Showing posts with label harpsichord. Show all posts
Showing posts with label harpsichord. Show all posts

Wednesday, May 13, 2009

Anatomy of a Virginal: the Jack

N.B.: I am deeply indebted to Mr. Carey Beebe of Carey Beebe Harpsichords Australia for information and insight concerning modern jacks. At his site can be found his own lucid description of the jack mechanism, as well as many photographs of both modern and traditional jacks.


The jacks perform two crucial roles in the virginal: first, a jack plucks the string when a key is depressed, and second, it damps the string to stop the note when the key is released.

Modern jacks are usually made from molded plastic, although some makers offer traditional jacks in hardwood (at increased cost). Installed in each jack is a tongue, an independent piece which rides on a hinge so that it may rock back and forth. And into each tongue is installed a plectrum, the little strip of thin material that actually plucks the string.

One of the great breakthroughs of instrument design must be the simple yet ingenious design of the harpsichord/virginal jack. The trick is that the tongue must hold the plectrum rigidly in place on the way up when the key is depressed, so that the plectrum plucks the string, causing it to vibrate and produce the sound. But at the conclusion of the note, as the jack comes back down to its resting place, the tongue must allow the plectrum to pivot backwards and slip around the string. If this did not happen, after the initial pluck the plectrum would simply come to rest on top of the string and the jack would have to be reset manually before that note could be played again. In such a case, each note could only be played once in each piece; twelve-tone serial music probably would have evolved much earlier in history.



Fortunately for us, someone (no one knows who) figured out that placing a little stop in front of the tongue keeps the tongue immobilized as the jack rides up, holding the plectrum steady in order to pluck the string. On the way back down, the hinge allows the tongue to pivot back out of the way of the string, and then a small spring (this was traditionally made of brass wire, or more usually, hog bristle) attached to the back of the tongue pops it back into place under the string, ready to pluck again. In real time, this whole operation takes only a tiny fraction of a second, allowing a single note to be repeated quite rapidly.

Nowadays the springs, along with most of the rest of the jack, are normally made of the same plastic (usually Delrin) as the rest of the jack, rescuing many thousands of innocent hogs from the cruel practice of annual depilation.

A small felt damper is inserted into a slot in the top of the jack, with the bottom of the felt positioned just above the level of the plectrum. This felt damps the string, stopping the note when the key is released. Since the plectrum and damper are part of the same mechanism, there is no way to lift all the dampers simultaneously, as with a piano damper pedal, and each note ends as soon as its key is released. In terms of practical musical impact, this means that any "smoothing" or "blending" of one note into the next to achieve a legato effect must be done entirely by the fingers; no cheating with the damper pedal. Sicut erat in principio, et nunc, et semper …

The jack pictured here is not from the Zuckermann virginal kit, but from a French-style harpsichord built from a Hubbard kit. It is a little more sophisticated than the old Zuckermann jack, although the modern touches also create some practical problems. The major differences are the tongue stop, the bottom adjustment screw, and the hinge. In the Zuckermann jack, the hinge is dispensed with altogether and the spring takes over its function. This sounds a bit suspect but it works well enough in practice. We will see how it holds up to constant use once the instrument is completed. As you can see in the photos, the tongue stop in the Hubbard jack is actually the end of a small set screw, which allows the projection of the plectrum to be adjusted. While it is a fine idea to be able to control the plectrum projection so conveniently, it also changes the angle of the plectrum from 90 degrees to a more acute angle, which can adversely affect the voicing. In addition, this raises the height of the plectrum, meaning the string is plucked earlier in the keystroke. Ideally, all of the plectra should pluck their strings at the same point in the stroke - otherwise the response of the instrument becomes inconsistent, undermining the precise rhythmic control which is a hallmark of fine harpsichord playing. Finally, the Hubbard jack includes a bottom adjustment screw, allowing the overall height of the jack to be adjusted. Again, it's a nice convenience, but the small screw heads can dig into the end felt on the key levers, wearing through and making it impossible to remove the keyboard without removing all the jacks first. (Remember, a typical large harpsichord can have 189 or more jacks. Removing and then replacing them all can easily add up to an hour to an otherwise quick operation.) The Zuckermann jacks are supplied slightly too long, and the builder simply files them down until their height is correct. If you go too far and file too much off the bottom, you can always use masking tape or glue some felt to the bottom to raise it again. Historically, jacks never had the bottom adjustment screw; it has been theorized (I believe by Frank Hubbard) that screws with sufficiently fine threads were generally unavailable until the 19th century, well after the "golden age" of harpsichord manufacture (approximately 1500-1800).

A brief video of the harpsichord/virginal action will probably make clear its operation more effectively than any number of words. Here is a clip showing a jack plucking a single note. Since it takes only a tiny fraction of a second, it is difficult to show the tongue pivoting backwards on its hinge to allow the plectrum to go around the string on the way down, so I have included a slow-motion replay of this stage of the action at the end. Choose the High Quality (HQ) setting for better video detail.

Tuesday, May 12, 2009

Anatomy of a Virginal: the Case

At this point an explanation of the parts of the virginal is in order so that the descriptions of the assembly make sense. The case is essentially a rectangular box, open at the top, with a thin soundboard installed about 2/3 of the way up from the bottom of the case, and an opening in the front where the keyboard goes in. The keyboard is screwed into the case from the bottom, and can be removed fairly easily for service.


In this photo, the case is almost complete, and the strings have been installed. Several pieces are missing from the photo: the keyboard, jacks, jack rail, and lid. But let's focus on the case for the moment.

Here is a close-up of the left side of the virginal. Descriptions of the labeled parts are below. Please ignore the horrific barrel distortion in my camera lens. The sides of the instrument really are straight in real life.




At the left end of the instrument, the end of each string is twisted into a loop, and the loop is secured around a hitch pin. This pin holds the string in place, and is buried deep in a block of wood to withstand the tension put on it when the strings are tightened. The longer bass strings attach to hitch pins along the edge of the virginal, but the shorter treble strings attach to the pins running in a diagonal line. A block of thick wood runs underneath the soundboard underneath this line; the soundboard itself is only 1/8" thick, and could never hold up under the tension of the strings. Here are three hitchpin loops, illustrating the range of wire gauges used. The rightmost wire is 8 thousandths of an inch in diameter.


Moving toward the right from the hitch pins, we come to the left bridge. The purpose of the bridge is twofold: first, it defines the speaking length of the strings; and second, it transfers vibrations from the string to the soundboard, which amplifies these vibrations into the sound you hear.

To the right of the bridge is the top jack guide, also called the register. As you can see, this guide consists of 54 slots (one for each note), each of which houses a jack. There is actually another guide inside the instrument, about two inches below the top guide, which is not visible here. Together, the upper and lower guides keep each jack in the correct plane as it travels up and down to pluck the string. Without these guides the jacks would wiggle back and forth, making it impossible to voice the instrument consistently. In historical Italian virginals and harpsichords, a single, deep guide that served the purpose of both upper and lower guides was painstakingly cut from a single piece of hardwood. If the slots were not cut perfectly, the jacks would rattle in them and make a sound one author likened to "the grunting of pigs." The finest instruments did not of course suffer from this porcine affliction, although Italian harpsichords generally suffered a bad rap for many years.

The right side of the virginal is a sort of inexact mirror image of the left side, the major difference being the presence of tuning pins rather than hitch pins.


On the left side of this photo, we see the termination of the hitch pins, left bridge, and jack guide from the last photo. The virginal is unique in that it has not one but two bridges, one at each end of the speaking length of the strings. Harpsichords normally have one bridge on the soundboard, and a nut that runs straight across the instrument, perpendicular to the strings. This nut is not fixed to the soundboard and does not transmit vibrations from the strings.


On this end, the strings are wrapped tightly around tuning pins, which are embedded in a block of hardwood called a wrest plank (underneath the soundboard). On a piano, this block is called a pin block. You use a tuning lever to twist the tuning pins, tightening or loosening the wire to tune each note.



Once the keyboard and jacks are finished, a plank of wood called the jack rail will be installed directly above the jack guide. This limits how far the jacks can travel and keeps them from flying out of the instrument and into the listeners' soup.

A decorative ornament called a rose, usually made from wood or heavy parchment, is often installed in a hole cut into the soundboard. I did not feel quite confident enough to attack my soundboard with the Black & Decker as of yet, so I decided to skip this for the time being. I may decide to add it later, although it will mean removing several strings to get at the soundboard. In any case, several authors note that the rose makes little if any difference to the sound, so it can wait. Here are some links to representative roses in the meantime:

From an Italian virginal of 1672
More information on this instrument

From a French harpsichord of 1683
More information on this instrument

The small hole in the upper right corner of the Zuckermann virginal is not a rose; a small toolbox with its cover lies in this corner, and the hole lets the player lift the cover off.

Monday, April 20, 2009

The Zuckermann Italian Virginal III

The kit instrument I am assembling is the Italian Virginal III, manufactured by Zuckermann Harpsichords, Inc. Zuckermann are still very much in business, and although they no longer offer the Italian Virginal as a kit, the finished instrument is still available. If after reading this you are just dying to start on your own virginal, a similar kit is available from The Paris Workshop, or you can try your hand at a Flemish virginal kit from Hubbard Harpsichords.

According to recent correspondence with Zuckermann, the Italian Virginal III kit was available from "the mid to late 1970s" which makes my kit between 29-34 years old. As I mentioned previously, I purchased the partially-assembled kit from a colleague of mine at my place of employment. He in turn "inherited" the kit from the previous owner of his house. The previous two owners of the kit had brought the construction fairly far along, completing the case and installing the soundboard, so most of the large assembly steps were done before I took over. Incredibly, all of the parts are still together after all these years. Unsurprisingly, the one item that seems to have migrated away from the kit is the included X-acto knife.

This virginal exemplifies the "false inner-outer" construction that Italian makers favored for centuries. The earliest Italian harpsichords were made of extremely thin cypress; while this made the instruments remarkably light (some single Italian harpsichords are reputedly light enough to carry under one arm) they must also have been very fragile. These harpsichords were kitted out with protective outer cases of heavier construction, which could be decorated with varnish, tooled leather, gilt, and so on. Over time, the separate inner/outer construction was abandoned, but the conceit of a smaller, fragile instrument inside a larger case was retained by the strategic use of veneer and molding on the (single) case.

In terms of modern harpsichord kit-building, the Italian Virginal represents one of the first Zuckermann instruments designed by then-owner David Jacques Way. The Zuckermann business had been successful purveying a harpsichord kit that represented a middle road between historical authenticity and economical modern manufacturing, as well as realistic expectations of amateur assembly. Way gradually took the kit business firmly in the direction of historical authenticity by eliminating modern materials such as plywood, reinstating the bentside in the harpsichord kit, and other less visible changes. The Zuckermann Italian Virginal represents an early stage in this process. The design is very close to that of historical models. While most of the construction is from plank wood (rather than plywood), and the soundboard appears to be solid wood, the jack register and lower guide appear to be made of plastic, and the jacks themselves, as is nearly universal in kits nowadays, are made of Delrin, the same plastic the plectra are made from. The most unfortunate concession to modern manufacturing lies in the plastic sharps on the keyboard. These look just awful next to all the beautiful bare wood; I may replace these with ebony or stained fruitwood if time and budget allow.

At this point there is a lot of little fiddly work to do, mostly consisting of small jobs that have to be done 54 times each (once for each note on the instrument). As the manual points out, though, some large double harpsichords have 189 strings, so the virginal should come together comparatively quickly.

Speaking of notes, the range of the keyboard is BB-e'''. That means the lowest note is the B two octaves and a half-step below middle C, and the highest note is the E two octaves and a major third above middle C. Looking at historical examples of this type of instrument, this seems to be a common enough range (if anything, the bass has a bit more range than many 16th and 17th century virginals and harpsichords). However, that low BB is somewhat deceiving. Many virginals and harpsichords prior to the 18th century were tuned to what is now called the "short octave," meaning that the last five or so notes do not descend chromatically; instead the sharps are tuned to diatonic notes, usually a major third lower than their "normal" pitches.

For example, a very common keyboard layout for a small instrument such as a spinet or virginal would have a bass range extending down to the E one octave and a sixth below middle c; but this note would be tuned to the C two octaves below middle c. Likewise, the lowest F# was tuned down to D, and the G# was tuned down to E. While this extended the bass range slightly, it unfortunately also eliminated four accidentals from the lowest octave (C# and D#, which are not on the keyboard at all, and F# and G# which sound D and E). But most keyboard music of the time doesn't require accidentals in this low octave, so the musical impact is not damaging. This may sound confusing but it actually works out quite well in practice, as long as the music is not too chromatic.

The Zuckermann Italian Virginal III also incorporates a short octave, but it begins lower down on the keyboard. The lowest note, BB, is tuned down to GG (two octaves and a fourth below middle c), the C# down to AA, and the D# down to BB. The impression I get from reading books on historical harpsichord construction is that the first variety of short octave was by far the more common; the deeper range of the kit may be a concession to modern players who may want to play later, more chromatic music on the virginal.

Since my own musical interests skew closer to J. S. Bach, I will probably go ahead and tune the lowest notes chromatically, in order to have the low chromatic notes necessary for the Well-tempered Clavier and the French Suites. Other Bach harpsichord works such as the Goldberg Variations and Teil I of Clavierübung require a fully chromatic bass range down to GG and are not possible to play on this instrument without altering a few notes of Bach's music.

The current Italian Virginals available from Zuckermann and the Paris Workshop appear to have further bowed to modern tastes and are fully chromatic down to GG, giving up two notes at the top end of the keyboard making the top note d'''. It's probably not an accident that this exactly matches the range of Bach's harpsichord music.

At this point, I have the virginal strung (except for two notes - I ran out of the correct wire for these and will have to purchase some more) and have a couple of parallel tasks going. First, now that the strings are on, it's time to pin the left-hand bridge. The small metal pins guide the strings so that they run in exactly the right place, as well as binding the strings to the bridge. The bridge is responsible for transmitting the strings' vibrations to the soundboard, so a good, solid connection between the strings and the bridge is essential for a full tone. Second, I need to finish the keyboard by gluing the slips of boxwood veneer to the softwood key levers, shaping the keys, varnishing the boxwood, leveling and aligning the keys, and finally installing the completed keyboard into the virginal. Then, it will be time to file the jacks so that each one is a custom fit in its respective slot, install the Delrin plectra in the jacks, voice the jacks (trim the plectra so that each note speaks at the same volume level), install the jack rail, attach the lid with hinges, apply the decorative molding to the inside of the case, and paint the case. Piece of cake.

Saturday, April 11, 2009

What is a virginal, anyway?

Basically, a virginal is a sort of sideways harpsichord. The strings are plucked by little plectra, like tiny guitar picks, rather than struck by hammers as in a piano. The major difference between a harpsichord and a virginal is that where the harpsichord strings run straight back away from the player (as in a grand piano), virginal strings run left-to-right. I suppose the idea is to save space in a domestic environment, although the actual space savings are not as large as you might hope, since the size of either instrument is largely determined by the length of the bass strings. Still, the virginal keyboard is situated on the side of the instrument, rather than at one end as with a harpsichord or grand piano, which makes placement of the instrument a little more flexible since you don't have to provide room at the end for the player.

Due to the left-right orientation of the strings, there is usually only space to equip a virginal with a single choir of strings (a "choir" just means a single set of strings, one for each note on the keyboard). In a typical harpsichord, there would be two choirs for a single-manual (one-keyboard) instrument, or three or more choirs for a double-manual instrument. This not only provides greater volume of sound, but allows the player to vary the tone color somewhat, since each choir can be engaged or disengaged individually. Since virginals have only one choir, it is not possible to change tone color in this way, but that doesn't mean there is no variation in the timbre. The jacks (mechanisms that hold the plectra) in a virginal do not stretch across the instrument in a straight line (as they do in a harpsichord); rather, the jacks are arranged on the diagonal, so that the relative plucking point of each string is slightly different that its neighbors. The effect is a bit like the similar yet distinct timbres of violin, viola, and cello in a string quartet or orchestra; although each instrument has a distinctive voice, they can blend together logically almost like a single instrument.

There are several styles of virginals. The kit I am assembling is representative of the Italian style, which consists of a basically rectangular box with the keyboard jutting out from one long side of the instrument, slightly to the left. The Flemish style virginals are also basically rectangular, but the keyboard is recessed into the side of the instrument. Flemish virginals can be further divided into "spinets" and "muselaars;" spinets have the keyboard on the left side and muselaars have the keyboard on the right. This has a radical effect on the sound produced. Spinets pluck the string close to the nut (in other words, towards one end of the string) and as a result the sound is rather thin and nasal. Muselaars, in contrast, pluck the string closer to its middle and the resulting sound has been described as "round," "flute-like," and "plummy." If you play the guitar, you can simulate the effect of either type by plucking near the bridge or over the fingerboard.

The most elaborate virginals were made by the Flemish makers such as the famous Ruckers family. Called "mother and child," these included a second, smaller instrument that slid into a recess in the side of the larger instrument and could be removed from the recess for performance. This smaller instrument was called an "ottavino" since it sounded one octave higher than the larger virginal. If desired, the "child" could be placed on top of the "mother's" jacks (after the jackrail had been removed) and slots in the bottom of the child would allow the mother's jacks to operate the child's keyboard, coupling the two mechanisms. When coupled this way, the mother's keyboard would play both instruments.



Other types of virginals exist, which eliminate the "dead space" in the corners, resulting in an instrument shaped like an irregular pentagon, or even an elongated triangle. While slightly smaller and lighter, according to some authorities the tone of these instruments is not as full or attractive as the rectangular virginals.

So why is it called a virginal? No one is really sure. Sadly, there is no truth to the notion that it was named for the Virgin Queen, Elizabeth I of England (although she was, by the way, reportedly an excellent player of the instrument). It seems likely that as a domestic instrument, it would have been associated with young girls and unmarried women, hence the name.

One more note about the name: in Elizabethan English usage, you often come across the term "virginals" or "pair of virginals." Much like a "pair of scissors" or a "pair of pants," this refers, paradoxically, to a single instrument. Furthermore, at that time, the term "virginal" could refer to any plucked-string keyboard instrument, such as a spinet, muselaar, or harpsichord.

In my next post, I'll have some more information about the specific kit I am assembling, as well as some photos.
Recently, I purchased an unfinished Italian virginal kit from a colleague. To the best of my knowledge, the kit was manufactured sometime between 1975 and 1980. As I work on the instrument, I'll document my progress on this blog.