Carving skills

My latest violin received a very special treat this summer by a fantastic french wood carver who kindly put his skilled hands to it during the Oberlin violin makers workshop in Oberlin, Ohio.






Olivier, who is a fine wood carver working at Bois d´Harmonie in France first drew up the design and than carefully decorated the finished ebony crown.

Holding everything together- Animal Glue

A violin is made up of various parts and the stuff which holds them all together is animal glue.

There is a long tradition for this kind of glue. The first and earliest finds of pieces of furniture glued with animal glue and descriptions of how to glue pieces of wood with animal glue date back to 1400 B.C Egypt. Greek philosophers wrote about the preparation and use of this glue (e.g Homer towards the end of the 8th century B.C). Also the Romans knew about the techniques and practices. The Middleages left us manuscripts dealing with the making and useage of various animal glues (11/12th century)

For violin making and restoration the material is absolutely perfect and indispensable with any modern type of glue. The biggest advantage is its reversibility- once the glue is dry and set it can always be returned to gel with water. In practice it means you can for example reopen an old crack if necessary e.g when it became dirty with time or was not glued level. For glueing the back and front to the ribs a weaker kind or more diluted glue is used to ensure the instrument can be reopened for repair or more vital the plates can pop up at the seams when shrinkage of the plates occurs during the drier seasons of the year- this loosens tension in the wood and avoids the energy built up releasing in a crack further in the plate (dry-crack). Animal glue fairly colourless which is good for glueing cracks invisibly and it does not creep with time.

The strength of animal glues is very good too. Different kinds of glues are available and it´s strenght can easily be manipulated by how much it is diluted with water. Types of animal glue are Bone glue (made e.g from Cattle bone), Skin glue (e.g from Cattle-skin), Rabbit glue (made of rabbit skin), fish glue (fish skin) and techical gelatine (made of cattle skin and comparing well to gelatine from the food industry).

Glue in granular form is soaked with water for 1-2 hours or even overnight, heated up to about 65C, taken out the water bath to cool once and than reheated for ideal strength. I store the glue in the fridge if not used to make it last longer and not getting mouldy- it is a natural material after all. It is liquid when warm and sets when cold.

The process from raw material to the glue-granulate we use takes about 90 to 100 days. The animal parts are being cleaned, cooked and degreased, cooking releases the collagen from bone or skin into the cooking water. Filtering, steaming down and drying brings it into the customary form.

The secret of the glue are the extremely though fibres of the collagen. Collagen is protein found in human and animal fascia. Our cartilage, skin, tendons and ligaments are mainly consisting of collagen.

One “measure” for sorting animal glue is the so called Bloom number. According to Dr. Kremer ( the bloom number is defined by the flexibility of the glue. The granulate is cooked up with a certain glue to water ratio and poured into a plastic cube of a set volume. Once cooled and gelled the glue-cube is taken out and set under pressure with a metal ball pressing down on the cube until it starts to tear/disrupt. This measured pressure is the Bloom number. The higher the bloom number the more flexible the glue. A flexible glue can be good e.g for glueing a patch in a repair- since the glue bond between original wood and the new wood is constantly vibrating when an instrument is being played.

Glue adhesiveness is greatly affected by its fat content. Less fat means a better bond.

I like to use Körnerhautleim in Würfeln (cattle-skin granulate) and a new Kremer product Rabbit skin glue for stronger bonds.

Here are some photos to show what the granulate can look like before it is soaked in water and cooked up and a picture of the prepared glue ready to use. (First picture: cattle-skin glue on the right and a japanese glue Harima S1, very light in colour on the left)










Ebony crown

Violinist and violin have a close relationship. Although the player seems to age faster, the violin does too!
One part of the instrument which is unwillingly shaped and worn through the decades of use is the button.
The button is part of the maple back being left as an attachment to the upper curve of the outline to reinforce and hold the neck in place. Original cremonese well preserved buttons are rare to find, mostly the details and initial shape have been rounded off and smoothed out or in many other cases an ebony crown has been fitted.
Installing an ebony crown can become a necessity if the button is too small and/or too low. The hight of the button determines the shape of the neck root and if too low neck heel strength may be affected or the look becomes an issue. I quite like the look of a nicely shaped ebony crown but the drawaback always is that original wood has to be removed through the process.
For my new instruments I don´t have any conflict with removing original wood, my crowns are part of the antiquing.

I would like to show a few photos of how an ebony crown is made and also some pictures of original Stradivari buttons with wear and with a crown.

Stradivari Joachim

Here is a slightly worn Stradivari button. You can still see the decorative comma cuts on the left and right where the button and the outline come together. The shape is completely rounded away and smoothed over the centuries.





This is an early Stradivari button on  a violin from 1683.

Note how large the maple part still is and how off-centre the centre joint sits. Todays restorers probably would have tried to save an original button in this condition




Ebony circles cut out

Making a button crown starts with a piece of ebony about 6mm thick, I use a forstner drill bit to drill a hole on the pillow drill.







crown template Guy HarrisonThis is a template I use to mark out the crown on the maple and also the ebony part.






crown fitted and glued

Here you can see the crown fitted to a violin of mine.
Since I fit the neck after varnishing the neck is still rough and slightly large to be reshaped after the crown is installed.




the crown from above

This is the same crown from above.







Playing jig

Since I´m not only making instruments but also playing violin I´m always seeking ways on how to improve my technique and tone.
Working with my teacher on the ease and quietness of slides from one position to another she recently suggested to cover the left thumb with something soft and smooth which could help reducing the slight tension between thumb and the violins´ neck and getting a feel for a smoother shift. She made me try with a glove, wearing only the thumb part having the rest hanging aside. It felt good and worked quite well. To get rid of the four useless fingers of the glove I didn´t want to destroy my glove but decided to knit a little thumb-cover for practising.
I can only recommend to try this!

practising violin helper

Stradivari scroll- totally scratched!

Tool marks left by the old masters are always fascinating to me and I keep looking for them trying to interpret the type of tool which created them, the way of handling it (direction and depth of the cuts for example) which may at the end give me a clue about the overall working method.

On this Stradivari violin from the makers golden period some interesting scratches can be seen, mainly on the pegbox walls. My assumption is that they are the result of a very strong and agressive abrasive, possibly sharkskin.

Very prominent marks on the pegbox walls of this golden period Stradivari violin

Very prominent marks on the pegbox walls of this golden period Stradivari violin.









For a while now I have been useing dried skarkskin on my own work, experimenting how deep the marks need to be in the white wood to be still apparent after the whole varnishing process. Most of the scratches disappear during wetting and scarping the wood- the first step which is done to the raw wood in order to let it swell up, raising the grain a few times before putting the ground and varnish on to have better control over the final surface texture.

Pebgox wall of my own scroll trying to get a similar effect

Pegbox wall of my own scroll, trying to get a similar effect. Most of these stratches will go away during the wood preparation, ground- and varnishwork.










If you´re interested in how sharkskin can be prepared from fish to finish please visit this blog by Charline Dequincey, a colleague in Canada who presents a detailed photo essay on the subject. Shark skin

This is a closeup shot of the skarkskin I am useing

This is a closeup shot of the skarkskin I am useing.







A new Strad model- new to me!

I had been working with my 1703 Stradivari violin model for a while and got the urge of starting a new instrument, I chose a well preserved 1714 Stradivari which I know sounds great and about which I have all the information I need.

Steward Pollens book “The violin forms of Antonio Stradivari” is an excellent resource when it comes to identifying an outline tracing of a Strad- which means to find out on which specific mould the  violin had been made. The original moulds being preserved in the Museo Civico in Cremona together with other artefacts of the Strad workshop were descriptively named by their creator- one form called “PG” would stand for “piu grande”- “a little larger”. The name, often the date and notes for each mould are either written on the wood with ink or cut into the surface. Here is a photo from the book showing the P-form.










To find out which original mould I am working with when choosing a model  I check my tracing and outline templates by imposing them over the original moulds presented in the book.

Since I use tracings of the original outlines which are typically assymetrical in their left and right half I try to correct some of these quirks. These deviations are naturally created during the making process rather than being part of the initial design. All the curves in a cremonese violin are made up of segments of circles- to have this in mind makes it easier to understand how the shapes come together.










Between my 1703 model and this new 1714 violin some slight differences in the outline are to be seen. To make it clear I did a tracing of one template and placed the other violin template above it. I hope the differences show up well enough. (1714 drawn on below, 1703 template placed on top)































The 1714 violin fits very well with the P-form (this is the form pictured far above). Upper and lower bout of both my models correspond well with one another. The main difference is to be seen in the C-bouts, which for the later violin are a bit longer as well as more upright and straighter, which gives the violin a more masculin look. I assume they have been made on different forms (since the P-form is dated 1705- too early for my 1703 model). However it is important not to forget the influence of the blocks. If you look at the last photo, this change in outline shape could easily be created by leaving the corner block slightly proud of the from. It is often subject to speculation matching a violin to its form. A lot of the changes in shape from the mould to the instrument happen during the making process which is another thing to consider. Stradivari was a very meticulous maker, over his long working life he made up a large number of moulds form his busy workshop- as the foundation for each new model if he felt the need to change the outline from one to the next instrument. Guarneri del Gesu on the other hand is said to have used 2 or 3 moulds for his whole career (none of them still exists). The outlines in some of his instruments show drastic variations. One explanation which sounds plausible is that he changed his models by simply working the blocks differently rather than spending the time on making a new form.

Problematic Bassbar

Recently I worked on this violin which was made in Mittenwald around 1850 and had some issues which needed attention. Several cracks in the front, one of them running along the bassbar as well as the bridge and soundpost being not ideal in any sense. The closest sound description was “tin can without at lid on”- very hollow, uneven, a sharp e-string and some wolfyness.

Once the top was separated from the rib garland the reason for its bad sound was revealed- the bass bar was coming off. Since I posted about bass bars in my last blog entry I thought this would be an interesting addition.

The violin is now back together with all the cracks repaired and a new bass bar installed. 200% sound improvement! It will be played by a young violin student.

violin Mittenwald 1850








violin Mittenwald 1850

The Bassbar

One structurally and sonically very important feature of violins, violas and cellos which is hidden on the inside of each instruments box is the bassbar. A bar, traditionally of the same material as the front, is glued to the inside of the belly, placed somewhat parallel to the centre joint but outwards crossing the bass foot of the bridge. It counteracts the downward pressure from the strings through the bridge and is an important acoustic support opposite the soundpost. Material choice, placement, fit and its shape are critical to achieve the best tonal outcome of an instrument.

Historical bassbars are rarely found in their original place, we are lucky that some have been carefully removed and being kept in museums or with their instrument. They look rather different from a typical modern bar, a lot shorter and lower sometimes made of slab cut spruce.

I will show a few photos of how I install a bass bar in one of my violins.

This is a finished Stradivari model violin front, the next step is to fit a bassbar. I use light spruce (with a density of about 0,38) perfectly split and with medium grain, my preferred thickness is 5,5mm.

Stradivari model Martina Hawe belly with bassbar

I start the process by finding my desired position, depending how much I want to stiffen up the front I can alter the inclination. Traditionally I determine the length of the bar by marking 4cm from each end of the violin plate. I than divide the upper and lower bout width by 7 and use the first mark away from the centre joint for a rough guide. Most important placement guide is the bridgefoot, I want the foot to outreach the bar by 1 to 1,5mm.

Stradivari model Martina Hawe

Once I have marked the position I use a chisel to roughly fit the bar to the shape of the front.

Stradivari model Martina Hawe

This is a frame I like to use- the belly plate is kept in a flat position while fitting and glueing the bar. I want to fit the bar to the belly and not the belly to the bar which would happen without the frame as the front is more flexible than the bar. Without useing a frame there is a risk of deforming the long arch when the bar is glued in.

Stradivari model Martina Hawe

Here you can see the bar in the fitting process, the profile is roughly cut down.

Stradivari model Martina Hawe

Once perfect fit is achieved I use hide glue as reversible glue and quite a few clamps to spread the forces as evenly as possible.

Stradivari model Martina Hawe

After a drying time of 8 hours plus I take off the clamps and shape the bars profile. Flexing the plate with my hands during the shaping process tells me how stiff the structure still is and where to take off more wood. Here it is nearly finished, the ends still need to be cut down in a slope.

Stradivari model Martina Hawe

Some points on purfling

Purfling is the name of the black and white inlay which was traditionally used by most classical instrument makers. Besides its decorative appeal it also protects the plates. By running across the grain at the top and bottom end it can stop a hard impact to the instruments edge from traveling forward into the plate along the grain line as a crack.

Historical purfling shows a wide variety in material, width and coloration.

Different species of woods can be found, maple, pear, poplar, walnut, beech, spindle wood, boxwood, ebony etc. The Dutch tradition was to use whalebone. Makers of the Naples school have often used paper- the city was big in paper manufacturing at that time. Occasionally you see purfling being only scratched in and filled with black ink, the Testore family comes to my mind- it saves time. Stradivari has used poplar for the white middle section and black stained pear wood for the dark sides. The material of choice has an effect on the final look but also on the usability, as the strips are bent into shape like the ribs on a hot bending iron it is nice to work with a flexible material. Ebony for example being brittle makes bending a big challenge and once it is bent and set in it gives a very stiff look to the lines and a strong look for its inherent blackness- ebony was traditionally used by some French makers.

Purfling is commercially available in various materials and thicknesses but I prefer to make my own as to me it looks more authentic. My method is to cut up a thin poplar strip on a bandsaw, plane it down to the desired thickness- this makes the white part. For the black sides I stain pear strips with ferrous sulphate and logwood chips to my desired blackness and glue the strips black-white-black together with hide glue. Once it is dried I cut it up into strips. The width of the black is adjusted at the bench for each individual instrument- I use a scraper and thin the black part down until I like it and/or it fits the model I copy. Thinning them down by hand also adds natural variations in thickness which is most times to be seen in original Stradivari purfling.

martina hawe purfling

The inlay is done by hand, some makers use a purfling dremel to cut the groove. I have made my own purfling cutters, the way they work is historically approved. A hand held shaft with a fixed metal bar which rests against the plates edge- an adjustable steel blade which can be set to the desired distance from the edge and a depth stop to control the cutting depth. I have made two of those cutters, one for each side cut of the purfling groove.

purfling tool

cutting the purfling channel

Once the cuts are made for the groove the wood in between has to be dug out. I have two tools called purfling pic for this task.

Purfling pic

Before marking out the purfling groove on the instruments plates I like to do a pretest. I use a piece of maple to inlay a pice of my purfling. This ensures my purfling cutters are set correctly, I want to avoid having a groove that is too tight or too loose for the purfling material. I also glue size my purfling groove one time to make sure the grain is saturated with glue before I do the real glue job. This is what the test looks like. I also like to keep this as a reference for the next instrument.

glue test

I have chosen some photos of historical purfling just to show you the variation that can be seen.
Guarneri del Gesu 1736
Guarneri Ysaye violin

Two photos above of Guarneri del Gesu violins of 1736 and 1740. Del Gesu purfling typically shows variation in thickness and the material used has a soft look and feel to it- it seems to bow to the slight unevenness in the groove.

Below you can see some purfling samples of the Brescian school, a Gasparo da Salo viola and a Maggini violin. Rather crudely made and inserted- matching the rest of the instrument very well- they have their own charm!

Gasparo da Salo viola
Maggini violin
Finally a corner of a Stradivari violin. Width and colour of the black strip give a strong impression and nicely accentuate the lines of the outline. There is a little mess in the purfling mitre going on, some black filler being used to fill gaps there. Looking at purfling in detail and close up almost exclusively shows imperfections and mistakes- however from a distance the impression is excellent. A little bit like a masters oil painting which doesn´t make sense when you look at the single brush strokes but is magnificent being looked at from a certain distance.

Violin Neck Reinforcement

String instruments are under constant tension and subject to changes.
The forces of the strings pulling the neck forward, their downforce being transfered through the bridge into the body of the instrument, humidity changes generating expansion and shrinkage of the wood in various ways and of course players inducing energy causing the system to vibrate which creates the sound hear.
A weakness in the structure is the point where the neck meets the body which is called the neck root. The modern neck is set into the top block with a dovetail joint, as opposed to being nailed onto the flattened top rib as in the baroque construction method.
The neck itself is stabilised by the ebony fingerboard which in theory, if the wood has been seasoned well and has a decent thickness, should stay the way it is glued on- flat.
The parts which are more likely to bend under tension and humidity changes is the neck heel- the underside of the neck root and the curved endgrain area where the thumb lands in third position, which can result in a slight change of the neck angle which than affects the playability and sound.
This problem is seasonal and usually more dramatic on the cello than on the violin. Still this can happen and in order to prevent that I reinforce my necks with a wooden dowel.
It is firmly glued into a predrilled hole, cut down and at the end covered by the fingerboard. I have been useing this method for a few years now and it seems to help against unwanted changements of the neck angle.
Another area where elevation changes are born is underneath the fingerboard at the point where the spruce front leaves the top block glueing surface rising into the long arch- I carefully work this area when I shape my archings to gain ideal stability there.

Below I include a short photo series to show the simple process of installing a dowel in a violin neck.

The neck root is now ready to receive some extra strength. I have drawn on the angle to show you where the dowel will go. For ease of drilling and to ensure the correct angle I have made a wooden block with a pilot hole which is attached onto the fingerboard surface during drilling. The dowel is of beech and 6mm in diameter.

The guidance block is clamped in place and I am about to drill the hole with my power drill.

The hole is drilled.

I use Titebond as my preferred glue. A hammer is needed to drive the dowel into the tight fitting hole.

Once the dowel is in I cut it down with a gouge, slightly hollow to allow some swelling when it comes in contact with hide glue when the fingerboard is glued on.