Friday, December 20, 2019

A Special Kaiser Theorbo

After nearly two years I have finished an elaborate model of the Martin Kaiser theorbo, E.24, conserved in Musée de la musique, Paris.

Please take a look at the original Kaiser by using the MIMO portal. My story will make a lot more sense.

Enter "Martin Kaiser theorbo" in the search box and then click on the image of the lute. Be sure to click through all of the options. The images expand and can be scrolled through with the magnifying icon for a close detailed view. Keep this portal open while you read my post because I refer often to specific photos of the original lute.

The Kaiser model theorbo is not new to me. I have built over a dozen since 1997 but I never attempted to copy the detailed ivory work on the neck and extension. My clients have been young professional theorboists starting their careers so I always kept the aesthetic concept simple.

The story of this instrument is different. The client contacted me with the suggestion that I build a close copy of the Kaiser. He also wanted an instrument with low tension, low action, double stringing on the fretted strings, double tied frets and narrow spacing on the bridge and nut.

I have felt that I have drifted away from historical practice in the sense that modern players need to be heard and need to travel. I adjusted by adding a few tenths to the rib thickness or the belly and bars and adopted single stringing even though few of the many surviving historical theorbos are single strung.

I eagerly accepted the commission unaware that it would deplete my stock of black ebony, challenge  my skill and try my patience. But along the way I learned new skills and techniques and improvised on old ones. I felt I needed to do this. This is the story, in picture and words.

Where to start? I built the first few models of the theorbo using plans that I created from images of the instrument that you have seen on the museum's website and published measurements. A description of the instrument by Joël Dugot had also been published in Journal of the Lute Society of America, Volumes XVII & XVIII 1984 &1985.  Once I was able to examine the original ( March 2000) I improved the accuracy of my models.

All of my Kaiser models have been built on this mold although I have changed my technique recently. I no longer use the lines denoting the rib positions as the only means to determine the shape of each rib. Once I have fitted the edge of a new rib to the previous one I measure the distance to the edge rib of the bowl at various locations along the length of the bowl. These numbers are divided by the number of remaining ribs. I mark the results on the new rib and then plane off the excess material to the marks and then smooth them on a sanding board.

The widths of the ivory rib spacers of the original theorbo are not uniform. I recorded various widths  during my examination. I chose a median width of 2.2 mm. Visually this seems right as I compare it with photographs of the original. I also used white holly for the spacers throughout rather than ivory.

The green masking tape denotes the depth of the end cap and the two vertical lines on each side mark the two end points of the middle third of this important feature. This is a visual guide that helps me keep the ends of the ribs uniform as they pass under the end cap.

The original lute has a one piece ivory end cap with an incised black line (see the museum's photos). It is not original. I chose to use the standard early 17th century design composed of multiple pieces.

I usually build up my end caps by gluing them separately on the bowl one at a time and then cutting the finials. The work goes well but cutting the finials in place is tricky. Another technique, that I find awkward, is to construct the cap separately and glue it as one piece.

I tried a yet another method. I assembled the cap, fitting and tack gluing each piece separately on the bowl including the stepped top piece. The ends of the cap, not shown here, are left uneven. I planed the bottom edge of the cap flush with the bowl rim and then glued a piece of cotton cloth over the cap.

In the photo I have slipped a hot spatula between the bowl and the cap. The heat melts the weak glue joint and the cap comes off in one piece while also retaining its shape. I then cut the finials and glued the cap back onto the bowl using full strength glue.

The cloth support is then removed with warm water and a hot iron.


The procedure takes more time but I am happy with the result.

Before I release a new bowl from the mold  I cover it completely with clear plastic packing tape. If the bowl is too tight to the mold the tape will hold it all together while a little more force is applied. The rear end of the ribs are lightly glued to the mold but their integrity is protected by the end cap. A spatula slipped between the mold and the ribs releases the joint. I glue the front block to the front of the mold with three small drops of glue. These release when a spatula is forced into the joint.

With the plastic tape still protecting the bowl I use a curved scraper to smooth the inside of the bowl. The ribs joints are then covered with a light weight water color paper soaked in glue. The internal cap is spruce bent to shape and trimmed. A broad quarter moon shaped piece of water color paper covers the rear. This area is often the thinnest part of the bowl due to the amount of levelling that is necessary to fit the end cap and I think it needs extra protection. Only after these procedures are completed is the plastic tape removed.


I use quartered Canadian sitka spruce for the core of my necks. This is shaped and fitted to the front of the bowl. It is then covered with a hard, dense veneer in one of the many styles or designs that are represented on historical theorbos. The neck veneer of the original Kaiser is unique. It is composed of nearly 200 individual pieces of ebony and ivory arranged in chevrons.

Optical illusions created with black and white lines are popular mind-teasers. I spent a lot of time looking at photographs of the original design and decided that the white strips were slightly broader than the black even though they appear to be equal. I figured out the necessary dimensions and cut the strips of ebony and holly from larger stock.

 I laid out the design on a thin piece of plywood and using various templates to insure uniformity I glued the individual pieces of ebony and holly to the board using normal strength hot glue.

As I proceeded it became obvious that the slightest error in preparing the width of each piece would result in a misalignment of the mitred center joint.
To correct this, I constantly had to adjust the width of the individual pieces. It was a long, laborious process but the finished  result was satisfying.

Using the same technique as with the end cap I glued a single piece of cloth over the pattern. This remained in place while I planed off the plywood backing, bent the pattern to the curvature of the neck core and glued it to the neck core.

 I had second thoughts. My bending iron which is a 3 inch diameter metal pipe heated with a propane torch wasn't long enough to bend the entire pattern all at once which I thought was necessary to prevent the pattern from becoming distorted. I decided to split the pattern in half, left and right, and bend and glue separately. I also bevelled the abutting edges so the two halves would join without the possibility of a gap forming.

In the photo I am removing the cloth support after gluing the pattern to the neck core.


The result was fairly successful. I was unhappy with the alignment of several pieces so I cut them out and corrected the problem. 

I finished the neck with two white lines of holly along the edges of the chevrons and then two broader pieces of ebony that extended slightly beyond the edge of the neck core. The neck was then fitted to the bowl as shown in a previous photo.

The next step in my construction sequence was the preparation of the soundboard. The top is 6-year-old Swiss spruce. I had been working on it while I was assembling the neck veneer so it was nearly ready.

Large Italian theorbo soundboards are always impressive but I think the Kaiser soundboard is particularly harmonious due to the graceful flow of the contour, the proportions of the triple rose and the delicately thin bridge. The original Kaiser does not have a heart, but this instrument deserved one.

The Kaiser triple rose uses Leonardo's Knot as an elaborate center piece. I have always liked this design and what's better then cutting three of them?

The bridge is conserved separately from the instrument and I missed seeing it during my visit.
It is made of black ebony and is asymmetrical, larger on the bass side (13.6 mm front to back and 11 mm on the treble side) and higher on the bass side (8.5 and 7.4). The tips of the bridge are missing and whatever style they might have been can only be conjectured. The footprint of the original bridge is clearly visible and can be seen in the photographs of the instrument. To me the circular outline suggests a sunflower or rosebud. This style of finial is usually seen in paintings of 16th century lutes and survives on several surviving lutes. The top of the bridge is decorated with thirteen  6 pointed "stars" scribed into its top surface between courses. There are two scribed lines parallel to the front and back edges. There is no lip on the front edge of the bridge and the rear does not appear to be angled. I do not know if the rear face of the bridge is sculpted in any way to accommodate the string ends.

During my examination of the instrument I measured the footprint of the bridge that remained on the soundboard ( 14.2 / 11.8) and was surprised when I saw the published measurements some time later.
A very poor cleanup of the excess glue could explain this discrepancy but the lack of refinement in the construction of the bridge leads me to believe that it dates from a more recent time.

I constructed my bridge from European plum stained black and sealed with tung oil.  I created my "stars" with a small chisel, pressing it into the top of the bridge in a criss-cross fashion. I cut a string channel for the string ends to sit in.

The ebony veneered extension is tastefully decorated with triple bands of ivory/ebony/ivory and a single ivory line dividing both the face and rear veneers in half. Furthermore, on the face of the extension the triple bands are joined  in front of the ebony/ivory striped theorbo head by a crossover band with mitred corners.


On the rear the extension narrows so that all of the elements come together as nine equal bands.


The outermost bands of ivory wrap around the theorbo head and meet in a cross-over piece behind the nut.

Bending tight curves on a hot iron is a challenge. The thought occurred to me that I am substituting  a medium dense, fine grain wood for ivory and I remembered that ivory becomes pliable after soaking in white vinegar. So I cut a test strip of holly, submerged it in white vinegar (5% acetic acid by volume) for a couple of hours, and washed it off. After I allowed the strip to partially dry, it bent easily into the necessary shape with only finger pressure.

For the finished instrument I followed the same procedure but strapped the bent bands  onto the theorbo head with dressmakers elastic and allowed them to rest overnight.

I turn my pegs by hand without templates, shaping by eye and checking diameters with Vernier calipers. Many of the problems encountered with sticking pegs are caused by insufficient drying of the wood before they are turned. I use Castelo boxwood that I cut into turning squares. These are stored for at least a year. When I start a new lute I turn a set of squares to a rough shape. A while before I start applying the finish to the completed lute I turn the pegs to their near final shape. The heads and shanks are finished last and then fitted to the lute as part of the initial setup. If they are to be black, I ebonize them with a  process using quebracho extract and iron acetate. I use the process described in:  After the pegs have dried for several days I seal them with tung oil sealer.

The theorbo's bowl, neck and extension were varnished with several light coats of French polish applied in the traditional way. This was allowed to dry thoroughly and then was rubbed out with 4F Pumice stone. The soundboard received two coats of tung oil sealer, allowed to thoroughly dry and  was rubbed out by hand.

The string disposition: 8x1/6x2
The string lengths:  88.5/171 cm
Strung in gut
Information added 12/29/19
Instrument weight : 1970 grams ( fully strung)
Original instrument's weight: 1942 (without strings)
Instrument balances at the fourth fret 
All photographs by the author.         

Monday, July 22, 2019

July Update - Theorbos and Early guitars

   It has been almost a year since I last posted. That doesn't mean that I haven't been busy. Here is an update on what I've been doing.

I designed and built a small  travel friendly theorbo to be used by a client who is playing a lot of the late seventeenth century French repertoire.

I thought this instrument should be different than the early seventeenth century Italian theorbos, even the small ones.  So I used construction features that I associate with the iconography of French lute/lutenists of the era or that anticipate lute designs that became popular in the early eighteenth century.

The fretted string length of my instrument is 78 centimeters while the diapasons are 105 cm.  With an extra peg for the fretted strings the  lute can be strung either 7/7 or 8/6 as well as 7/7.  My client's choice of diapason length was so short I could not mount the diapasons in the usual way - on a straight extension. the choices were either an extension with the theorbo head off-set from a straight extension, or  one with the extension angled precariously off the end of the fingerboard. I settled on a solution that Joachim Tielke and others started to use at the end of the eighteenth century for out-fitting  angélique or baroque lutes.

I thought an archaic rose design would be appropriate so I chose a pattern that is very close to the 1550 era Hans Frei's Warwick lute.

My client has liked several of my Venere model theorbos from the 90s so I used the same cross -sectional arching as the original Venere but with eleven ribs. I also thought that maple, in this case bird's eye, would be the most suitable wood for a lute of that era.

Here is a model of one of my Venere theorbo circa 2004 with the historically correct number of ribs (27) for comparison.

As you know I have been avoiding the use of ebony or other exotic woods when possible. The neck veneer is Claro walnut, Juglans hindsii. from California or Oregon. The board I used has a diagonal curl. I cut thin, narrow strips, and reversed the direction of every other one to achieve the pattern.

The extension is English walnut, Juglans regia, cut on the slab.

The pegs are Castelo boxwood stained black.  Smooth operating pegs are essential. Therefore, I prefer to turn my own pegs because I can have control of every step of their creation. I buy the wood as lumber and cut it into manageable boards if necessary. From these I manufacture 1 inch square lengths about 3 feet long which are sticked and stacked for storage. When I begin a new lute I cut the number of peg blanks I need from these lengths and turn them to a rough form. Each peg is indexed to fit back on my lathe. When I need a break from carving the rose I put the pegs back on my lathe and finish turning the form, free hand and by eye using only calipers to check the size of the various components. Inevitably, the heads vary in size but I use this to advantage and fit them in descending order from largest to smallest along the length of the pegbox- a feature often observed in historical lutes.

Guitar Connoisseur is an online magazine that publishes interviews and articles accompanied by high quality photos of guitarists and luthiers. Occasionally, it delves into the world of classical guitar and recently even back to the nineteenth century.


The Spring issue of GC includes this interview by client and staff writer Pat Bianculli.

For more than a year I have been working on a copy of the Martinus Kaiser theorbo, E.24 Musée de la musique, Paris. I have built a dozen models of this instrument but this is the first that I have copied all of the aesthetic details.

 In April I was nearly finished. I had strung the instrument and was letting it settle before I started the final adjustments when the bridge exploded. I took a photo (see below) and put the lute in its case. I notified the client that there would be a significant delay and offered to return his deposit. He kindly agreed to wait until I replaced the soundboard. I needed some time away from the instrument but now I am anxious to begin work on it again.

This is type of failure is troubling because I am not sure of the reason; perhaps the glue on the treble side of the bridge jelled before I applied the clamps resulting in a loss of strength while the bond on the bass side held firm, or perhaps my support caul was not positioned properly. Whatever the reason,  it took a great bite out of the belly.

I have just about finished a model of the 1813 Josef Pagés guitar that is conserved in St Cecilia's, University of Edinburgh. I spent a wonderful day with the original two years ago. You can read my post from that visit at:
In this photo I have strung the guitar with six single strings from a set for romantic guitar tuned to 432Hz.

There are twelve planetary pegs and the bridge is drilled for six courses. A second nut is grooved for an eleven string (single first) set-up.

The back and ribs of the original Pagés are Lacewood. I was unable to obtain suitable Lacewood boards so I got into my stock of planation grown Cuban Mahogany for the first time. The neck is Spanish Cedar.

The finish on the entire guitar is multiple coats of tung oil sealer. The oil has been drying since early June. Soon I will French polish the back, sides and neck. The top will be left with the oil finish.

interior of my Pagés model

As promised, some time ago, I am making a drawing of this guitar and I will publish it here along with the story of building this model.

I will soon be finished with another Buchenberg.
This one in yew at 93 and 170. A story will follow.

In June Susan and I exhibited at the Boston Early Music Festival. On the table is the Pagés and a Baroque chitarriglia after Matteo Sellas that I borrowed from a client.

We always look forward to renewing friendships from previous years, attending shows, some featuring our friends and talking to an enthusiastic public. This year was as exciting and interesting as all the others.

I wanted to have something special for a back drop so Susan created these wonderful posters of my clients when they visited us to picked up their instruments.

All photos by the author

Tuesday, August 7, 2018

Matheus Buchenberg MIM #1570 Brussels

At the end of June I was in Brussels for an appointment at the Musical Instrument Museum (MIM) Brussels to examine their large theorbo by Matheus Buchenberg, Rome 1610, Inventory number 1570.

My interest in MIM #1570 goes back to the early seventies when I was first collecting information on historical lutes. The theorbo was one of several pictured in Anthony Baines' pictorial survey: European and American Musical Instruments, Viking Press, 1966.  I still find this book inspiring and I turned to it again when I wrote to MIM to ask for an appointment.

Matheus Buchenberg was born in 1568 and worked in Rome at least from 1591 where he died in 1628 .  E.G. Baron (1727) describes him as a builder of theorbos of large proportions with a delicate penetrating metallic tone (Bletschacher).

The theorbo suffered a serious accident in modern times damaging the top and bowl. Repairs were carried out in the 1990s. The top is conserved separately which offers the opportunity to examine the internal features and construction of the top and bowl. On the other hand, it makes it difficult or impossible to accurately record some measurements. 

Two features of the theorbo stand out. It is large. The length of the fretted strings and of the diapasons are approximately 99 cm and 171 cm. The bowl is approximately 71 cm long, 43.5 cm wide and 16.5 cm deep. and it is built entirely in maple, with the exception of the spruce soundboard.

My intent was to collect enough information to enable me to build a good model of the original. This included making tracings of the contour of the bowl, neck and theorbo extension; templates of the cross-sections of the bowl and of the longitudinal axis, as in the photo below; numerous external and internal measurements and numerous photographs. Throughout my examination I recorded  impressions, opinions or reservations of various observable features.

All examination implements were made of plastic, card stock with painters tape or cloth.
Most photographs were taken while the lute was inverted on the examination table.

The museum's photographs and description of the theorbo can be seen on the MIMO portal.

The bowl is constructed with twenty-one ribs of lightly figured maple, mostly sawn well off the quarter. The ribs are separated by dark wood spacers that are less than 1 mm in width. The two edge ribs are significantly wider than the other nineteen which are somewhat uniform in width. The finish on the bowl, where patches of it remain, has some substance. Perhaps it is a varnish applied over a ground. This photo includes the cardboard jig that I use to create a profile of the bowl. More on that later. Since the top is not attached to the bowl the lute lies flat on the examination table. I made a tracing and measured the result; 435 mm wide and 710 mm long to the neck joint. Note - lute bowls are prone to distortion without their soundboard glued in place, therefore my measurements should be considered approximations of the bowl's original  dimensions.

The flattening of the bowl is extreme. Its deepest point is level with the ends of the end clasp and measures 166 mm.

The bowl has been expertly assembled: the joints are tight and the ribs widen uniformly, from the center out, as they disappear under the clasp.

The end clasp is a single piece of maple measuring 740 mm tip to tip. Its maximum height measured at the center is 48 mm. The clasp is visibly asymmetrical in that the treble side (photograph's left) sits lower than the bass side. The discrepancy is not the result of the treble rib having been cut down in an attempt to improve the lute's playing quality because the clasp's finials are the same size. Also, the two edge rib's widths are nearly the same height.

For clarity this photo shows the treble side step in the end clasp. It also demonstrates the rigorous working life this instrument endured. The surface of the bowl is full of dents, scratches, repaired cracks and a few worm holes.


The front of the bowl displays a high level of craftsmanship as the ribs pass smoothly over the front block with joints intact and little evidence of the type of distortion caused by the pull of the strings. The rib widths widen uniformly along the arch of the neck joint and precisely abut the neck.

While the instrument was inverted I made two templates; the cross- section at the widest and deepest point of the bowl and a second template of the longitudinal axis (shown in a previous photo).

I made the template forms from two pieces of heavy card stock glued together to form an over-sized arch. Small rectangular tabs made from  file cards are then placed against each rib of the lute bowl while the other end overlaps the template arch. Each tab is held in place with painter's tape. Once the two edge pieces and the center piece are secured the assembly is self-supporting and the remaining tabs can be easily added . When completed the structure is light weight and accurate.

The inside of the bowl reveals an interesting history. Most of the original printed paper that covers and reinforces the rib joints remains. Also, a sizeable piece of printed paper is glued over the area just below the internal clasp. Numerous wood cleats reinforce several rib joints. The clasp itself is roughly made and is about 10 mm thick in the middle. New rib material (yew wood) has been inserted to repair  a modern day accident - the soundboard was damaged as well.

The upper left corner of the photo reveals a strip of thin wood has been applied over, and covers much of the width and length of the edge rib. The same addition was made to the opposite edge rib.

The  front block, although appearing to have been roughly made, affords a snug fit for the ribs. A large wrought iron nail pierces the block and helps to secure the neck.

The two auxiliary strips that reinforce the edge ribs terminate at the front block as seen here.

What appears to be the end of the fingerboard is visible in this photo. It is stained black to mimic the appearance of ebony. However, it is not a separate  piece of wood glued to the top of the neck but is part of the same neck material, creating a faux fingerboard.

The evidence for this is, in part, the lack of discernible glue lines around the edge and ends of the of the neck. The surface is cambered. I estimate it to be less than 2 mm in the center tapering to a little over 1 mm at the edges.

This photograph of the neck joint area shows several interesting features.

The arrow-shaped notch at the neck joint is obviously a center line indicator and an idiosyncratic example of Buchenberg's methods.

The top edge of the auxiliary strips that were applied to the edge ribs are visible. They are about 1.5 mm thick. By comparison, this this photo reveals how very thin the two edge ribs are.

I was very disappointed that my magnetic thickness gauge failed to work and I was unable to measure the thickness of the lute's ribs.

The neck is 410 mm long, 101 mm wide and 27 mm thick at the joint with the bowl. It is 81 mm wide at the nut and 20 mm thick at the point of transition to the extension.

The extension is constructed in solid maple 807 mm long, 848 mm if the length of the over-lapping glue joint with the fretted neck is included. It is 73 mm wide at the point it emerges from the front of the neck.  The sides are angled in slightly so that the width of the underside at this point is 68.5 mm. The thickness of the extension at the first peg is 25.6 mm. The rear of the extension is cambered less than 3 mm diminishing to about half that under the theorbo head.

It is interesting to see the sloping contour of the extension as it blends into the neck . This is a style that is always associated with theorbos built with solid wood extensions. This is the earliest example that I have seen.

I can think of no rationale for the two depressions on each side of the extension.

There is a darkened incised line that borders the top and bottom edges of the extension.

The string trough is roughly made with obvious chisel marks, although the recess for the treble strings has a nice profile. The trough is 15.5 cm long, 14 mm wide at the narrow end, 19 mm deep and 11 mm wide at the bottom of the trough.

A darkened incised line borders both edges of the top of the extension.

The peg heads are generally 20 mm wide with a maximum shank diameter of 8 mm.

The string nut is made of a light color wood stained black. It is 43 mm wide. The six string grooves span 35.6 mm. There is a 2 mm camber to the top of the faux fingerboard.

Photographed face down.

The theorbo head, shown inverted, is roughly made. It is 53 mm high, 106 mm long and 71 mm wide at the nut support. The nut is 33 mm high and 85 mm long. The darkened incised lines that are found on the extension continue onto the theorbo head, somewhat clumsily.

Photographed face down.

This view (also inverted) shows the angle of the sides of the head. The width of the head at the widest part of the crook (see the arrow indicator in the previous photo) is 33 mm at the top and 19 mm at the bottom.

The damaged areas of the soundboard have been stablized. It is constructed from straight close grain spruce. The discoloration, cracks, scratches and other defects of age obscure the fact that this a fine soundboard that probably possessed superior acoustical properties.

The ornate triple rose is beautifully cut and is well preserved.


 It deserves a close-up view.


The bridge is made of a light hardwood, probably maple, stained black and accurately drilled for fourteen single strings that span 132 mm. The rear face is indented to help hold the string ends while the front face is under-cut above the string holes creating a slight over-hang that helps to settle the string loops.  It shows signs of being quickly made; there are tooth-plane marks on the top surface and an incomplete chisel cut on the treble side wing. It is still a fine example of the standard design of the era combining functionality with cool aesthetics.

Some may find the treble wing proportionally too long and open, but I am convinced this is intentional. Whereas the compact design of the bass wing condenses the mass, the length and openest of the treble wing spreads less mass over a wider area. This mirrors the dimensions of the bridge which progressively concentrates the mass from treble to bass.

The bridge is 227 mm long. The top of the string block is 155 mm long on the front and 158 mm on the rear edge. The top slopes down from the rear so that the treble side at the first string hole is 6.9 mm high while the rear is 8.15 mm. At the fourteenth string hole the heights are 8.8 and 9.6 mm. The gluing surface under the first string is 11 mm front to back while under the fourteenth string it is 13.3. The string holes are generally drilled 3.5 mm above the soundboard.

An excellent technical drawing of the top made by Stephen Gottlieb is available from the MIM museum shop . The drawing includes all bar positions and their dimensions and measurements of the soundboard thickness at numerous locations.

I bought a copy of the drawing while I was in the museum and since I had it available during my study for consultation and future use I did not examine the interior of the soundboard in detail.

All the surviving bars appear to be original. Unfortunately, the two treble bars and the J-bar are missing although their original positions are indicated by glue marks. There is also a scribed circle whose top tangent indicates the position of the first bar. The opposite tangent usually indicates the position of the missing J-bar, while the focal point indicates the position of the bridge.

In closing I would like to thank Anne-Emmanuelle Ceulemans, curator of stringed instruments, who kindly arranged my visit and to Joris De Valck, restorer and my host in the museum's laboratory.

All photos by the author