Arias Restoration II

The back of this double back guitar had been removed and replaced in order to deal with some cracks which had developed over time. It was done many years ago and by someone who did not take the care that this instrument deserves. The first indication of this is a black veneer line which appears and disappears as it runs next to the binding.


The back was removed it seems by sawing through the purfling and sides on the black line. Although it must be extremely difficult to keep the saw within the limits of the black purfling, it would have been easy enough to remove all of the black and replace it once the top was glued back on. The other mistake made back then was a failure to register the curve of the back before taking it off. This can be done by checking the relationship of the back to the sides before taking the back off. Actually I compare the lift or angle which a straightedge placed on the edge forms with the plane of the back (as if the back were flat). Even a back (or top) which is sunken will conserve this relationship out at the edge where the reinforcements don’t sink. In this case the sides were planed flat to receive the flat, sunken back when it was replaced. I say all this not to criticize the repairperson but rather to explain later how I restored these elements.

IMG_6094Firstly I removed the rosewood pins which had been used to locate the back in the previous operation by drilling them out. Knowing that I would be replacing the black veneer I choose to remove the binding from the guitar and take the back off by detaching it from the reinforcements instead of sawing through the way it had been done before. As soon as I did that I realized that sawing through had not been such a bad idea. IMG_6096When I got the back off I saw that Arias had used triangular spruce blocks instead of reinforcements. The thin edges of these blocks are quite delicate and so sawing through them might have been better than separating the joint on each one. It took a lot of time but finally the back was off and only a few of the blocks had to be replaced. The outer back had three cracks and the seam was separated causing ugly splits in the white centre purfling. I chose to replace the centre strip with a slightly wider one which also served to widen the back enough in order to allow a curve and to keep it from spittling in the future. The back of a guitar tends to shrink over time and this will cause either splits or for the back to end up too small for the outline formed by the sides and the damage is visible where they meet. IMG_6103With the back off I could see the repairs that had been made and the possibilities that I had for restoration to the original state. The priorities were to secure any cracks, recuperate the width of and to restore the curve of the back. The back bars had to be removed because they were concave over their length and were holding the back into a negative curve. In removing the longest one I saw that cyanoacrylate glue had been wicked in through a hole in the centrestrip to deal with a loose spot between back and brace. That stuff is hard to remove! In order to reglue the two sides of the back to get clean purfling in the middle it was necessary to remove the cleats as well. IMG_6118Once the two halves of the back were free I prepared a new centrestrip and jointed the back with the help of some old tricks to make sure that there would be no unevenness between the different elements. All efforts must be made in a restoration to avoid having to thin any further to even out irregularities and also having to re-varnish. The marks of Arias’ toothing plane were evident through the inside of the guitar so I didn’t feel too bad about using mine to ensure flatness of the surface before glueing on the cleats and the bars. This instrument seems to have been made with a very clear idea to make it as light as possible so I respected that idea and made the cleats the same size and shape and used cloth tape and hide glue to secure the cracks.IMG_6130

The bars could now be shaped to the curve I wanted and glued to the back. I had to guess at the curve but I have seen many instruments from this period and I chose to use as slight a curve as possible which is in keeping with what it might have had after 116 years of shrinking. Just two big jobs left to report on now: securing the interior back and preparing the sides for the outer back and glueing it on.IMG_6132


arias-3My position on restorations is that instruments should be conserved in the original state as much as possible. It puts me at odds with many musicians who feel (justifiably) that a guitar should be returned to playability above all else. We have learned so much from antique instruments and we could learn so much more, but with every modification we lose information about the original instrument and its maker. Documenting a restoration is a way to perserve some of this information but trusting the eyes and opinions of someone who worked on the guitar many years ago is often not enough. This problem is exacerbated by the fact that some restorations or repairs are undertaken by untrained individuals or even in the case of ethical professionals the criteria for restoration has changed over time and may continue to change in the future. What I might do today with the best intentions might in the future be deemed woefully misguided.img_6131

With all that in mind, I will tell you a bit about a recent restoration that I performed more in a way that made the instrument a better tool for the musician as opposed to prioritizing the preservation of the instrument in its original state. The first time I came across this guitar in 2010 it was suggested that I take the back off in order to properly restore it. I chose not to do that but rather to do only what was absolutely necessary to preserve the instrument. Since it was already playable that was good enough for Javier Riba (the owner) at the time. I re-fretted, secured some small cracks on either side of the fretboard, removed a tap plate and re-clamped the bridge which was lifting (although I did not remove it completely beforehand).The guitar has been used in many concerts since then and on an excellent recording for Tritó. Aljibe de Madera

The problems I chose not to address at that time included some cracks in the back which I was sure had been repaired, some back seam separation img_6084and a general fragile nature to the guitar (worsened by the fact that the curve on the outer back had been destroyed by the back removal and the two backs sometimes touched). When the back was removed by someone else many years ago, some of the purfling had been destroyed as well. Now the bridge was once again lifting and Javier was feeling nervous every time he played the guitar. It was time to take the bridge off, remove the back and take care of all the problems there might be inside.

First I removed the bridge  very carefully so as not to damage the top. The combination of a few factors made this job easier than I had thought it would be. The bridge is ebony which is notoriously difficult to glue well. The bridge was already lifting and I believe that this was not the first time that the bridge had been taken off. I say this because the top had some old glue-filled “divots” under the bridge. For a great joint you need both surfaces to be very even. img_1048Once the bridge was off I was able to see just how deformed it was and how unrealistic it was of me to hope for a lasting repair just by adding glue and clamping as I did back in 2010.  So, how to obtain two smooth and correctly shaped surfaces which I could then glue together? If I were to work the top smooth and shape the bridge to it as we do in construction, the bridge would end up lower causing incorrect action as well as losing some of its edges as deformation was quite considerable. Sanding the top is always unadvisable with a historically important guitar and in the case of a top with a thickness of 1.7mm, completely out of the question.img_6077 I was unwilling to glue ebony to ebony in order to make the bridge slightly higher so I used european beech to laminate a layer on to the bottom of the bridge which allowed me to work the bottom of the bridge into a curve which was mated to the top. The minimal edges of beech which were left showing were coloured black. In the photo below you can see the beech layer before trimming, shaping and dyeing.img_1053

The top where the bridge was to be glued was not smooth enough to allow for a good glue joint so I used a small router to deepen the outline of the bridge just enough to get a surface which was in its majority smooth and free of irregularities. That turned out to be about 4 tenths of a milimetre.  The flexibility of the top was so great that I didn’t think I could glue the bridge using only outside pressure the way we do here so I used bridge clamps. From what I could tell from the “grab” of the glue, the squeeze out and the finished product I think this joint will last the rest of the guitar’s life. I took pictures of the restoration of the rest of the instrument as well so I hope to share those in future posts.img_1063




A friend just finished a stint working for a local ice cream parlour and was telling us the other night about how they work.  Just a few examples:  The almonds they use are hand-sorted to ensure quality.  The local strawberries are sourced from a single farmer.  (Most of the strawberries in Spain are tasteless monsters often produced in greenhouses near Portugal.)  This year they didn’t make their very popular blackberry ice cream because they couldn’t get enough of the fresh berries; frozen is not good enough for them.  I myself have asked for orange sherbet and found on one occasion that it was not orange season so they didn’t have it and on another it was the beginning of the season and the oranges were still a bit tart.  When I tasted it I was surprised to that this characteristic showed up in the delicious final product.  This friend was very impressed with other aspects of how the business was run.

The spanish word “artesania” is much used and much abused but in its origin it referred to the traditional workshops where a trained professional produced functional articles that by their nature were unique and creatively made.  These workshops were often associated and regulated through guilds.  Ceramics, carvings, ornate furniture, leather goods, wrought iron, goldsmiths, blown glass, marquetry and instrument-making are just a few examples.  These artisans persisted because there was a need for the product, the quality was very high and of course cheap industrial processes did not exist.  In different historic periods there have been resurgences of artisanry for reasons of increased wealth and patronage by the nobility.  Those examples that have survived into today’s industrial world have done so usually because of the quality of the work and the demanding clientele who are not satisfied with what industrial production has to offer.  In order to make a living at such a time-intensive occupation the artisan must be well-organized, efficient and must be able to charge much higher prices for his products than the same product produced in an industrial setting.

My favourite example is, of course, guitar-making and one of the oldest examples of this tradition can be found right here in Granada.  Fine guitars are often referred to as “handmade” in part as a translation of the spanish term “artesanal” but the definition of the spanish term is not about the absence of machines but rather the absence of industrial processes.  Artisans of old used the best technology they could and often showed great ingenuity in developing new ways to lend more efficiency to the work without detriment to quality.  I dislike the term handmade because it gives the impression that we use no machinery in our work.  We too use the best technology we can but we control every step and integrate it into the whole process.  In most cases the guitar-makers here work alone unless they have an apprentice.  It is this hands-on, total responsibility for the process that makes these guitars so much better than a factory product.  As I have said before, the factories will never reach the quality that we can for the simple fact that profits will always be the driving force.  They can theoretically train someone to make a guitar as good as ours but when they see the time it takes that one person to produce one guitar while their production lines do nothing (for that guitar) it doesn’t make economic sense.  So they will take advantage of the production lines for the less important tasks until they realize that in guitar-making every task is of the utmost importance.  Now back to the ice cream analogy:  the quality of the ice cream I was talking about earlier is fantastic, their dedication to quality, sustainability and eco-friendliness pays off and everyone wants their ice cream (much more expensive than the competition of course).  In my mind it is this constant striving for quality that makes the handmade guitar so much better.  So, the definition of handmade must be qualified to refer to the process of one maker choosing the wood, seasoning it, using his tools (electric and otherwise) to thickness the components, design and bring to fruition the aesthetics of the instrument, and most importantly transform the pieces of wood into something magical which has the potential to stir our emotions like nothing else.

Here is a video made with a guitar which inspired me to make a copy of it this year.

Deuxième Arabesque de Claude DEBUSSY (arrangement: J. Riba)

“In the spring of 1913, Andrés Segovia decided to seek his fortune in Madrid, and gave a recital in the Ateneo the evening of May 6 […]  The programme did not seem so different from those of the followers of Tárrega but it contained a surprising novelty:  The transcriptionby Segovia himselfof the second of the Deux Arabesques for piano by Claude Debussy […]  This audacious move revealed very clearly the aspiracions of the young Segovia to place himself in a different category than the rest of the guitarists looked down on by musicians and practically ignored by learned audiences—and his intention to play a leading role in the artistic circles of his time.”

Angelo Gilardino,
extract from the liner notes to the CD “Aljibe de Madera,
Homenaje a Andrés Segovia” Tritó TD0094

Up close with the 1900 Arias

This is some information from my examination of the 1900 Vicente Arias.  The photos of this guitar can be found here and here.  Other articles have details of the building of the first replica.  The guitar has a label which tells us that Vicente Arias made it in 1900 shortly after moving to Madrid.  It is a relatively simple guitar with good quality wood and excellent craftsmanship but without the obsession with perfection that seems so desirable these days.  There is a double back probably 1 to 1.5mm thick with very light bars across it.  According to those who have studied other guitars of his he did this to get the sound of a very shallow instrument while giving the physical sensation of playing a full-sized instrument.  This explanation doesn’t fit with the two sound-port sized holes in the inside back just below the bridge. Since the outer back has been taken off it is very possible that these holes were made by someone else.  It should be noted that the back was taken off and replaced by an amateur or someone with no regard for the original work.  The back curve was ruined and no attempt was made to replace the filleting which was destroyed in the cutting. The headstock has also suffered some modification as can be seen in the photos and I think the original holes were plugged and redrilled at some point.  The added pieces of cedar on the head might make one think that this head once had tuning machines but I can find no evidence of that.  The top is extremely flexible, more so than anything I have ever seen.  However, the deformation in front of and behind the bridge is within acceptable limits (the curve in front remains positive) and there are no cracks on the top except for the typical ones along the fretboard.  There are three cracks on the outside back, one under the central reinforcement and the other two very likely cleated (perhaps the reason for the back removal?) as they show no movement when stressed.  There are some marks of slight burning on the underside of the top perhaps where the top was reheated after glueing the bridge on.  The peones are glued with no spaces in between them.

The following are some measurements in mm:

Depth: endblock   93
neck  92
upperbout 91
waist 82
lower bout 92
These would originally have been different but hard to tell by how much after the back removal.

bridge  25 wide and 28.5 with the lip
tie block is 8mm high off the soundboard and the front of the bridge is 6 high
wings are 3 high

soundhole diameter is 86
rosette width is 24.5

three bars on the top
the closest one to the neck is 3 high X6
next one is 15.5 X 6
last one “below the soundhole” is 14 X 6
not scallopped

the bars on the double back are 4 X 5.5

There are 7 fans 3 X 4.5 at the highest point, seemingly planed straight while sitting in the solera so are not uniform height.
they have a rounded profile.  The projection of these fans meet very approximately at the 10th fret.  I prefer to measure the separation of the fans at the third top bar and then again at the back edge of the bridge.  the first measurement is 30mm and the second is 45mm.  These round figures indicate that he used these measurements too as opposed to using a converging .  Please note on the pictures of the illuminated top the short finger braces, they are shaped just like the fans.

The space between the top and the inside back as measured at the “lower” edge of the soundhole is 79.

the length is 479, upper bout is 264, waist 228 and lower bout 365.

I finally got my hands on a Hacklinger gauge at the same time as I had the Arias in the workshop.  I checked it against my caliper on a few pieces of wood and found it was reading 1.6 for 1.7 so I would add 0.1 to the following measurements.   At first blush it seems that Mr. Arias settled on 1.6 mm as an appropriate thickness for top, back, double back and sides.  The sides in places go down to 1.4 but then anyone who has ever worked with Brazillian can understand that.  The back is very close to the double back so I was unable to get the Hacklinger to measure anything but the stretch between the second and third bars on the back and only that thanks to the “soundholes” in the double back.  The back showed 1.6 and 1.5 in the central area and 1.3 out towards the sides with lots of variation in the thickness which makes it hard to say with any certainty that this was his intention.  Who knows what was done to this back when it was removed.  The top shows one very thin area (1.2) where a tap plate was present and you can even feel a depression in that area, I think it is obvious that someone ripped it off, damaged the top and sanded it down and then someone covered it up with another tap plate. There is also a thick area behind the bridge thickest(2.0) near the end block behind the A string.  This area thins out gradually in all directions to the 1.6 thickness of the top within a distance of 4 or 5 centimetres.  I have no explanation for this although it does make for a thicker are in general behind the bridge.  One might think that the rest of the top was sanded down later in the guitar’s life and this part left thick but there are no signs of that, no sharp rising or increased thickness around the bridge and fretboard.  The rest of the top shows 1.4 to 1.6 very consistently, a bit heavier on the 1.6 readings (and even two 1.7 readings) between bridge and soundhole.  As you approach the edges of the guitar there is no decrease in thickness, actually in the lower bout there is even a tendency of maximum readings around the edges whereas as you move inwards you start to get about 0.1mm less. The top is extremely flexible and I get a resonance at just below F#.  I couldn’t get a very accurate measurement of the thickness of the double back but by making a saw cut in a small block of wood that just allowed the “soundport” edge to enter convinced me that thickness is 1.6 as well.  The holes in the double back are symmetrical and 32 mm in diameter with a separation of their centres of 96mm.  The double back has these tiny braces and then seems to have a second set on the “bottom” side as well, I can only just touch one with a finger.  The back brace (at least number 2) is about 15mm closer to the soundhole than the one we see on the double back.  The braces on the double back are scalloped but I think the ones on the back are not because I couldn’t get the magnet from the Hacklinger to pass through anywhere. As I think I mentioned before, whoever took the back off destroyed the back curve so assuming that the double back curve is still there then the distance between the two backs was originally greater than it is now (10.5 mm at the holes)  At one point I got a vibration by tapping on the back and was very worried about a crack until I realized that the bars of one of the backs is almost touching the other one and vibrated sometimes.  I can with very little pressure make them touch.  That gives us a pretty good idea how high the back bars are.  One of the things I try to do when I study an important instrument is get an idea of what the original archings were.  Usually an arch will flatten out under tension after 100 years but the edges of the top (or back) will still have their original angle upwards with respect to the plane of the top.  If you place a straight edge along the tangent at the edge you can read the distance the straight edge projects above the opposite edge. This allows us to reproduce a curve for the solera which should allow for the same reading on the finished copy.  The idea is that the arch will be the same as the original was in its beginings.  I placed my straight edge at the end block just off-centre so as to be able to take a reading at the top edge-23mm. Then at the fretboard and read at the endblock-7mm.  Transverse readings just in front of the bridge-13.5mm and on the brace below the soundhole-6mm and mid soundhole-7.5mm.  The back had no angles so I have to assume that the edges of the sides and linings were altered when the back was glued on again.  Of course it is possible that the original back had no arching at the edges but I sincerely doubt it.

First Arias

I finally finished the Vicente Arias copy!  Since I wasn’t making this guitar for anyone in particular it kept getting pushed back by the guitars people had ordered.  Here you have a video and some photos of it “in the white”  The sound changes once it is varnished but it is a very slight and predictable change and the builders among you will hear the lack of varnish.  Javier Riba, who will be opening the Cordoba Guitar Festival this year, came by and played a little Sor for us, study no. 12 op. 6.  If you remember, he owns the original and let me do an extensive examination of it.  When you make a copy you are usually trying for both an aesthetic and an acoustic match but the aesthetics are usually easier than the acoustics.  In this case I was very pleased to hear Javier say that the feel of the guitar is very similar to the original and the sweetness of the tone is there too.  The video was made with an audio recorder which also does video but it seems that for a really good audio take it needs to be quite close.  Thanks to Toni Valls and Carlos Juan Busquiel for the advice on finding a recorder and on making the recording itself.



Closing the guitar

In examining a guitar with a double back; in preparing the pieces and planning out the construction of a copy you obviously end up thinking a lot about what effect it will have on the sound of the guitar once it has strings on it.  Volume, projection, the character of the the different notes, sustain; any one of these might be affected by the physical changes in the instrument.  Well, I will try to get a number of different opinions on this guitar once it is finished but for now the biggest difference is the rigidity of the structure.  It makes sense once you think about it but I never had.  To have the sides connected around the perimeter of three plates instead of two is a big change.  I didn’t flex things too much but with just the internal back glued in with its reinforcements above and below it felt and sounded like a drum.

      I usually make more than one guitar at a time and I am used to seeing them hanging there but this one sure looks out of place.

It took me ages to decide how to glue in the internal back; I never could get a look inside to see how Arias had done it.  The fit has to be very good but I had to have room to get it into the body and imobilize it at the neckblock.  I had cut shallow slots in the neck block beforehand to accept a cutout of the back but now it was very tricky getting the back into the body and at the same time Into the slots.  Next time I think I will make the cutout the same size as the block and then glue supports above and below it.  The other thing I will do differently is the endblock.  I cut the endblock short to fit only between top and internal back  and there is a small deformation where it ends.  This photo shows the back being glued on and the last one is after the string has been taken off. 


So here is the moment of truth.  Bracing patterns are not really so important, the sound of a guitar depends so much more on things like:

-the tension of the top when the guitar is finished
-the humidity at the time of gluing
-the sequence of steps and the procedure followed

However, since I really don’t know how Arias got the results he did (a fantastic instrument with a sound and architecture that are very much alive over 100 years later) I am trying to replicate exactly what I see in the original guitar.  Many of the antiques I have played have been dead and/or the soundboard badly bellied but not this one.  For example, I can’t imagine what good the mini-braces in the foto below will do but I am using them.  Please note that they show every sign of having been glued on during construction of the guitar.  They are not crack repairs, they have the same shape as the fan braces and they are scalloped after being glued on (I can’t imagine anyone scalloping them through the soundhole).