Restoration of Egyptian Oud - part 3, Soundboard

The soundboard is a real mess but I shall attempt to undertake restoration/ repairs making as little change to the original as possible. Part 1 of this series includes images of some of the problems and damage to be corrected.
Repair work will include:
- removal, repair, reinstallation of bracing.
- removal, repair, reinstallation of rosettes.
- re-gluing of splits and separated joints.
- patching of broken areas
- rebuilding of the edge tiles.

Before starting, I made a complete digital record of the damaged edges of the soundboard for future reference. The attached images give an indication of the extent of repairs required around the soundboard edge. On this oud the tiles were originally rebated into the edge for about half the depth of the soundboard or about 3/4 of a millimeter - similar to edge banding on a lute. Another (unusual?) feature is that the tiles do not extend around the entire periphery of the soundboard but stop short at the bottom edge of the belly, This part was originally covered by a strip of leathercloth glued in place.
The soundboard has been subject to repairs and removal in the past so little now remains of the foundation of the original tile installation. Tile replacement will, therefore, have to be for the full soundboard thickness. The tiles, themselves have also been subject to repairs and replacement in the past but enough of them remain to indicate the original tile design.

One consequence of leaving the oud unrepaired for many years (mea culpa!) is that there is consequential damage.
For example, in order to keep everything together in storage, I made the mistake of using 'scotch tape' to 'temporarily' keep the soundboard and bowl together. The result is that glue from the tape has - over time - been absorbed into the wood and has stained the surface of the soundboard in places. To remove the staining and 'lift' the glue residues from the wood I used a wax solvent, of the household "Super Goop" variety, applied locally to the affected area with a Q-tip rubbed in the grain direction of the soundboard and then rubbed dry with a paper towel. This procedure was repeated several times until most of the stain was removed. The absorbed glue residues were then removed by first coating the area with the solvent, covering with a piece of brown (absorbant) paper and then applying a hot iron to the area - the softened residues being absorbed by the paper ( an old trick for removing wax stains). The affected area was then rubbed again with a paper towel moistened in solvent to remove the last traces staining.
This process revealed the bare wood of the soundboard where the glue residue had been removed indicating that the soundboard had been originally finished in some way to prevent dirt absorption - most likely using a light wax finish(?)

As the belly is now in a very fragile state it has been temporarily fixed with masking tape to a flat working surface of particle board to protect it against any damage during handling. Cut outs for the bridge and plectrum guard have been made in the working surface so that the belly can lie flat and fully supported while being worked upon. Splits in the belly have also been temporarily taped to prevent further damage.
As the original braces are to be preserved, they have been numbered and reference marked so that they may be replaced in their original positions after repair.
To preserve the oud maker's signature on the rosettes, they have been coated with a clear oil varnish. Without this protection there is a danger that the signatures might be 'smeared' or otherwise damaged by moisture during the process of un-gluing the braces and rosettes for repair.

The bars were removed by brushing warm water along the glue joint on both sides of each bar, applying heat locally with a hair dryer and, at the same time, carefully working a fine artists spatula under and along the glue joint after the glue had softened sufficiently. This had to be done slowly and with great care to avoid damaging the bars or sound board.
The large quantity of old glue used and left behind by previous luthiers who had undertaken repairs made bar removal a lot more difficult. Full strength carpenters' hide glue had been used - no thought of any facilitating future restoration work there.!Lots of clean up of old glue remains to be done.
Many of the bar ends are split or otherwise mangled so they will require glue repairs or patching.
The same strong glue has been used on the rosettes so I have decided to attempt their removal another day,

I have decided not to try to remove the shams for repair as they are too strongly glued in place and there is too great a risk of doing damage to them and the soundboard. I shall, therefore, carry out repairs with the shams still glued in place.
In the meantime, does anyone recognise the design which seems to me to be a bit unusual - although there are some 'Arabic' elements? The workmanship is quite rough.

The attached image is a sketch drawn to scale of the sound board viewed from underside to show the bracing layout.
The maximum height and width of each brace is shown measured in mm. The cross section of each brace is also shown tapering from maximum width to about 2 mm at the top of each brace (a truncated triangle). The braces are cut away in the usual fashion reducing the ends to a depth of about 8 mm where each brace meets the bowl - the cut away starts about 7 cm in from the end of each brace. All work rather roughly executed.

The wood of the braces is quite light (low in density) - similar to cedar wood. The grain of the braces in cross section runs roughly parallel to the face of the sound board i.e across the brace - as is usually found in old surviving lutes (not vertical as in modern guitars).

After having made progress in clearing some workspace today I have decided to pick up the thread again on this project (the work needs to be done sooner or later).

The next step is to clean all residual glue from the braces and sound board - starting with the braces. Once the glue has been cleaned off it will be possible to assess the extent of damage to the braces and the repair work required. The intent is to use the original braces.

The glue residues are hard and brittle and have been applied in excess. Chipping away the glue with a knife or chisel is not an option due to the risk of tearing and damaging the relatively soft but brittle wood of the braces.

The method for glue removal is to first brush water onto the heavy glue deposits, apply a wet thin cotton cloth over the glue and then apply a hot iron to the cloth. The hot iron (at about 450°C) instantly generates steam that quickly softens the glue. The softened glue is absorbed into the cloth, the cloth then sticks to the iron allowing the soft glue to be wiped off the brace with a sliding action of the iron. This procedure is repeated bit by bit using fresh cloth as it becomes contaminated with the absorbed glue.
This is a slow but quite efficient process.

The cleaned braces will be left to dry for several days.

The braces have been numbered in pencil (Arabic numerals from 1 to 6 - not counting the half braces on either side of the large sound hole). Every effort will be made to preserve these original markings. Brace #1 is below the bridge and #6 is at the neck block.

Note that the cross grain of the braces is not perfectly parallel to the sound board surfaces but varies between about 30° to about 45° from horizontal.

With the braces cleaned up some of the typical damage requiring repair can be seen.

Brace splits occur when the sound board is carelessly removed - a combination of driving a hot spatula into the end of the brace and attempting to pry the sound board loose before adequately softening the glue at the end of the brace with moisture and heat. The damage is easily repaired however.

Not sure how the ends of two of the braces became so 'chewed up' over time but this will be fixed with a small 'splice' repair.

With the braces cleaned up, next step is to clean up the underside of the sound board to remove all surplus glue. This was achieved using the same damp cloth/ hot iron procedure used for the brace clean up as previously mentioned.
The attached image shows the area under the fingerboard inlay before and after clean up with glue and the original paper reinforcement removed
Removal of the glue is made more difficult because the whole underside of the sound board had been originally scored with a toothing plane. Not sure why the sound board underside was finished in this way - possibly just to ensure a uniform flatness but without then going to the extra trouble of planing out the grooves as would be the usual practice in sound board thicknessing.

More to follow.

The shams were covered in excess glue and had been 'repaired' in the past with 'globs' of glue spread over the splits and cracks.
Due to the fragility of the sound board the shams are to be repaired in situ.

Excess glue was again easily removed with the damp cloth/hot iron technique. The glue 'globs' were picked out with a dentists pick after the glue had been softened. The open cracks will eventually be filled with matching wood.

One problem with use of water and heat for removing braces and excess glue is that the accumulated dirt of ages - in grained in the sound board wood - mixed with the water can cause unsightly 'watermark' staining on the sound board outer surface.
To remove these stains (and a great deal of accumulated dirt) the offending areas are rubbed with a moistened cloth dipped in Pumice powder. Pumice powder is a very fine abrasive (volcanic ash) used for final polishing of varnished finishes. The fine powder picks up the dirt producing debris similar to that produced by a pencil eraser. Eventually the whole outer surface of the sound board will be cleaned up in this manner.

The damp cloth/ hot iron technique requires a plentiful supply of rags that are discarded when contaminated with the old glue.
We are, of course, removing old hot hide glue in this project.

The surface of the hot iron also picks up a lot of the glue so must be cleaned at intervals. This is easily done - after the iron has cooled and disconnected from the power supply (!) - with hot soapy water and nylon bristle scrubbing brush.
This work could just as easily be carried out without a fancy, thermostatically controlled, electric iron - using just a plain iron heated over a flame but - of course - an electric iron is more convenient and a bit safer.

The average sound board thickness is probably less than that and the edge - due to repeated removal of the sound board for past repairs - is thinner still. The 2mm edging is just to give me room to play with before final thicknessing.

I cannot recall if I have previously surveyed and recorded the sound board thickness so will do this for the record and post the results later for information.
Note that the measured thickness (about 1.5 mm if I recall correctly) will be somewhat greater than the minimum average due to the toothed surface - a caliper measurement only recording the 'peaks'.
Did the maker use a toothing plane to produce a more flexible sound board without significantly reducing overall sound board mass or was he just taking a short cut to save construction time? Another possible reason for a toothed surface is that it provides a greater surface area for gluing the braces.

For those unfamiliar with a toothing plane, attached is an image of my antique toothing plane. Note that the blade is set at a very high angle and is scored on the back face with vertical grooves to produce 'teeth' along the cutting edge. In use the teeth score the surface of the wood into fine grooves. These planes were once used to work wood with difficult grain before final finishing using a regular plane or cabinet scraper (to remove the grooves). They were also used to prepare a wood surface prior to veneering the plane levelling the surface and grooving it to provide a greater surface area for gluing. Some luthiers use the toothing plane as part of the sound board levelling process.
Toothed plane blades can be purchased from Lee Valley Ottawa designed to fit their regular planes but at one time the old blades came in three grades - Fine, Medium and Coarse.

In this case the oud maker failed to remove all of the original saw marks from the underside of the sound board - again a sign of rough workmanship - out of sight out of mind!

Here is the sound board thickness plot giving the nominal thickness in millimeters - measured with sound board calipers.

Due to the age of the sound board it is possible that the thicknesses were even greater than measured here.

These thicknesses are of the same order as those found on surviving lutes of the 16th/17th C.

Again...

Thanks Yaron! There is some way to go yet before it can be confirmed that the old sound board is viable for restoration to playing condition as planned.
At present the sound board - after moistening the underside uniformly by covering with a slightly damp cloth - has been clamped to a flat surface where it will remain for about a week to dry out slowly. This procedure is necessary in order to prevent the unbraced sound board curling due to shrinkage of any glue residues that remain after cleanup. If this is successful then it will be possible to check the fit of the sound board to the bowl that hopefully has not distorted significantly over the years.

In the meantime work can continue on the bowl which has its own set of issues.

The maker"s label glued inside the bowl shows that the oud was made by apprentice luthier Hosain Muhi AlDeen Helmy according to Arabic speaking members of the forum.I posted a low resolution image of the label some years ago on the forum but here is a much clearer image.

I cannot read Arabic but was wondering about the meaning of the writing in the extreme bottom left hand corner of the label? Is it a number or a date or the maker's initials perhaps?

An update on the full translation of the label - given the clearer image - would be welcome.
Thanks

The damaged areas and missing parts of the sound board will be repaired with Eastern White Cedar - a wood of similar density and hardness to the original which may also be a type of Cedar (Cedar of Lebanon perhaps - but hard to know for sure).

I have an abundant supply of Eastern White Cedar - grown on my own property. It has a pleasant aromatic smell and is resistant to rot and insect attack. Several pieces have been cut on a bandsaw to provide more than enough selection of material to allow close matching of wood grain with the original. The colour of E.W. Cedar varies from creamy white to pale brown - but will darken over time to match the original wood.

One of the half braces (located on either side of the sound hole) is missing so a blank for a replacement has been cut to closely match the longitudinal and cross section grain direction of the surviving original. The brace will be finally cut to shape after it has been glued to the sound board.
Note the notch that fits over the rosette - confirmation that the braces were originally fitted after the rosette had been glued in place and before the sound board had been glued to the bowl.

The original braces with split ends have been repaired by running hot hide glue into the splits with a thin artists spatula. Inexpensive small spring clamps are used to hold the splits together until the glue has cured.

The sound board is in a fragile state so must first be 'stabilised' to facilitate safe handling by gluing over the many small cracks with paper strips. This repair work is to be done progressively - crack by crack - clamping each repair to maintain sound board flatness as the hot hide glue cures (and shrinks).

The paper used to reinforce the cracks is a hand made 'laid' paper manufactured in 1803 according to the watermark.

Starting with the largest of the repairs - the area covering the fingerboard inlay, which in this case also includes two cracks starting at the sound board edge. This patch replaces the original paper reinforcement under the fingerboard extension inlay.
After application of the hot hide glue and correct placement of the paper patch the glue was reheated with a hot iron after covering the patch with a piece of heat proof 'oven paper' (readily available from grocery stores). The re-liquified glue was then pressed into the cracks and crevices with finger pressure and the repaired area clamped flat and allowed to dry and cure for a couple of days.

The badly damaged ends of some of the braces must be repaired by cutting away the damage and splicing in new wood.
After cutting away the damage with a razor saw, the splice joint surfaces are cut by pulling the brace over an inverted plane held in a vice. Due to the small surface area of the splice joint this procedure ensures joint accuracy and perfect fit. The repair wood is similarly treated so that the grain patterns of the wood are in alignment.

The pieces of repair wood are glued in place with hot hide glue of high Bloom strength - strong glue that starts to gel within a few seconds. After application of the glue, the joint is held together under finger pressure until the glue gels after which time the spring clamps are applied. This way the joint surfaces will not slide apart under clamping pressure.

The repair wood has been cut well oversize so the excess wood will be planed off after 24 hours when the glue has fully cured.

With the glue dry the spliced wood patches at the ends of the braces have been planed to size. Excess material was first removed with a chisel and the patches were then planed flat on an inverted jointer plane moving the work over the plane blade for precise cuts.
The repaired ends have been left with a little extra material to allow a precise joint when refitting the sound board. Small chips of wood missing from the sides of the brace ends will not be repaired.

Before the braces were removed from the sound board reference marks were made on each brace and sound board surface. These marks and the original brace location marks made in pencil by luthier Hosain Muhi AlDeen Helmy will enable the braces to be re-glued in their original position on the sound board.

Quote: Originally posted by jdowning

I .. was wondering about the meaning of the writing in the extreme bottom left hand corner of the label? Is it a number or a date or the maker's initials perhaps ?

I came across this thread again. So here's the full text of the Arabic label:

The School of Decorative Professions
Work of
Hosein Mohyeddine Helmy
Department of
Musical Instruments Carpentry
Third Year
And in the left corner it says Mabrouk ! :
Congratulations

Seems it was the final project of a 3rd year student !
Worth restoration per se ? Maybe for the fun and learning some new techniques.

Yours indeed
Alfaraby