How
to make silk strings for early instruments
Why silk?
I am not going to make claims of authenticity for silk strings being
used on medieval, renaissance or baroque instruments. This discussion seems to be ongoing (see Comm 1796 by John
Downing vs Comm 1767 by E. Segermann).
Missing from this discussion are the actual strings. Unless silk strings are made, tried by
musicians, and collectively developed, it is very easy to dismiss their
usefulness or their attractive qualities.
These same qualities should be a part of the argument of whether our
predecessors would be interested in going to the trouble of making silk
strings, or they would be totally content with sheep gut. Some of the best arguments for early gut
strings are made as result of direct experience of modern gut string
makers. Similar arguments cannot be
made for silk strings because there appears to be only one person, who is not a
commercial string maker, making silk strings for early instruments at the
moment. Whatever discoveries I have
made in the process, are staying right where they were made. I believe that by changing this, and sharing
the practical process itself in as much detail as I can muster, I can give the
silk strings a better chance, in addition to contributing to the discussion.
There are additional considerations.
As a process, making silk strings is definitely simpler, less time
consuming, and can be mastered by a willing musician. The results, however, can be very satisfying. Not only can silk strings match in their
qualities the gut ones, but a wide range of twisting techniques and twist
angles can be tried. Only recently we
noticed a discourse on gut strings with higher twist than the one once
preferred by the 19th century romantic violinists. Such higher degree of twist offers
possibilities that may not be experienced with gut strings now considered the
norm. Music making for the enthusiasts
especially can become much more satisfying experience, as the more flexible
strings are definitely less demanding and more forgiving, if slightly less
loud.
Not the least, at this
moment in history, raw silk is extremely cheap. I am buying it for about $40 a
pound. That is a POUND! Enough to make
strings for over 50 bass viols. The
price of gut by the weight sometimes exceeds that of gold. For the business minded, silk strings can be
made by machines as easily as modern fly-fishing lines are made. They also can be made by hand with some $20
worth of equipment.
Silk components,
processing and their use in string making.
I will speak primarily about Bombyx mori silk, the cultured silk
used by modern textile industry. This
is the one I used mostly. I also
experimented with the Tussah, a wild silk. This helped me realize that wild silks of the Saturnae
family (very widely spread around the world, including most of Europe, primarily
southern France and Mediterranean, with some recently discovered Celtic burial
sites containing fabrics made of wild silk) can make good strings as well.
Any silk consists of two major proteins, Fibroin and Sericin. Fibroin, as its name clearly implies, is the actual fiber, with a very strong molecular structure (there is an abundance of printed source material on silk). Raw silk, the silk simply reeled from cocoons without any processing, usually comprises from 65% to 80% of fibroin. The remainder is sericin, a gummy glue, designed to keep fibroin fibers together and the cocoon intact, protected from the elements including the ultraviolet radiation from the sun. To exit from the cocoon, the caterpillar secretes an enzyme that hydrolyses sericin. Removal of the sericin from silk makes it vulnerable to UV and in the wild, leads to its quick decomposition. Fibroin, however, is relatively chemically inert and insoluble in practically all of the organic solvents. Both Fibroin and Sericin have about the same specific gravity.
The raw silk is produced by placing cocoons into a tub with hot water to soften the cocoons. The temperature varies depending on the process, but commonly is from 65C to 98C. After combining about 12 strands together, the thread is reeled off the softened cocoons. Sericin glues these strands together to produce what appears to be one continuous hair-thin filament. The usual industry standard is a 21 denier thread (Denier is the weight in grams of 9000 meters of filament... don't ask...). This is the filament I use to make silk strings.
I will describe the
process of a string being made and explain the reasons for procedures as we go.
I re-reeled the raw silk from industrial skeins onto spools
myself. Whereas many silk suppliers will
provide the silk on cones, one can never be sure how long the filament has
spent on the cone, or how deformed and weakened it has become. Also you can not be there to supervise the
tension setting on the machine. Silk
coming in loose skeins is guaranteed to be undamaged and at its strongest. However, if anyone considers reeling the 21
denier silk from a skein onto the spool, he should not undertake it
lightly. A well made vertical
adjustable reel (to put the skein on and spread it) as well as reeling device
with electric motor and oscillator (you just can’t sit there for a few hours
moving filament right and left) needs to be designed. Now to our string.
I have made a gadget to hold 26 spools with 21 denier silk (an
arbitrary number, just happened to be so).
After a couple of years I have enough data to predict how many filaments
are needed for a particular string diameter.
Incidentally the final diameters coincide with those of gut strings of
the same pitch.
We start with making a smooth, directly twisted string (i. e.
twisted in one direction), say a top string for the bass gamba, diameter 0.68
mm. First a few calculations. According to my data I will need 208
filaments of 21 denier thickness. I
need to know how long the finished string needs to be, L cm, and the
coefficient of twist, Co. The first
relates to the length of the dry filaments assembly; the second I found in
practice to vary from 25 to 37.6, with the number 30 in my experience working
well in most situations. Of course,
this number was determined experimentally from the thickness of the
filaments. The formula I use to
calculate the number of turns, N, to give the filament assembly is:
N = Co × L / #½,
where #½ is
the square root of the number of filaments.
This formula effectively gives a string of any diameter about the same
twist angle, as determined by the coefficient of twist, with the thicker ones
getting a little more twist, as they end up shorter than the thin ones.
I want to make three string lengths, so I start with L = 305 cm. I
choose a twist coefficient of 32 (enough body, but good flexibility as
well). It might be important to touch
slightly on the question of the amount of twist for silk strings. Whereas overall the strings twisted to a
degree related to gut string twist will exhibit similar qualities as gut, the
fibroin in silk has more lateral stiffness than collagen, and appears to favor
slightly more twist without much loss of strength. There is however a question of string compression on the bridge
or the nut at the lower twist degrees.
I intend to touch on this later.
The number of filaments is, as I mentioned, 208. This gives N = 677 turns to be put into the
string. I use a few hand cranked gear
boxes with different gear ratios. This
way I can count the number of turns precisely.
Another possibility is to use a computer controlled step motor, though I
personally found this to be bothersome to program and use.
After calculations I can assemble the filaments into a string. I have a board with a number of sturdy
hooks. Some hooks are needed to make
rope twisted strings, some hold weights for stretching silk etc. The board is attached firmly to the wall,
and from the opposite wall a chain is connected tightly to this board. The tightness is important. As I start assembling the string, I use
small (about 1.5 cm) brass hooks to which to tie the silk. I also have some heavier hooks for really
thick and tight strings. With all 26 filaments
(see above) tied firmly (!) to one hook, I run them back and forth between a
hook at the 305 cm position on the chain and another larger hook attached to my
wall board. 8 passes gives me 208
filaments. Running such thin filaments
requires attention, clean hands (watch for any rough skin that can break
filaments) and well fixed end points.
With every new bunch added the assembly tightens, and if the chain is
not tight to begin with, the filaments will be of different lengths. This would distribute the tension unevenly
in the finished string. After
assembling all 208 filaments, I tie off the end (learned the way from our
friendly fly fishing makers), of course without losing any tension on the last
bunch.
Now we have a bunch of filaments, which we need to moisten. For this I attach the small hook to a rope
with movable hooks, instead of the chain.
This way I do not get the chain wet, and can control the stretch of the
silk. I use distilled water to moisten
the filaments THOROUGHLY. Minerals
in water can affect silk. Japanese silk
string makers leave their silk in water overnight, to be sure (of course, they
use this also as a gentle degumming technique). In any case, the sericin takes a few minutes to soften. It is a good idea to moisten silk, let it
hang maybe 20 minutes, and moisten again.
As the silk becomes moist, it starts stretching. This provides a good opportunity to work
some filaments around the hook to ensure all the filaments have the same
general tension. Theoretically, silk
can stretch up to 25%. After allowing
the silk to soak in water thoroughly, I
start twisting, while stretching. The
twist adds handsomely to the stretching process. Twisting can be done all at once, or in stages. The important thing is that by the end, the
string is twisted and stretched as tightly as possible. Of course, with thinner strings it requires
good judgment, not to break them. Also,
it is very important that the string stays moist. If some parts of it manage to dry out, the string will not be
uniform. I moisten it regularly
throughout the process.
One particularly attractive technique of twisting, used in both China
and Japan, calls for hanging the silk assembly vertically. They build high wooden structures for this,
or use barn-like high buildings. What
is good about this, is the use of measured weights, which are turned to give
the twist. These measured weights give a
consistent stretch to a twisted string.
After being twisted, the Chinese cook the string. The reason for this is that, whereas the
fibroin will hold applied twist after being heated, for the string to be
uniform and stable the sericin must melt to penetrate the fibers. This process will create a very uniform
string, having a smooth gut-like appearance.
It is possible to make strings from degummed silk or silk thread, more
on this later.
One important aspect of cooking is to find the right time and
temperature for cooking. An undercooked
string will have one set of problems; overcooked string will have another set
of problems, one of which will be reduced strength, due to the breaking of
outer filaments. So, especially thin
and highly stressed strings need to be watched closely. I do not cook the strings without a good
timer involved. In general it is
preferable to keep the temperature 15-20 degrees C below boiling point of
water. My take on Bombyx mori
silk is from 18 minutes for thin strings, to 28 minutes for the thick
ones. There will be some variations
depending on the variety of silk used.
As mentioned earlier, the string has its own natural glue in it
already. It has to be melted and has to
penetrate the string. When it is done
under tension, the string is better packed, as the filaments get closer
together and force more sericin to the outer layers. For this it can be steamed.
I use my own steamer (a traveling clothes steamer is a good bet) where I
can put a steam tube about 30 cm long around a string, and move it over the
whole string length, with steam going inside.
Now, if the temperature is too high, the string will develop waves, so
watch out if you try this, the pipe diameter that works well for me is about 4
- 5 cm. If it is smaller, the steam will be hotter. Steaming is just one possible technique, and needs to be mastered
to work well. The advantage arises from
forcing sericin to the outside of the string which creates a natural
"finish". Alternatively this
sericin can be wiped off some thinner strings, reducing the weight of the
string. This helps the string to be
thinner for a given strength.
It is possible to go from twisting to cooking without steaming in
between. First, it is a good idea to
let the string, wrapped, let us say, on a copper cylinder, rest a few minutes
in cold distilled water. To cook a
string, this copper cylinder will be placed in olive oil at 80 -
85C. The quality of olive oil has to be
food grade, as some cheaper varieties contain undesirable ingredients. The temperature is important, as lower
temperature will need longer exposure, and higher temperature may damage some
outer filaments. Oil cooking allows
already moistened sericin to melt and shape nicely without loosing any water in
the process. Plus, it allows a string
to dry slowly under tension. For a
string of such diameter (of 208 filaments), about 20 minutes in 85C oil will be
enough. A second hand deep fryer with
temperature control makes good equipment for cooking. I use a thermometer, to be sure of the right temperature.
As the string is
twisted very tightly, very little oil penetrates to the inside. The sericin also prevents this from
happening, as it is designed to protect a caterpillar inside cocoon in natural
conditions from all kinds of substances save water. Thus the oil conditions the very outside of the string and helps
the string to dry just at the right pace.
After coming out of oil (HOT, be careful!) the string is stretched. For this particular thickness, I would use
about 4 kilos. While on the instrument
the tension can reach somewhere to 6+ kilos, but at this stage the string is
still wet, and is more fragile. The
stretching part is made possible by a system of hooks and chains, firmly
attached to the walls or ceiling, as well as some weight lifting equipment stolen
from exercise prone children.
Here comes an important part.
After anywhere from 6 to 12 hours, depending on air temperature,
humidity and string diameter, the string will become practically dry. When removed from its stretching rack, a
beginning string maker will encounter a strange phenomenon. The string will be quite stiff: Melted sericin acts exactly like
beeswax. If you ever tried to bend a
wax candle wick, you know it would be possible only if it is warmed. Otherwise it will crack. The same will happen to this freshly dried
string. It will crack if sharply
bent. While overall it is not going to
damage the string, it can be inconvenient.
To avoid this problem I roll the dried and still oily string around a
piece of wood dowel of about 5mm diameter.
I make one string loop around the dowel and roll it, still under
tension. You will hear a distinct
cracking sound, as the sericin will develop regularly spaced cracks. After this, the string can be straightened
and stretched a little more. Again, the
sericin does act like beeswax, and the longer a string is left to stretch after
rolling, the less flexible it will become as the sericin will flow slowly and
reconnect. Fortunately oil helps, and
the string will never become as hard as when merely cooked in water. The rolling also introduces some oil inside
the string, conditioning it. It is
worth mentioning here, that olive oil does not seem to present any problems to
the bows or fingers unless there is too much of it. It mixes happily with rosin, and leaves skin
soft and conditioned. String balm and
cosmetic in one !
To seal in success, the string can be cooked a few more minutes in oil
after being “crunched”. This produces a very smooth pliable string where it is
hard to see its twisted structure even under magnifying glass. If the string is intended for a bowed
instrument, at this point it is worth to start watching how much oil is left on
the string. It can get to that too
much point. A simple rag and good
judgment will do the work.
In Chinese string cooking, instead of oil, a mixture of animal and
vegetable glues is used. One difficulty
in this process is, again, that high temperature of cooking reduces the final
string strength, primarily by damaging some outer filaments. This may compromise the outside of a smooth,
directly twisted string, as well as create a certain "hairiness" in
the strings (ironically, a quality ascribed to “Acribelle” silk violin strings
fashionable in the 19th century). It is
especially of concern for the top lute and violin strings. If the temperature of the glue mixture is
reduced, the glue has a hard time penetrating tightly twisted string. The second problem is that personally, I
could not find any glue combination that would match the sound quality produced
by the sericin. It does not mean, of
course, that such a glue does not exist.
The third problem arises after twisting and cooking, as the string needs
to be stretched and dried. If it dries
too quickly, some of the outside filaments will break. In China they make their strings during the
rainy season, to slow down drying. I
felt I needed something to naturally slow down the drying, without waiting for
rain. Good glues for cooking include
hide glue, rabbit skin glue, sturgeon glue (isinglass), rice starch, gum arabic
in any combination. Adding sugar or
honey to the mixture has its own good qualities. They make hide glue and isinglass more flexible. Much can be learned and applied from such
disciplines as candy making. The
possibilities are limitless, and I encourage others to experiment.
This concludes the basic procedure for making a smooth, directly
twisted string. After drying and
stretching I measure the final stretched length, and record it - 209 cm. The diameter measured as 26 thou, 0.66 mm.
Water and… honey.
Following all the same procedures, but using distilled water at 80-85C instead of olive oil will create a very nice string as well. However the drying speed needs to be watched closely. It may be slowed down by drying in damp air. A good trick is to drop the string right after cooking into cold water for a few minutes. The brittle quality of melted sericin in the water-finished string definitely will be more pronounced. The string will have a certain dryness in sound as well. It appears that when cooked in oil, more sericin is forced outside the string, and there it flows and reconnects, creating a shiny protective layer. The combination of oil-water cooking can be tried in different ways. One observation, oil cooked strings have more stability when humidity and temperature changes, and are less inclined to collect finger oils and dirt. Silk (including sericin) is not affected by common enzymes, therefore stays happy with certain personal skin chemistry. As some know, gut can start disintegrating mysteriously.
Overall I personally could not find a way to keep the sericin at least as flexible as the collagen glue in gut strings. Adding a little honey to the water bath acts as a humectant, and lowers the flexibility-humidity barrier by about 5%, but still leaves the string “crunchy” at regular levels of humidity. For fly-fishing the Japanese made a silk-worm gut substitute, a very transparent and uniform string, in a similar way to that described above. The silk is first degummed (removing the sericin) and then cooked in a flexible glue, often a seaweed gel. However it is not easy to degum raw silk without loosing some of its strength. This, and the fact that the sericin prevents destructive affect of UV on silk, may explain the easy breaking of such a string (as experimented by both John Downing and myself). Such a string snapped at about 310 Hz (d#) on a 62 cm lute, while a raw silk top string lasts over a month of good use at 370 Hz (f#).
The “crunchiness” of sericin is one peculiar aspect that would be important to resolve. If sericin is left in the string dry and inflexible, it will crack at the points of stress, the string holder, lute bridge, or nut. The stress distribution will become very different. Thin strings, like the top lute or violin string, will very likely fail much sooner than in a well conditioned string. In bowed strings, especially at low humidity levels, dry sericin on the string surface might feel somewhat slippery. Thus the importance of finding a way of keeping the string unified and flexible. Again, there is a beauty to the sound of sericin, and a degummed string has quite a different sound. I will return to this.
Rope-twist strings.
We are accustomed to strings being smooth. Modern gut strings are polished for example, though that might not be the case for medieval or renaissance times. Above I gave the technique for making such a smooth string of silk. However there is a different way of putting a silk string together, in the manner of a thread or rope. Chinese string makers definitely preferred this way. The advantages are these: although the string is twisted, at the same time it has minimal inner tensions and pulls. Such a string settles on the instrument very quickly and stays very stable. Secondly, the outer filaments, which on a directly twisted string run on the outside, create hairiness when they fail whereas in a rope twisted string the failed filaments tend to hide inside. Thus if the outer layer is damaged by say friction, like a pick used on medieval lute or oud, the string does not start getting loose outer filaments. On Chinese instruments like the qin, the way of playing itself puts so much stress on the string that it has to be twisted like a rope. Thirdly, the rope twist has its own structural integrity, so the question of sericin, or glue in string cracking, becomes quite unimportant.
The disadvantages are: the “bumpiness” of the string (which makes it easier to pluck a lute, as fingers grab the string effortlessly, but requires some adjustment to bow). Also, the rope twisted string has lower strength limit comparing with a smooth one. For example on a 62 cm lute I would not use a rope twisted string above the second (d) string, as it will not last long.
I will explain how such a string is made using an example of a string equivalent to one described above, with the same final diameter, length and characteristics. Start with three bunches of 60 filaments each, 210 cm long. You can see that the number of filaments in this case is just 180, not 208 as in case of directly twisted string. Moreover the starting length is greater, as in case of this particular string, I will get in the end 209-210 cm out of the starting 210. It is easy to deduce that what is missing in length, goes into diameter.
I will not go into the design and principle of a rope. Information is abundant. One difference I need to mention here. Whereas in the rope making the three (or
two, or four) primary bunches after twisting counterclockwise are allowed to
twist clockwise on their own to guarantee the optimal secondary twist, in the
string making we are dealing with some very thin lines, and we want to be
completely in control of the final twist arrangement, so we will twist every
bit of it by hand (or an appropriate machine).
First the calculations. We will make an S twist string as
opposed to a Z twist. To have a
well relaxed twist, as described above, we will give the same amount of twist
to the initial bunches and to the string assembled of all three. I set the twist degree Coefficient at
29, lower than 32 used for the direct
string. Calculations give me 786 turns
for 210 cm and 60 filaments (same formula as above). I assemble three bunches of 60 filaments each, each hooked up
equidistantly from the central hook, where they all come together at 210 cm
from the board on the wall mentioned above. After being moistened
thoroughly, the bunches are twisted each counterclockwise 786 turns (it can
be done individually or to all three at once with a three gear gadget, if you
happen to have one lying around) and hooked up individually to 1 kg
weights. This puts each bunch under
exactly the same tension while twisting the assembly clockwise, otherwise the
string will be no good. For those
who will really get to making such a string,
I will mention that the spread of the bunches and weights should be such
as to allow the ends come together at about 1/3 of the final number of turns
(in the above example, somewhere after 260 turns). After clockwise twist of the
string, the three-hook end is tied together to one hook, the string is put on
copper cylinder and cooked according to individual taste. It is worth mentioning that while the equal
amounts of twist are suggested above, the primary bunches can be twisted more
or less than the final string. Every
variation will offer some differences to the final string. For example by twisting the primary bunches
much less than the final twist it is possible to get an almost smooth in
appearance string, while keeping some of the advantages of rope twist. Experiment!
Obviously, the thicker the string, the more twist it needs to have to
work properly. A string made of two
rope parts instead of usual three, will take more twist, and be more flexible.
Strings made of
degummed silk and silk thread.
It is entirely possible to make some very good strings from degummed
silk. Some musicians may prefer the
airy, wispy sound of silk without the gum in it. Such a sound is clearly preferred by the traditional Japanese
musicians. The issue becomes how to
keep the string reasonably manageable.
If the string is made by twisting directly, some gluing substance needs
to be applied, as the outer filaments will break and attempt to unravel. This substance needs to be light and strong
at the same time. One definite oriental favorite is the so called Chinese Wax,
now impossible to find in the West.
This wax melts at slightly above human body temperature. Also, as the string will be very pliable and
soft, the twist in it, especially in lower twist coefficient numbers (on my above
described scale, from 25 to 31) will tend to redistribute from the high stress
places like nut or bridge, increasing the possibility of breaking filaments
there. This will be more pronounced at
the lower twist degrees, as the string will tend to “flatten” on the nut or
bridge without sericin helping it to stay together. Certain heat-steam-water treatments can be devised to make the
filaments stay better together. It will
require fine judgment, however, as it will be easy to stress and damage the outer
filaments, and wind up with extremely hairy string. To remedy this Japanese string makers polish the string with rice
paste, which removes fine broken filaments and fills in a starchy paste. While
this procedure sounds simple, it requires a good amount of practice or
instruction. The rope twisted string,
however, does not have all these problems, and might be a better way for
degummed silk. There are some threads
on the market made of continuous filament silk, for example Japanese “Tire”
thread. (a common silk thread, like cotton or linen, is made of short broken
filaments, called spun silk thread).
The price of such a thread (while it is a very good quality silk) is
astronomical compared with raw silk.
Spun silk thread can make some reasonable strings, but you always need
to consider direction of twist already in the thread and calculate the string
accordingly. I should mention that
cotton, linen and hemp can make some surprisingly good strings, as hard as it
might be to believe, following the techniques described above for the silk.
These, in my opinion, work better on bowed instruments than plucked.
Other string
treatments worth trying.
One possibility is a tanning process, using tannic acid or tea. The string is still cooked for about usual amount of time, in strong tea, or using tannic acid, stretched, dried and “crunched”. But then, it needs to be left alone for a couple of weeks (read up on tanning process, if curious). I found that tanning in silk continues at the usual humidity levels. The result is a very flexible, uniform, well sounding and strong string, good for the top string of a lute or violin. The color is nice, too. All together such a string reminds somehow of a well tanned leather.
Another treatment is linseed oil. Linseed oil has been used for a few
hundred years to treat silk fishing lines.
As some hundred + years old silk fishing lines are being used even now
(by crazy enthusiasts, but aren’t we all, aren’t we all…), this means that
linseed oil offers silk a good protection against the elements. One definite fact (I claim this not only
from my experience): the silk needs to be degummed before any serious
interaction with the linseed oil (save just light rub-on finish), otherwise it
will spread unevenly and the string will be a mess. (The silk is degummed
usually with soap-detergents in water.
However I find that about 30-45 min cooking in water with about 5-10%
alcohol in it, degums nicely, without as much damage to the filaments. This method is not used in industry probably
because of the cost, and maybe the addictive nature of alcohol.). I would suggest cooking at 80-85C in linseed
oil, but working with linseed oil is much more difficult affair than
with olive, or almond, or even walnut oil.
It is neither very safe, nor could I come to a definite conclusion as to
how well it works under a bow, or finger.
On bowed instruments I did not use linseed oil treated strings in all
the seasons. It works very nicely
outside of the heating season (low humidity).
In plucked instruments there was an element of whistle – finger against
string, possibly needing the addition of a varnish material, like copal resin,
or such. The quality of oil might be
very important, using raw oil instead of boiled or vise versa, but then the
whole affair becomes so complex as to discourage one with other things to do
(to play music, for example). Another oil with similar potential and
difficulties is tung oil, or a mixture with linseed oil. Definitely both are easier to use as a
finish rather than in a penetrating bath, but this does not reduce possible
advantages. Linseed oil was
favored in many uses all the way back to medieval times.
Loaded bass strings
and Demi-File.
It is claimed by textile specialists, historical preservationists and
chemists that silk readily absorbs up to 300% of its own weight in
metallic salts. This is rather large
number. A string made of such silk
would sound more than an octave lower.
Apparently many old silk American flags from Civil War era were made of
silk weighted with lead and mercury salts.
Makes you shiver, doesn’t it !
Starting with the 20th century the industry standard became
tin salts or aluminum salts. Both are
not as heavy, but also not as poisonous.
I am not chemically inclined and did not go for the whole weighting
process, including acid baths and such.
But I did try a simple watercolor-like technique, using a copper powder
pigment in a gum arabic base. To work, the silk needs to be degummed, and then
what really is a copper paint, is applied to filaments which are then
twisted. There is a reasonable gain in
mass of the string, and a certain interest in the sound In my experiments, pitch was not as stable
with temperature change as with plain silk strings. Nevertheless metal weighting of silk, followed by finishing with
linseed oil, would be an easy way to make smooth bass strings that would match
in appearance the iconographic evidence from the 17th century.
If such a procedure was ever used to make bass strings (I am threading
this ground very carefully, with a respectful hand wave to Mimmo Peruffo and
Ephraim Segermann), those strings in all likelihood would follow the fate of
all the old weighted silks – they disintegrate disastrously after mere hundred
years or so, like those old Civil War flags do.
While silk weighting requires some chemical witchcraft, there is a good
technique of making bass strings that would use a rather simple machine. Mine is made of two old kitchen mixers on a
frame and a sewing machine pedal switch.
They are arranged against one another, to turn the string core, on which
winding is made. Older versions would
use small boys turning a handle.
Definitely a simple way to make say a nice low D for gamba using an old
C string. No wonder 17th
century instrument makers and even some musicians used to do it right at
home. I am speaking of the Demi-File
technique. It can be done on a smooth
string, which does not require my expertise, or the wire can be inlaid in a
rope twist. To make the string tighter,
I usually inlay the wire (the 2 section rope twist seems to be better) a few
turns before the desired amount of twist, and then twisting the string together
with the wire in it. The wire does not
stretch, really, so it can break if it is twisted too much. There is a way to make a Demi-File string
without kitchen appliances. First the
silk string sections are twisted (counterclockwise usually) and the wire is
twisted the same way as well, then they are brought together and twisted
clockwise. This requires a bit of
practice and experimentation, and results can be varied. Again, experiment.
Overspun basses of
silk.
Everyone is familiar with wire-overspun bass strings, with wire wound
around gut or synthetic core. If you
aren’t, you are reading the wrong magazine.
Silk overspun basses, with silk wound around silk core, were made in
China for at least 4000 years. As I
found some problems with traditional technique, (which can be seen at http://www.silkqin.com/02qnpu/05tydq/tystring.htm),
as far as applied to western instruments (for example the strings are wrapped
while wet, after drying and silk expanding, the wrapping becomes a separate
buzzing entity), I developed my own technique, described here briefly.
The core and wrapping are calculated and made separately. If the core is thick and the winding thin,
the string will be less flexible but have more ‘body’. If
the proportion is changed towards more winding thickness, the string will
have more flex and sustain, but less body.
The core can be made in a variety of ways,- smooth, rope of three, of
two or four parts. Different
coefficients of twist will produce different results.
Also, a demi-file core can be used, to reduce the final string
diameter. The wrapping is calculated as to diameter and length and twisted
separately from the core. To work well
the wrapping has to be twisted to the highest degree it can take. For silk and above formula, Co would be
37-37.6. A silk or even cotton, linen or hemp thread can be used as wrapping,
as long as it is uniform and tight, and the results will be good. The important part is to do all the wrapping
dry, since wet silk still expands, and buzzing will result. The ends of wrapping can be fixed with some
thin silk filaments and a bit of glue (I use gum arabic). Wrapping with silk and wire next to each
other makes a good quality strong string.
In conclusion.
I hope this brief description will convince more people to try making strings of silk, as the process is not of such complexity as gut string making. As these strings are made and tried, some different acoustical possibilities will be discovered and new inspirations created. The author has tried to be brief but clear. If some reasonable questions arise, he would be willing to offer clarifications. The preferred way of contact would be by e-mail at vokaria@yahoo.com. I intend to set up a web site with some pictures of strings, and other relevant information, such as suggested numbers of filaments for different instruments etc. The link to this site is http://www.globalissuesgroup.com/silkStrings/
I would like to express my appreciation to Alan Dobson, John Downing, and everyone who gave me this or that piece of information to ponder on.