Recent developments. November 3 2004

 

The document How to make silk strings for early instruments originally was intended to be presented as a commentary in the FoMRHI quarterly. This explains some references, like the one to the ongoing discussion between Segermann and Peruffo, as well as a general hands-on presentation intended for people with skills and knowledge. I still keep this last intention. This web site is not intended to advertise anything, and/or entertain anyone.

In the past few months I realized that certain points in descriptions of string making processes needed clarification. Some of these are in a domain of developing skills and practice.

One such task is twisting a silk string without developing waves in it, i.e. making it round and smooth without material bulking up in regular patterns. I ran into this problem a few times, and a couple of people trying to make strings on my encouragement questioned about this as well.

I apologize for my explanation being rather empirical and by no means complete.

Making a smooth string of silk, where the filaments are twisted in one direction, appears to be a simple task. In practice, however if certain conditions are not known and understood, the silk will bulk up in regular intervals, making the string “wavy” and less pure in sound. I am speaking of the “raw” or gummed silk. When it is moistened, the sericin gradually changes its nature from dry and quite slippery to somewhat gummy, and if moistened some more, to quite slippery. I had the best twisting results in the initial moistening stage. The waves developed if the string was allowed to get “soggy”. In reality it is easier done than explained.  The string is moistened just until it retains the water, which happens in a few seconds of application, and twisted right away under stretch.

The second aspect that developed some more, is the degree of twist applied. Basically it turned out that most gamba and violin players are accustomed to the response that a string of rather low twist is offering. This degree of twist corresponds with one favored by romantic violinists. As heavy vibrato and “fat” sound are in general so effective in covering the natural string qualities, a few shivered when the violin family went “metal”. For a viol player such low twist string offers about the maximum dynamic volume, with less interesting tembral quality. It probably offers the best in playing French viol music. However in viol consort it requires much more ability to control and much more mastery, and still makes the ensemble sound somewhat rude. I’ve been running a few amateur viol groups, and the only time they did not irritate the listeners was while strung with lower diameter – tension, higher twist strings. The tuning problems became far less pronounced as well, while the blend was often exquisite.  Personally I am convinced such considerations have to be taken into account when deciding on diameters and twist degree. Baroque violinists who tried silk strings overall favored the sound produced by higher twist.

One interesting technical device discovered, was the use of starch in loading silk strings. In traditional Chinese string making starch is an important part of string glue. This glue commonly is prepared for one time use. It is disposed of after making a batch of strings. However I realized that starch (wheat starch is my favorite) can be used many times without a risk of spoilage, and can be applied before the silk is twisted, thus giving new possibilities. To this end I make a wheat starch paste. Not cooked. Just suspended granular starch in water. The paste can be made at needed thickness. For that matter, as the starch quickly falls out on the bottom, I just mix it in with my fingers to a needed consistency and also apply with the fingers. The silk is prepared in the usual manner. Depending on how thick the paste is (and as the starch has a similar gravity mode, it basically replaces some silk) up to 8% less fiber can be used. For the top lute, violin, guitar and such strings I can use a very light slightly milky mixture, without reduction in the silk amount. The silk is moistened slightly, as in the usual technique (makes absorption easier) and then the paste is applied with the fingers. The string is twisted to a needed degree and wrapped on a copper drum. After that my preference at the moment is just to cook it in near boiling water for 2-3 minutes, until the white starch turns transparent. This can be easily observed (we all know how the starch works). This can be done as well in olive oil, or linseed oil if necessary. The idea is to have starch burst and turn to permanent glue condition. Other substances just provide the finish for the surface. The string is stretched and wiped of any unneeded starch, then dried and rolled. That's all. Resulting string shows very good sustain and is very even.

Different starches provide different amounts of loading and adhesion. It can be clearly felt with the fingers, depending on a grittiness of the paste. Flour works, but being slippery, less of it adheres to the silk.  The starch paste can be allowed to dry and remoistened numerous times.

With this technique I had phenomenal success using spun silk thread, especially for the top strings. They practically don't stretch, and last longer. The reeled silk has a tendency to loose long filaments on the frets and then come undone, the spun silk does not have this problem, filaments being short.

If the mixture of wheat (other starches work much worse) starch powder and copper powder is used, the string retains an amazing amount of copper, making successful smooth basses. Also I make rope-twisted basses this way and core-winding basses. All of them are very stable, loud and clear.

I personally find this technique very easy, quick and clean. I don't even bother much about using distilled water, the more grittiness, the better! The starch distribution is very even after twisting.

I tried mixing in hide glue, gum arabic etc, but all of them seemed to reduce the initial advantage of the starch granules sticking to silk filaments.

However one use of hide glue seems to produce very good results. It is an application of hot glue after a string was cooked and before stretching and drying. This sort of finish appears to improve the sound and the string resilience. Right now I have no data on how well it interacts with the fingers of that unfortunate group whose finger sweat destroys gut strings. Both gut strings and hide glue consist of collagen, but no assumptions can be made without an actual use data.

If rabbit skin glue is used instead of hide glue, the string remains flexible, and even more so if technical gelatin is used (to those who are concerned with BSE, please do your own research). Hot gelatin can be applied to silk even before twisting. The resulting string by all parameters is reminiscent of a gut string.

There is a danger of the string coming out not completely round when it is cooked on a solid cylinder, like copper. Using a rubber sleeve over the copper will keep the silk string from losing its round shape in cooking. This is especially important for thin strings.