January - February 2006
Click Here to go Back
Newsletter #9: Restoring the 1809 Broadwood; Restoring a lute; Recording the vihuela
The reader will recall the previous newsletter’s account of the early history of the Broadwood firm, established in the last decades of the eighteenth century. Cabinetmaker John Broadwood went to London, apprenticed with the harpsichord maker Burkat Tschudi, married his daughter (in time honoured tradition), and carried on making fine harpsichords. Eventually he dropped the harpsichord, around 1793, to concentrate on making “cottage” pianos, the early squares so beloved of the emerging merchant class of London. He established a firm that would make some of the finest keyboard instruments of the first half of the 19th century.
In an inspired marketing move, he gave the middle-aged Ludwig van Beethoven a fine grand at the request of several important German-speaking composer-pianists working in London, shipping it to Vienna. It proved an important event in Beethoven’s career as a composer, and he went on to write music which specifically exploited its capabilities. These strengths included an expanded range, a more robust and well-engineered action (derived from the work of Broadwood’s colleague, Stodard), a large, heavily-braced case, and longer bass strings made possible by a split bridge. The improvements added up to an instrument that departed significantly from earlier Viennese and continental pianos, having a greater dynamic range, better tuning stability and a more pungent stage personality.
Last fall an opportunity came up for us to acquire an unrestored 1809 Broadwood grand, a rare instrument (far from being in playing condition) that still stood on a stand, rather than legs. It had barely been touched by poor early restoration, having all of its action intact, if impossibly moth-eaten. It was very dirty, and had the usual case of “cheek warp”.
[nearly 200 years of dust and grime]
A word of explanation: Before the advent of cast iron frames (starting in the 1820’s) the (grand) piano suffered from a fundamental problem stemming from its burgeoning string tension. The strain would slowly deform the case in both square pianos, which twisted longitudinally, and grands, whose bentside was unable to withstand the pull of the larger-gauge iron wire. Eventually the tension would pull the cheek (the short piece of the case on the right of the keywell) downward, lifting the strings off the bridge and compromising the necessary “downbearing”. The strings must press down on the soundboard in order to make the proper mechanical contact, thereby effecting the acoustical amplification of the thin soundboard, which acts as a rocking membrane stiffly suspended over a resonant cavity. Having seen quite a few of these early pianos, I can report that this is a consistently observed problem. Practically every picture of museum Broadwood grands that we looked at showed evidence of the ubiquitous cheek warp.
These instruments, especially the squares, were not constructed to last for centuries; that would have compromised their design in another way, making them too heavy and too large. Among the grands, the surprising resonance of the early pianos is partly due to the absence of iron, the feeling that the piano is balanced on the edge of its tolerances and built as lightly as possible, like a violin or a lute. Only later were pianos built with industrial standards of durability and weight, when iron was introduced. The later pianos are not better, just different.
Our Broadwood revealed itself when it was opened: Virtually all the casework is supported by elaborately crossed members and liners made entirely of spruce or pine—softwoods! The belly rail, which takes all of the tension of the strings at a right angle, forming the end of the cavity at the keyboard end, is made of a 2 cm (less than one inch) oak plank, into which the transverse members are mortised.
[a view of the interior of the Broadwood with the bottom removed, showing crossed spruce framing and underside of the soundboard]
This design does not mean that the instrument cannot work again, far from it. It simply means that the instrument must be disassembled, the joinery restored and if necessary re-inforced. Marinus van Prattenburg, our restorer, had observed many of these early grands and squares with similar problems of restoration, and he had a solution. Apparently this problem was also under consideration in the evolving design by Broadwood himself.
Marinus had observed several different, increasingly built-up design solutions, including one Collard and Collard of a later date, in which the problem was solved by the introduction of a large high-density block. Early Viennese pianos, such as our Stein copy, also often have a bentside whose treble end is built up, sawn from a larger block of dense timber.
Broadwood appears to have resisted the introduction of metal, as he may have felt that it would cause the treble response to suffer. According to Marinus, Broadwood and other makers tried large solid pieces with holes drilled in them (to minimize weight), rather than heavier frame members, but this also failed. Eventually he followed his competitors and introduced the first composite metal (plates with bars, not unit-cast) frames. But our problem is whether to introduce a re-inforcement that would have to be considered an anachronism into this instrument, in order to be sure that it would never be endangered by having to be taken apart again in the near future.
Marinus felt, rightly I think, that if the fix were not seen, covered by the bottom and top, it would serve the design much better than any metal (or wood) introduced which would be apparent to the casual observer. Moreover, it would show that Broadwoods’ conservatism regarding the introduction of higher density was perhaps unjustified. An instrument which will still function in our time, living a new life after probably four or more generations of muteness, would have to look forward to newer technology for its renewal. But the fix must be partly an illusion, and not be evident.
Part of Marinus’ conviction of the necessity of the replacement of compromised treble framing by a replacement metal re-einforcement also came from seeing a poorly-restored Broadwood grand in which the short-sighted restorer had simply bolted a capo d’astro bar (seen in every modern grand) over the treble strings, creating a monster which worked poorly, disrespected the original, and looked terrible even to uninitiated observer.
Marinus’ solution involved immobilizing the instrument, cranking it square incrementally over weeks with a jack, and re-glueing the joinery of the cheek. Then he would install the re-inforcement, which would ensure the stability of the repair, as well as being entirely removable and invisible, with minimal alteration to the original. He had tried this in two previous comparable instruments and it looked and sounded perfect, with no hint of returning warping even after nearly 15 years. I had seen and heard both of these earlier successful restorations, so I agreed to this creative solution to the problem.
[Marinus’ "pastry box" re-inforcement with the iron gap stretcher in the foreground. The cheek has already been re-joined to the bentside. No part of the metal will touch the soundboard.]
Marinus also removed the soundboard and bottom (early pianos are like harpsichords in this respect, while modern pianos are bottomless) with minimal damage, repaired and replaced them in their original locations, re-using the hand-cut iron nails.
[square section, hand-made iron nails removed from the bottom]
Subsequent newletters will detail the further adventures of the Broadwood, as the action is restored, the instrument is strung, making its first sounds, and coming back to life. We hope to be recording some of the Beethoven Sonatas within a year.
Restoring a Lute by Hans Jordan, 1966
As a lutemaker, I usually refuse to fix instruments. “Doing repairs”, I am often heard to pontificate, “is fixing other people’s mistakes--and I only fix my own mistakes”. And there have been plenty of those, I am the first to admit. “A musical instrument is only a kit” is another of my favourite sayings, reflecting the reality that instruments always change under string tension, as we observed above. Further, the environment stresses the wood as the humidity changes with the seasons; the joinery is continuously under attack because wood expands and contracts differently along or across its grain. The material chosen for a particular function, its position with regard to grain, and the woods to which it is joined are all crucially important to long-term stability.
Recently a colleague visited me, one who was a skilled guitarist, and had had early exposure the lute, owned one which had fallen into disrepair, and wished to see whether it might be rescued. Quite often lutes are not worth repairing, since many poorer quality instruments have been made by makers anxious to establish their names and cultivate a market for beginners who will upgrade to better instruments. This was what I expected.
Imagine my surprise when confronted with a fine lute of the early years of the lute revival, a multi-rib ten-course lute by Hans Jordan of Markneukirchen, the well-known stringed instrument making centre of Southern Germany. Made from the finest materials, this lute was a wonderful thing to see, but time and tension had taken its inevitable toll, and the lute was no longer in playing condition. Its fingerboard showed that at one time it had been played, probably by a woman, whose fingernails had etched deep grooves in the soft rosewood.
Jordan lutes were played by many pioneering lutenists from his country, including Walter Gerwig (whose recordings are still available), Michael Schaeffer, and my teacher, Eugen Dombois. The latter two were students of Gerwig, and though they progressed to more historically accurate reproductions later in their careers, they began playing the lute on these ten-course Jordan lutes. Gerwig insisted that his students play everything on one lute, in “viel ton” or Renaissance tuning in G, with ten courses, from pitch notation. This enabled the lutenist to transcribe even the baroque repertoire and play it from the score, and this is how Gerwig’s diverse and wonderful recordings of Bach, Italian music and French music of the early 17th and middle 18th centuries were made. The first recordings of baroque lute music in the original tunings ("nouveau ton") on more historically accurate lutes were not made until the late seventies of the previous century.
[photo of unrestored lute, showing a palette knife inserted into the neck joint]
Jordan copied the smaller Tieffenbrucker and Venere collective lute shapes of Venice, a rounded body which clearly shows its Byzantine heritage, but he introduced several engineering improvements consistent with the prevailing thinking of his day. At the time the makers were not necessarily interested in copying old instruments in every detail because their customers generally were not interested in playing lighter instruments without nails in a historically aware manner. Gut strings were not used. Many lutes from the fifties were still using metal frets set into the fingerboard (though this one had tied nylon frets). This tradition of lute-making continued in Germany from the “Vogelweide” lutes (used in female choirs singing lute-accompanied folk songs), the “lutars” of Bruger and Neeman (early scholars and arrangers of the music of Weiss and Bach), and the mandora making of the late 18th century. The latter often had single strings, seven to ten courses in guitars tuning, with many of the features described hereafter.
[planing off the old nut bridge, in preparation for making a new one]
Jordan had copied the bowl of the historical lute, but had added several un-historic features: A full depth binding and “cheater” rib (an outside rib on the edge of the bowl) and lining, like the guitar. He made a bridge for 10 double courses with an ivory nut. The soundboard was about the normal thickness of 2 mm, but a liner or patch was found under the bridge, to prevent the cupping which I call the “potato chip” contour, the deformation of the soundboard in front of the bridge and behind the rose, now considered desirable. He had invented an ingenious wheeled ivory nut, supposed to improve the tuning facility. The instrument had a “theorboed” head, a separate pegbox which extends the string length of the lowest three courses. The neck was re-inforced by two rosewood rails, glued into it. Even the body frets had been inlaid into the soundboard; normally they are simply glued on.
The most remarkable design idea was observed after I took the instrument apart, removing its binding, soundboard and fingerboard. I discovered that it had a complicated and very difficult-to-execute dovetail joint between body and neck, where historic lutes simply have a mitred butt joint. This elaborate construction had not prevented the high-tension overspun strings from breaking the neck joint partially through, elevating the action (string height) to the point that the instrument was rendered unplayable. What to do?
The dilemma is this, remarkably similar to that of the Broadwood: Should I restore the instrument, which is a baroque lute built on a renaissance body, exactly as it was built, including its apparent liabilities? Or should I restore the instrument, respecting its essential features and integrity, with some kind of historical model in mind. I also needed to take into account that its owner must find some utility in it, having access to a repertoire that interests and engages him, rather than simply making the instrument work, regardless of its blindness to accuracy. It’s a rhetorical question.
I sought authority for my proposal from the owner: I would repair the lute, renovating its body joint, making the neck angle reasonable; replacing the nut bridge with a more historical one which added a pair of strings, removing the patch which stiffened the top unneccesarily; I would renovate the pegbox by adding two pegs for the extra course, and re-configure the theorboed head to reflect a lute design which actually did exist, in the transitional period of the late 17th century. It is this period in which lutes of this type were frequently included in the painting of the Dutch masters of genre painting: Vermeer, Leyster, Hals, Mitsou, et al.
They painted these lutes in so many pictures that this type of lute is often colloquially referred to as a “Dutch head” lute. The heads were designed to improve the sound of the lowest gut bass strings, and had a brief vogue in Northern Europe before the introduction of overspun and gimped strings which incorporated fine copper or silver wire.
I felt that this renovation would give the lute a useful life, providing access for the owner to a larger and more interesting literature than the relatively short historical span of the ten course lute, and more importantly, giving the extended bass strings something to do. I would actually be giving the instrument back its proper historic place, while making only minor changes, but also rendering it far more useful.
[a shot of the dovetailed neck and body joint before repair. The screw hole visible in the centre is the means by which the joint is clamped when glue was introduced. I re-used this hole to re-clamp the newly set neck joint.]
The first challenge was the restoration of the dovetail, which involved getting it apart. I soaked the joint with water, using a carefully cut and taped cotton wad. The end grain of the neck sopped up water like a sponge. Over a period of five days I slowly heated and knifed the joint apart. Then I lined the female part of the joint, and refitted its surfaces, including the all-important neck/body mitre, by lining them with pearwood veneer. Close-fitting a joint with four angled surfaces is quite a challenge.
This re-fit allowed for putting the neck angle back to about one degree less than 180. I made special blocks to keep the tops of the neck and body flush, and I used hide glue to re-assemble the joint. Then I made a new historical-style bridge and glued it to the top, after planning and scraping the old one off. Tapping and listening to the top told me that I should remove a small amount of wood from the bars. It was fairly robust and mortised into the linings, something historical lute makers never did. After re-fitting the top and glueing it on, I made a new ebony fingerboard, glued it on and scraped and planed it to a slightly convex contour. The old wheeled nut was replaced with new one, carefully shaped to minimize the string “choking” friction which Jordan tried to overcome. Then I made a new theorboed head plate, which allowed for a new course. Having made two copied pegs and drilling and reaming the pegbox for them, I re-fitted all the ebony pegs. I added new body frets made from re-cycled piano-key ivory and re-strung the instrument, after renewing the French-polished finish.
[the renovated pegbox; note the posts which support the string tension, new pegs and four new nuts, one for each course]
I was amazed to find that such a small string length (57 cm.) would give such a lovely sound. Normally baroque lutes have a string length of around 65 to 70 cm., with correspondingly longer bodies expressly designed from the bass lutes of earlier generations. But this lute worked wonderfully. The extra-length basses sounded finely graded in their response. The lute is easy to play, and stays very well in tune because of its overall heft. I am happy with my decisions, and I think the owner is well pleased. The overall appearance and utility are restored, with a measure of historicity re-introduced. I made a label for the interior, readable only with a light, which details all the changes, with the date and name of the restorer. On this lute the player can tour most of the French and early German repertoire of the mid to late 17th c., as well as the six early suites of Weiss, and some music by Bach. The short string length makes a lot of the literature a bit easier, if anything.
[the re-finished bowl, made from maple and cherry ribs]
Recording the Vihuela
I built the vihuelas, four of them, and their cases (see the photo essay [http://www.earlymusicstudio.com/building_vihuelas/ ] under Period Instruments and the last newsletter for more on this). Now, after going through the music that I’ve played for twenty years, it’s time to record some of it for the website. First, I wanted to have some well-known music, because one thing my website counter tells me is how often people use the search engines to find recordings of the best known music. Presumably they are playing it themselves, and they want some models. So I selected my pieces, choosing a balance of more obscure material (the marvellous and complex Milan Fantasia del Segundo Tono) and better known pieces (the Mudarra harp fantasia and the Conde Claros and Guardame las Vacas variations).
We made contact with a fine musician with much experience as a producer of movie and television score material. His name is Stu Goldberg. I warmed to his famous name (the eponymous variations of Bach!) immediately. He is a brilliant jazz pianist and composer as well as a producer. Having decided that perhaps they would like to move, his wife found a small vineyard and large house for sale in Penticton, in the south Okanagan valley, on the internet. He moved his studio into the basement after doing a complete retrofit. His new studio has a vocal recording room, mixing room with many keyboard synths, a blinding array of gear and a recording room with piano. His workplace is acoustically treated and fed with clean power, with the latest recording and editing suite, and as a bonus, the whole thing has a fantastic mountain and lake view through large windows. It's certainly not your average troglodytic recording studio.
Stu in his studio
We discussed the recording project and how it might sound, how we would produce and edit the raw material and create high-quality tracks for downloading which might reflect my particular ideas about the vihuela’s sound, its original soundscape and the values of the music.
It’s quite clear that the vihuela composers did not share the secular world of early Italian lute music, with its chordal dance style and freedom of form and content. The vihuela music is very Catholic, with idiomatic vocal inflection, and late medieval modality still a strong influence. It proceeds at its own pace, almost oblivious to the listener’s desire for concise narrative.
Having practised the session for weeks with my own very out-of-date digital tape machine (which still works perfectly), I had every reason to expect that I would be able to make properly paced recordings of several takes, then do a bit of clean up for the inevitable noises that plague closely-miked fretted instruments. Before I began each take, heart, as usual, in mouth, I took calm from looking at the snow-covered vineyards, with mountains and lake laid out breathtakingly before me. And I played.
[playing the vihuela at a table in the studio]
Stu made it easy. He made me feel right at home, with a musician’s sense of the timing and attention to detail. He was ready with the right supportive patter, and unfailingly bright. (“It’s a beautiful thing!”) I was amazed at his sixth sense for the correct and coherent assembly of the raw material, assisted by the very big ears of Susan, who has done this for me for many years now. Including breaks for dinner and lunch, we recorded and assembled 25 minutes of music in only 14 hours! Record time!
Finally we achieved a creditable reading for each of the pieces, some of them famously difficult. Then it’s time to master. This involved writing a processed file which places the highs and lows of pitch and volume in the right perspective, using equalization. The beginnings and endings are given proper attention. Then reverberation is added to make the harmonic and melodic contours more apparent to the human psychology of hearing. He rolled off the bass, he boosted the mid-range, in the frequency band of human speech, slightly. This gives the listener the impression that the music is right in her ear, that it speaks personally. It's rather like editing a movie, in many respects.
Then we auditioned several reverb software options. This is a field which has become much more sophisticated in the recent past. A Dutch company has taken the trouble to broadcast an acoustical signal and record it with different mike positions in several large venues which have been observed to have excellent acoustics for listening. By filtering the signal through these software processors, it’s possible to convincingly simulate hearing the instrument as if it were being heard in that environment. We selected a particularly fine church in Holland, the sort of place which I feel closely mimics the sound of the vihuela as it might have been heard in the contemporary experience of the composers who wrote for it. So that’s what we ended up with—fine recordings of my vihuela made at St. Joseph’s-in-the-Vineyard, in Penticton, B. C.
Thanks to Stu for his wonderful work! (www.stugoldberg.com)
Next time: More Broadwood restoration, the long-promised article on Arnold Dolmetsch, the changing face of music...
Click Here to go Back