saying that the greater the prevalence of silk in relation to metal, the less brilliant and more opaque the sonority is likely to be. What criteria were used to determine the right ratio between metal and silk in bass strings, one that would guarantee a balanced sound in terms of timbre and dynamics? In the guitar the ratio was more limited than in bowed string instruments: once the working tension had been decided on, the proportions between metal and silk were calculated so as to produce the greatest volume of sound, using the thickest possible metal wire and at the same time reducing the silk core to a minimum, almost to the breaking load of the string when in tension on the instrument. In spite of this measure, overspun silk strings - even those that have remained in their packets - sound rather percussive to our ears, and lacking in upper overtones.

 THE TWENTIETH CENTURY

The beginning of the twentieth century was characterised by an increase in the vibrating length of the guitar, which, with Torres, was extended, stabilising at about 65 cm. As for the tuning standard, it may be said to have fluctuated around 435 Hz, in spite of the frequent 'transgressions' reported in the literature. The extension of the vibrating length from about 62 cm to 65 cm - almost a semitone more - would entail an increase in the working tension - the string being equal - of almost a kilogram. It is nevertheless clear from the diameters provided by Pujol that the calibres used were thinner than those of the previous century and that the tension was about 'a semitone' less (bringing the working tensions close to those of the nineteenth century), but with an important innovation: the gradational tension of the first three strings, as opposed to the equal tension characteristic of the stringing of the nineteenth century:

Diámetros en décimas de milimetros *


Prima       de 12.5 a 13.5 (0.63-0.68 mm)

Segunda   de 16 a 17.5 (0.80-0.88 mm)

Tercera    de 20 a 21.5 (1.00-1.08 mm)

Cuarta      de 15 a 16 (0.75-0.80 mm esterno)

Quinta      de 18.5 a 19.5 (0.93-0.98 mm esterno)

Sexta        de 23 a 24 (1.15-1.20 mm esterno)


* In fact these are twentieths rather than tenths of millimetres, corresponding to the so-called 'PM' gradation, still in use in Pirastro calibres.


The tuning standard of A = 435 Hz and a vibrating length of 65 cm (Pujol owned a Torres guitar) give rise to the following ranges in working tension: **

E:     7.4-8.6 kg

B:     6.0-8.1 kg

G:     6.6-7.7 kg


** It is not possible to calculate the tension of the bass strings because the measurements provided by Pujol give only the total diameter of each string.


Finally, we give the measurements of three gut guitar strings dating from perhaps the 1940s or 1950s, discovered intact in their sealed packets (belonging to Lorenzo Frignani, Modena):
String Diameter Observations

E     0.64 mm 'Perfect' greaseproof paper packet, France. Medium torsion.

G     1.02 mm 'Perfect' greaseproof paper packet, France. Low torsion.

G     1.05 mm 'Celesta' greaseproof paper packet, France. Low torsion.


These measurements - even though the B string is missing - correspond perfectly to those given by Pujol, and confirm the greater gradation of the working tensions by comparison with the nineteenth century, as a result of which the strings feel equally stiff to the touch, a criterion still adhered to in the production of guitar 'sets'. In order to facilitate a comparison with current practice, the range of working tensions found in the 'medium tension' sets of a number of commercial firms are given below (A = 440 Hz, vibrating length = 65 cm):

E:     7.8-8.1 kg

B:     6.0-6.2 kg

G:     5.7-6.1 kg

D:     8.1-8.3 kg

A:     7.9-8.1 kg

E:     7.0-7.2 kg


THE ADVENT OF NYLON

With the appearance of polyamides - we are on the eve of the Second World War - the scenario changed drastically. It is well known that Andrés Segovia was a driving force behind the search for new synthetic materials that could be used as a substitute for gut at a time when, because of the war, all available gut was used in the production of surgical thread: it was quite impossible to find gut strings for musical instruments. It was against this background that the American Albert Augustine, one of Segovia's guitar makers, had the brilliant idea of trying to fit his guitars with a synthetic thread called 'nylon' recently invented by the Dupont Company, which was used for fishing, for women's tights and for the parachutes of American soldiers. The question arises whether the transition from gut to nylon affected the sound of the guitar to any significant degree: the answer is yes, but with a number of qualifications. The first consideration relates to the acoustic and mechanical properties of nylon: without wishing to go into too much technical detail, the advantages of this material are that it absorbs very little atmospheric humidity, it is inexpensive, its surface is perfectly smooth and it has a considerable resistance to abrasion. On the other hand, its specific weight is slightly less than that of natural material, and this translates into a sound that is less brilliant and spontaneous than that of gut, which, as rightly observed by Stefano Grondona, tends to be much more similar to PVDF (Polyvinil dilen -fluoride, incorrectly referred to as 'carbon') than to polyamide. The advent of nylon thus brought about a radical break with the acoustic tradition that had always been associated with gut strings. But the real revolution came in relation to the bass strings: in multifilament form, nylon has a resistance to traction that is so much greater than that of silk as to permit, for the first time in the history of this instrument, a consistent reduction in the diameter of the core in favour of a consistent increase in the thickness of the metal wire. The result was an astonishing improvement in the general acoustic characteristics of the bass strings (greater volume, richness of overtones and duration of sound), practically unknown to guitarists of the nineteenth and early twentieth centuries. The advent of synthetic materials and of metal strings dealt a fatal blow to the glorious, principally Italian tradition of string making: many skilled string makers emigrated to America, where they set up firms whose names still testify to their country of origin, while those who stayed behind turned to the production of surgical thread. And so it was that, with the disappearance of the last of the old string makers, a tradition that had endured without interruption for centuries - passed down orally and professionally from father to son - finally came to an end. But gut continues to be studied: nowadays research is directed not only towards the reconstruction of the strings of earlier times but also towards the development of materials that possess the acoustic characteristics of gut without being marred by its defects, such as its high cost of production, its instability in the face of climatic change and its poor durability. But that's another story.