Saturday, 3 February 2018

Because You Asked: Body Frets

I recently received a request for a "how to" blog post from Mr Bill Good, and I'm going to do it.
This time out, I'm going to describe my procedure for installing body frets.

It's one of the very last things I do when I'm making a lute, or a group of lutes. Everything else is pretty much done--all the woody bits have been glued together, a nice, shiny finish has been slapped on, I've fitted a set of tuning pegs (now there's a finicky procedure), made a nut, tied on a set of gut frets, and strung up the lute. I've tuned those strings and played the first, quivering, barely in-tune notes, and an indescribable thrill has begun to form in my being as I realize, again, that I've created a living, breathing thing. The lute is alive!

The Lute Maker, in the Throes of Creation

In such a state of creative fervour--and after such a long time putting the lute together, weeks and months of sometimes very difficult, demanding work--the impulse can be to... well... get into a bit of a hurry at this late stage, cut corners a little, just tape down some toothpicks, nobody will notice, nobody ever uses the body frets anyway....

No! Don't you even think about it! Calm down--you've still got work to do. An hour's work, if you do it right, and have a procedure to follow. Here's mine.

Finding the Octave

The first thing to do is to tune that lute. And I mean tune it, pretty much constantly, for a couple of days at least. Regardless of the type of strings you've used, they'll be continuously stretching and going out of tune for most of that time. I find it useful to actually play the lute while I'm tuning--in addition to waking the lute up and helping it to find its voice, my playing works and stretches the strings so that they will fall into tune that much faster. Besides, I'm a lute maker, it's my job, and I have built this thing to make music--at this point, what else would I rather be doing?

Once the strings are really well in tune, you need to find the octave, the point approximately halfway up the string where the 12th fret will be located. It's not exactly halfway as measured by a ruler--because of the physics of string stretchiness, the octave will always be somewhat less than half the distance between the nut and bridge. (Depending on a few different factors, the octave will be probably be approximately 1.5 to 2mm shorter than half the measured distance.)

How to find the octave? Allow me to demonstrate. This is a lute that I recently repaired--not one of mine, but one made in the early 1980s by a maker in British Columbia. Like many lutes of that vintage, the fingerboard had warped a bit over the years, and needed to be re-shaped in order for the lute to be playable. In order to do the re-shaping, I had to remove all the tied frets as well as the body frets that were on the lute. Now that the work on the fingerboard was finished, I needed to put a set of new body frets down. (By the way, the gut frets I've put on are placed provisionally, and will be adjusted when I find the octave and calculate all the fret positions.)

I have a small ebony body fret that I have shaped and keep just for the purpose of finding the octave (I'll talk about how to shape these body frets in a minute or two.) I put the fret down on the belly under the top string, and strike a note.

I move that fret up and down the string, making slight adjustments one way or the other, plucking notes until I've found the true octave. I use an electronic tuner to do this (because I am lazy), but if you're a purist you could probably do it by ear, comparing the fretted note with the natural harmonic produced at the 12th fret. Whichever way you do it, when you are convinced that you've got that fret and that octave in the right place, take a very sharp but not very hard pencil (HB works well), and make a crisp, light mark on the soundboard on each side of the fret. Move the fret away. Then, between those two marks, make another mark exactly in the middle. That is the precise position of the octave.

Next, get out your ruler, and underneath the top string, measure accurately the distance between the octave mark you've just made and the front edge of the nut. Now, double that number: this gives the lute's playing string length. Use it to calculate the position of all twelve frets, according to whatever system of temperament you are using (this lute, a 9 course, will be fretted in equal temperament.) Cut a piece of card stock about an inch wide and a bit longer than the length of your octave, and carefully mark out the fret positions on it. You'll use this card to lay out the positions of the body frets (there will be three on this lute), and to adjust the positions of the tied gut frets on the neck.

I also write down on this card the diameters of each of the tied frets, and when I'm done using it to lay out the body frets and adjust the tied frets, I put it in the lute case and keep it there. It's very handy to have when replacing worn out gut frets.

Making the frets

After erasing the marks I've made on the soundboard, I lay a strip of 1/2" masking tape down on the soundboard, so that the inside edge is more or less in line with the edge of the fingerboard (the strip of tape isn't quite parallel to the first course, but flares a little.) This tape will mark the location of one end of each of the body frets, so I lay my fret card down alongside it, and mark the locations of the body frets on the edge of the tape--like so:

I can then get out my ruler, and determine the length of each of the body frets.

I want this 12th fret to cover the first three courses (the 11th fret will cover 4 courses, and the 10th fret will cover 5), so I measure the distance between the edge of my masking tape and the mid-point between the 3rd and 4th course. (You can see my mark on the masking tape on the left side of the ruler; you can also see the length of this fret will be 33mm.)

On this lute only the 11th and 12th fret will be on the belly, and the 10th fret will be on the fingerboard. In order to determine the length of the 10th fret, I need to measure from the fingerboard edge to the mid-point between the 5th and 6th course. You can see the fingerboard edge clearly here: the length of this fret will be 51mm.

Once I have the length of the frets figured out (and written down someplace), I can decide on the thickness of frets that I want to use.

This can be a bit of a tricky calculation, and it deserves some thought. The thickness of the body frets you use will depend partially on the thickness of the tied frets on the neck, and partially on how much and how quickly the belly slopes down and away from the strings, once you get past the body-neck joint. The slope of the belly at this point depends on the amount and severity of the "belly scoop" that the luthier has planed into the edge rib of the lute: some makers like a deep belly scoop which slopes quickly away from the fingerboard, while others prefer a more gentle, gradual slope (that's generally the way I like to do it, and that's the way the maker of this lute decided to work.)

Let me explain how I decided on body fret thicknesses for this lute. My general method of fretting a lute (which I normally use unless a client asks for a different setup) is to begin with a thick fret for the first fret--say, 1.05mm for the first--and then gradually tie smaller frets all the way up the neck. So, for a 9-fret neck like this one, this is how the tied frets go:

1st: 1.05mm
3rd: 0.95
4th: 0.90
5th: 0.85
6th: 0.80
7th: 0.75
8th: 0.75
9th: 0.75

The 10th fret is my first wooden fret and it will be on the neck, so  I will make it 0.75 thick as well, to give some uniformity of "feel" with the last three tied frets. Then, because the next two frets are on the belly--which is gradually sloping away from the string, and creating a bit more clearance--I will grade the 11th and 12th slightly thicker--0.85 for the 11th, 0.95 for the 12th.

Now let's make some frets.

I make fret blanks out of ebony and english boxwood, two very hard woods that will withstand years of playing, and generally choose the light or dark wood depending on the overall look of the lute. (On this lute, the original body frets that I removed were ebony, so that's what I put back on.) I make my body fret blanks 1.4mm wide, and I'm able to thickness them quite precisely to various thicknesses, as you can see in the photo above. This is the tool I use to size them up, a pull-through scraper:

I have described this tool and its use in detail in a previous blog post from a few years back. You can visit that blog post here (you'll need to scroll to near the end of the post.) It's well worth a look--I use this unit to make sets of rib spacers, ebony edge bindings, strips for striped veneers, and much more. It's an incredibly handy tool in the lute maker's workshop.

I take the blank of desired thickness, and cut a segment 3 or 4mm longer than the final finished length. Then I dress the edges of the fret, using this jig, held in a benchtop vice.

It's just a piece of particle board with some arborite on top--I think it came from a piece of old kitchen countertop. The arborite edge is nice and sturdy for holding the fret. I have a thin piece of some kind of plastic sheet stuck down on top with 2-sided tape, forming a ledge that's as wide as the fret, where I can hold the fret securely while working it.

With a #2 flat file, I break the top edges on both sides of the fret. I file them at about a 45° angle, then round over the two new edges I've just created. I'm careful not to go over onto the top of the fret--I don't want to lower it, I just want to make those edges comfortable to press a string down onto.

Then I relieve the sharp edges at the end of the fret. I put the end of the fret at the edge of my jig, and with my #2 file, give a quick downward stroke to put a 45° angle on it. Then I relieve the edges on that angle by trailing the end of the fret across the file. I then cut the fret to its final length with a sharp chisel, and dress that end of the fret in the same way. Here's how the fret ends up:

Now that I've got all the frets made, i can finish laying them out.

Laying out the frets

I need full access to the end of the neck and top of the belly to lay out the frets, so I unstring the lute. I just loosen off all the strings without removing them from their pegs, and tie them away from the top of the soundboard.

I lay down the frets in their approximate positions, according to the marks I've put on my long piece of masking tape. I then stick down smaller pieces of tape at the other ends, and bring back my fret card to mark the fret positions on them. Then I move the frets away, and use my small ruler to draw a faint line on the wood of the belly from the mark on the treble side to the one on the bass side.

I then use a small, sharp, curved scraper to scrape away the lines I've just drawn. I'm not really scraping away very much wood--mostly just a thin layer of finish on the belly and neck so that the glue I use to stick the fret down will adhere to the wood below.

Gluing the frets

I use hide glue for this job, as I do for all the gluing jobs on a lute. You might be tempted to use yellow carpenter's glue for this, which would be fine, but I find that hide glue tacks faster and cleans up better--instead of waiting overnight for yellow glue to dry, you can use hide glue and be playing again in an hour.

I use a fairly thin glue for this job--you don't need a lot of glue to get these frets to stick. 

I wrap a little two-sided tape around the end of a small dowel, and just press that down onto top of the fret so I can lift it and hold it over the glue pot while I brush some glue on (the more dextrous among you could probably just as easily use tweezers for this job.) I just dip the brush in the glue, sweep upward onto the fret, then sweep back down to brush off excess. Then I stick the fret down quickly in position--locate one end of the fret close to its mark, then bring the other end down. It should be good--if not, adjust. Hold the fret down for about 30 seconds, until the glue begins to set. 

Then bring in the small metal ruler, and use it to press the fret down until the glue is well set (a minute or two of pressure should be good.)

Then move on to the next fret, and go through the same procedure, then the next. By the time you've got all three frets stuck down, the glue will have set well enough on the first one that you can go back to it and clean up the squeeze out. Just brush a little cold water on each side of the fret, give it a minute for the glue to soften, and clean up with a sharpened spruce stick. For a last measure, give each fret a quick wipe with a damp bit of paper towel, to remove any glue residue that might be lurking on the top of the fret.

And that, my friends, is that. Remove the masking tape bits, and start the process of tuning the strings. By the time the lute's back in tune, the body frets will be firmed up, and you're ready to play again.

So that's my procedure. It took a long time (and a lot of photos) to describe a fairly simple operation. Still, some might object that I go too far--that all this rounding and filing and fiddling is simply paying too much attention to a feature on the lute that isn't, after all, that important. To which I would reply: there is no detail on a lute that is too small to pay attention to. And in the case of these body frets, a lot rides on their position and shape. If they're not in the right place, the intonation will be wrong. If they're left square-edged, the note will sound thin, but more to the point (ahem), they will feel unfriendly to the fingertips. I've tried many lutes that had body frets with edges left square, and I've always had the uncomfortable feeling that I was stubbing my fingertips on them. Why would I want to leave something unfinished on the lute that I felt was impeding my playing?

When players and makers think of "playability", they're mainly thinking of big things like string spacings, action (string height above the fingerboard), fingerboard arching, and neck shape. But I would argue that apparently "small" things like body frets can make a significant contribution to playability too. There are a few other small details of finishing a lute that make similar contributions; perhaps I'll make them subjects of future posts.

Sunday, 17 December 2017

Here's a Thing That I Made, or, How to Hold A Lute Mold, or, The Return of the Betterizer

I have a question for all the lute makers out there: how do you hold a lute mold?

I've looked around at a lot of makers' websites and visited a few of their workshops, and I've seen a few different contraptions. Some are simple, elegant and useful, while others seem awkward and difficult. Here's a selection:

A plain stick of wood fitted in the underside of the mold, and held in a bench vice. This is the method used by Robert Lundberg, illustrated in his book Historical Lute Construction. Here is a photo of his protegé, Prof. Günter Mark, using the same method:

(Courtesy Liuto Forte)
I once used this method, for the second lute mold I ever made, and found it pretty inadequate. The mold can be tilted forward and back in the bench vice, but that's it. (I guess in a pinch you could turn it 90° in the vice, and tilt it side to side, if you really wanted to.) Not very convenient when you're carving a mold, and want to tilt it a little to scrape a rib facet, or slightly change the angle of light from your shop window or bench lamp. I'm sure this way of holding the mold worked fine for Lundberg (and does so for Prof. Mark), but my eye needs more angles, more ways of seeing. I want a way of holding a mold that allows a full range of motion, so I can set the mold in any position.

Here's a good example of a jig that I imagine works really well: two locking hinges. I've seen quite a few makers using this type of holder; here's a shot of Malcolm Prior with his. (Malcolm thinks this method might have originated with the renowned English maker Stephen Gottlieb.)

I once tried to make one of these jigs out of stacked layers of plywood (again, early on in my lute making career), but lacked the skill or resources to do a good job. Malcolm's version looks very well made and useful--note the tightening levers on the bolts through both hinges. The mold (and lute back in progress) is held very securely, but is fully adjustable, tilting forward and back, as well as side to side, and rotating 360° where it sits in the bench vice (you can't see that part in this photo). I can't comment on its ease of use, never having used one, but it does allow a full range of motion and orientation in any position.

Here's the slickest of all the lute mold holders that I've seen. It belongs to Grant Tomlinson, and he tells me he bought it years ago from a supplier in Vancouver who's no longer in business; it was called a "universal carver's vice."

There's a big steel ball (like a trailer hitch) that bolts into a plate screwed to the underside of the mold. The ball is sandwiched between two half-inch steel plates, one of which is sunk into a big block of hardwood, which is held in the bench vice. A couple of heavy bolts cinch the plates around the ball; one of the bolts may be loosened (note the nut with lever attached in the view below), and the mold may be moved freely, and then secured in any position.

There's the owner of the mold (and mold holder) himself, Grant Tomlinson. As you can see, he's just getting ready to fit up and carve a new top block for this mold. In fact, he's getting ready to make top blocks (and bodies) for a new group of lutes. Here are the other three molds in the group:

Each is fitted with its own steel plate, so it's a fairly small matter to remove the current mold from the holder, and set up another. In this way, Grant can fit and glue a rib on one back, and while that's drying, set up the next mold, and fit and glue another. I doubt that Grant will be working on all four backs at once, but he might work on two at a time--most probably they'll be made of the same material (for instance, yew wood), so that he can set up his planes and other tools a certain way, and not have to change for a harder or softer material.

I got a chance to use this mold holder, and Grant's molds, when I worked with him in 2008-2009 (once again, my thanks to the Canada Council for the Arts for the grant that financed this residency.) One of Grant's instructions about using this piece of equipment is that it's essential to hold onto the work securely when loosening the nut and releasing the vice--if you don't, the mold will just fall over, and the consequences to it, and the back in-progress, could be devastating. (And no, in case you're wondering, I never let the mold slip!) If you've got a good hold of it, everything's fine: just re-position it, tighten the nut, and you're ready to go. 

(By the way, I've looked online for a mold holder like Grant's, and while I have found carving vices that kind of look like it, I've never found one that looked heavy-duty enough to withstand the weight of a mold and the wear and tear of using it pretty constantly for years, as Grant has done with his. If anyone has found a good version, please leave a comment and link at the end of the post--I'd love to have a look.)

What I used for years was a setup that Grant showed me a long time ago--the rig he used before finding that space-age carver's vice that he has now. A block of hardwood is screwed to the underside of the mold, and a large ratchet clamp grips the ends of that block. The bar of the clamp is encased in a built-up block of plywood, and that plywood block is held in the bench vice. It's low-tech, but effective--just the kind of thing I like. Have a look.

Everything's very secure. You can see that I've drilled through the hardwood handle of the clamp and inserted a 3/8" bolt, which gives plenty of leverage when tightening up against the hardwood block. 

Here's a look at the plywood block that holds the clamp, and sits in the vice--it's an essential part of the rig. It's built up of layers of ply, and as you can see, some thicknesses of card to ensure the clamp's held tightly. 

The setup has some drawbacks, but a few things work really well. Once you've clamped up the block a few times, there gets to be a kind of recess pressed into both ends, where the jaws almost automatically fit whenever you clamp up. Also, the clamping pressure is so good that you can tilt the mold by simply pulling it toward you or pushing it away (as shown in the photo above), without having to loosen the clamp. I've found it very handy to have this small amount of flexibility available, when carving the mold or fitting a rib. It allows a quick glimpse at the work from a slightly different angle, which can be a great help in the middle of the process.

Now the drawbacks. The mold tilts away or toward you, or forward and back in the bench vice, but it doesn't rotate, which means it's always facing in the same direction--unless you release it from the clamp or the vice, lift it up, turn it about in your hands, and then clamp or vice it again. (That's what you have to do if, for instance, you've fitted and glued a rib on one side of the body, and want to do the same on the other.) It's a royal pain--I did it for years, and while it's a good workout for your biceps, it's also pretty awkward, and potentially hazardous to the ongoing back assembly.

The main problem, though, is that like the Lundberg/ Mark setup, there's no way to alter the orientation if you want a slightly different view when carving a mold--to see how light and shadow fall upon the complex shape, or to follow a rib facet as it comes up in a long, graceful curve from the neck joint toward the abrupt transition at bottom end. I find mold carving to be a very demanding and difficult process, and as I said above, I need as many different angles and ways of seeing as possible in order to do a decent job of it.

So, a little while ago, I put the betterizer part of my brain to work, and came up with a solution that performs very well. I've talked about "betterizing" before in this workshop diary--I define it as a specific kind of improvement one makes to a tool or a jig in response to a displeasure one has felt building up, almost unconsciously, over many years, toward a certain shortcoming in the tool. It's like a burr under the saddle or a stone in your shoe, an irritation that can be ignored or withstood for a long time, but once decisive action is taken... there is only the freedom and sheer joy of having a tool that performs exactly as you want it to.

Here's what I came up with--a turntable that allows the mold to rotate 360°, and can be locked firmly in any position. As you see, it fits in between the mold and the hardwood block.

Here's a look at it separate from the mold.

There are two knurled nuts, on two bolts, one on each side of the hardwood block. Remove those nuts, and the top plate (along with the hardwood block) can be lifted off, to show the inner workings.  

The turntable base is made up of two thicknesses of 1/2" birch plywood, with one extra thickness of 1/8" ply on the bottom. The upper layer of 1/2" material, as you see here, is cut in two pieces in the bandsaw--an inner circle and an outer ring--to leave a channel about 1/2" wide, wide enough for the shanks of a couple of carriage bolts to travel freely around it. The lower 1/2" layer is cut in two pieces in the same way, but the space between the outer ring and the inner circle is a little more than 1" wide--wide enough to allow the head of the carriage bolt to travel freely. If you flip the base over, this is what you see: the outer 1/8" plywood layer. All of these wood screws are what hold the layers and pieces of the base plate together.

One other crucial feature of the turntable is the length of 1/2" maple dowel which centres all the layers, connecting the top plate and base plate together. It's the axis upon which the turntable turns. This is the underside of the top plate, showing the centre hole into which the dowel fits; the #12 screws that attach the top plate to the hardwood block; and the two holes through which the carriage bolts will fit up when the two plates are put together.

Here are the carriage bolt, spacer disc (made of 3/8" ply), and knurled nut.

Just a word about the knurled nut: it's simply a hexagonal-head nut which I embed within a couple of layers of ply (you can see the threads of the nut winking in the photo below.) I "knurl" the outer edge of the disc in the band saw, and then relieve the rough corners with a file. I could use wing nuts for this purpose, I suppose, but I prefer these shop made nuts because they are secure and easy to twist, which you end up doing a lot with this rig. I think wing nuts would be harder on the fingers, after a long day at the workbench carving a mold.

And here's what it's like, with the plates fitted together, waiting to be fixed to the underside of the mold--

Which I do with a couple of stout #12 wood screws on each side of the turntable (so there are four screws altogether holding it to the mold.)

Once it's fixed to the underside of the mold, just clamp the hardwood block into the big orange ratchet clamp and get to work. To turn the mold around, just back off the nuts about a half-turn, rotate the mold to the desired position, and tighten the nuts back up again. It really is that simple.

I remember the day I first made one of these things and tried it out: I was so thrilled, I must have spent an hour just loosening the nuts, repositioning the mold, tightening the nuts again. The turntable, and the mold, could be spun effortlessly, but there was no danger of anything falling over or falling apart; and when I tightened up the nuts, with only moderate finger pressure, the mold would stay very securely in its position. I could tell I had made a good jig, one that worked and would be safe and durable to use, whether I was carving a mold, or fitting and gluing a rib, or as I'm doing here, carving a top block for a new lute back.

It's like a miracle--I can position this mold in any direction I want, except, perhaps, upside down (but I will work on that.)

So there it is, another troublesome operation in the shop all betterized. It really is not an exaggeration to say that this invention has made my work with molds and backs much easier and more satisfying in so many ways. It's like a toothache is suddenly gone. I can't believe I made lutes before I started using this rig.

If anyone reading wants to make one of these things for him or herself, please be my guest. If you need any further hints on how it's done, just leave a comment here. And if you make any improvements to this rig, please let me know.

Go forth and betterize!

Sunday, 4 June 2017

The Inset Lute Rose: A Contemporary Example, and Some Historical Ones

In my last post I described a lute that I and some of my colleagues here in Vancouver put together for loan to a deserving music student. We started work on it in the lute making class at the LSA Workshop West in 2015, and it had contributions from Grant Tomlinson, Ray Nurse, Wilma Van Berkel, and me: I dubbed our little group the "Commonwealth of Vancouver Lute Makers."

In the post, I mentioned that I didn't carve a new rose for this lute, but instead inset one that I had carved in 2003, for a project that I eventually set aside. I had no qualms about reusing it: the rose looked good (I must say I was a bit impressed by the carving skills of that young whippersnapper Travis Carey); and it saved me a fair amount of time. To give you an idea how much, I usually spend around 20 to 24 hours carving an average lute rose, while inlaying this old one took probably no more than 2 or 3.

A further reason I felt comfortable insetting my old rose was that, from the evidence I've seen, it was a trick of the trade used commonly by the old master lute makers. Just like modern makers, they obviously took a lot of pride and satisfaction in making a rose that was (in the words of Thomas Mace) "smoothly cut," yet they often seemed willing to re-use and recycle suitable old parts in new lutes, one supposes for reasons of economy, as well as aesthetics.  This is true of roses, just as it was true, for instance, of early 16th century Italian lute bodies (and bellies) by Hans Frei and Laux Maler that ended up in France a hundred years later, converted to 10 and 11 course lutes.

Today, what I'd like to do is just give a brief description of my technique for insetting an old rose in a new belly. It's not too complicated, but I think it's worth showing. In my experience as a lute repairer as well as maker, I've seen quite a few instances where otherwise healthy lutes have needed new bellies--and there's often no good reason why the fine original rose shouldn't be preserved and given new life.

After that, I'd like to show you some historical examples of inset roses in lutes that I've come across in my travels.

So first, the technique.

Take one rose, already carved, and in an appropriate place--perhaps the line deeply scored around the perimeter--use a sharp knife or rose cutting chisel to carefully cut the rose out of the surrounding soundboard.

The edge of the cut-out rose is bound to be a little rough, so smooth out the bumps by running a sanding block around the outside. I use a square-sided sanding block and place a block of wood under the rose, so it's raised about 10 mm off the table top. This ensures that the perimeter of the rose, while perhaps not perfectly round, is at least straight-sided (that is,  perpendicular to the surface of the pattern.)
Step two: cut a hole in the new soundboard. Lay out the location of the rose centre along the centreline of the belly, then use a compass to scribe a circle that's approximately 1mm smaller than the diameter of the rose to be inlaid (you should check the diameter of the rose very carefully in many places, since the rose might not be perfectly round.) Use a knife or rose chisel to cut out this circle...
then fashion a disc of, say, 3/4" mdf (medium density fibreboard), a millimetre or two smaller in diameter than the hole you've just cut. Use double-sided tape to stick a strip of 220 grit sand paper to the outside...
and use the disc to carefully shape the hole. Again, use spacers to raise the belly a few millimetres above the table, so when you use the sanding disc, the edges of the hole end up straight-sided.

Proceed very carefully, and check the fit often, from the back side of the belly. Make sure to mark the centreline of the rose, so that its grain orientation matches that of the belly. I found it helpful in the fitting process to work roughly quadrant by quadrant--get a decent fit on the upper right quadrant, then work into the upper left, then lower left, and so on around the circle. Eventually, the rose popped snugly into place.
Once it was fit, and everything looked good from the back and front sides, I just left it in place, and proceeded to gluing. I cut out some pieces of thin handmade paper to overlap from the soundboard partway onto the pattern--specifically, across the outer ring of semicircles to the solid ring partway in.  (I thought that would offer more security than just gluing to the fragile outer part of the pattern.) I applied hide glue to the paper--not too much, because I didn't want to have to clean up great gobs of dried glue later on--and stuck it down. On top of this went a layer of waxed paper; then a circle of thick card slightly bigger than the rose, that covered the whole gluing area; then a batten; then clamps.
Please note that I did not attempt to brush any glue around the outside edge of the rose and try to stick it into the hole in the belly. That would have been totally impractical, and created a terrible mess. I reasoned that the paper would do a good enough job of securing the rose, and that I might incidentally get some glue squeeze from the paper in the (very slight) gap between the rose and the belly. And eventually, the rose would be secured in place by three harmonic bars, as well as numerous other small bars, so I didn't think there was any danger that it might come loose.

This is the look from the back, after the glue dried: 
And the view from the front:
All that's left was to take my rose chisel and carefully cut out the paper that underlay the outer ring of the pattern. Once done, I could treat this belly as I would any other--gluing on a set of bars, fitting into the lute body, etc. Here's the final look:
Nice. I think the Travis of 2003 would be pleased.

Now--let's have a look at some slightly older examples.

I don't get to Europe too often these days--I'm pretty much chained to the workbench (sigh)--but when I do, I head straight for the musical instrument collections (heere bee the lutes!) Julia and I made a trip to Munich and Copenhagen during the Christmas season of 2015-16, and I got to ogle lutes galore. The museums in Munich were wonderful, but I must say the Carl Claudius Samling at the Danish Music Museum was a real revelation. They had just moved into a new space in the former home of the Danish Broadcasting Company. The exhibits were fine, the staff friendly and helpful, and the musical instruments on display were, simply, stunning. We visited on short notice, so I had to be content to look at the lutes through glass in their display cases; at some point in the future, though, I would love to make arrangements to do some research among the instruments in the collection. 

Here's my first example of an inset rose, from a lute in the Carl Claudius Samling, by Andrea Harton (no. 102A). Dated 1617, it was originally a theorbo (the body is ebony and ivory), but was converted at some point to its present configuration as a German Baroque lute. I have no information on when, or by whom, the conversion was done; from the look of things (from a distance, and through glass), to me it seems as though the body may be the only original part of the lute to have survived the conversion.

Or, perhaps, the rose as well?

It's a lovely triple rose, and it is indeed "smoothly cut"--and it is definitely inset. Here's a little closer look:

This view shows clearly a ring of dark-light-dark-light purfling around each of the three parts of the rose. It's an unusual look, but I think it's quite well-done: not only are the roses neatly fitted, but the purfling rings are actually cut into each other slightly, elbowing each other's space a tiny bit, which helps to integrate them together visually. (On original old triple roses, the chip-carved border will often wind and snake and overlap to integrate the three parts.) 

Is this the original rose? I don't know, but at the very least it looks to me like a rose carved in what I would call a high-Renaissance style--the elements are beautifully proportioned, the lines are cleanly and deeply cut, and the work overall gives an impression of efficiency and mastery. If it is not the original rose, it certainly is an original rose, and whoever did the conversion made an inspired choice to include it.

(Of course, it's also an open question whether the three parts are actually from the same original rose--the pattern of the top element is different from the two below it, and though they all appear to be cut by the same steady hand, the chip carved border at the top is a little bit compressed compared to the other two...)

Two more examples, both from the National Music Museum at the University of South Dakota in Vermillion, SD, USA. These are the two superb 13 course baroque lutes of the 1720s by Thomas Edlinger of Prague (nos. 10213 and 10214), both, again, conversions of earlier lutes. They've travelled an interesting path: from the 18th century they were kept stored in the attic of a castle in Bohemia, and were "rediscovered" by the castle's owners only in 1907. From the 1950s until the late 1970s they were loaned to the Kunsthistorisches Museum in Vienna (Robert Lundberg examined and photographed them there; some photos appear in his book "Historical Lute Construction.") In 2002 they were acquired by the National Music Museum. Supported by a research grant from the Saskatchewan Arts Board, I was able to visit the Museum and measure and photograph these lutes in the summer of 2006.

I'm not sure I can adequately describe the thrill I got from looking at and handling these lutes. It was one of those rare instances in my life where all my senses were heightened, my eyes and hands and mind totally engaged in observing the object before me: sensing its weight and balance, the way it shimmered in my hands when I simply spoke a word; observing the woods used and the joints fitted, the curves, cuts, colours and proportions; trying with all my concentration to notice, and commit to memory, every single detail about the lutes.

It's a good thing I did, too, because I sure wasn't able to get much use out of the photos I took. I was working with a (film) 35mm SLR camera, and did not have the expertise to photograph in the rather harsh lighting conditions in the museum's basement (I would love to go back there with a good digital camera.) Most of my shots of the roses were a wash, unfortunately--so for my purposes today, I'll rely on the museum's own photos.

First the larger of the two, NMM 10214. This lute, attributed to the famous Tieffenbrucher family of lute makers in Padua (a label inside reads, "In Padua. Vendelinus Tieffenbruker"), has a beautiful back of 21 striped yew ribs, and probably began life as a bass lute in the late 16th century. Edlinger's label inside shows that the 13 course conversion was done in 1724.

Photo courtesy NMM
Photo courtesy NMM
The rose, one of the most intricate and beautifully carved examples I've ever seen, is inset. From the original lute? Difficult to say--but that doesn't diminish its glory.

Photo courtesy NMM
The pattern of interlocking rings is a fairly common motif in surviving old lute roses, but the carver has added some interesting features--first, a set of eight flower patterns forming a square within the circle; and second, and most intriguingly, a pattern of a sun with a human face at the very centre.

According to Wikipedia, that figure is actually a very common heraldic charge called "the sun in splendour" or "the sun in his glory." In its most usual form, the figure "consists of a round disc with the features of a human face, surrounded by twelve or sixteen rays, alternating wavy and straight...The alternating straight and wavy rays are often said to represent the light and heat of the sun respectively." That's exactly what we have here--the human face, and 16 rays, alternating wavy and straight (the carver's skill in accomplishing this detail in the fragile spruce is simply astonishing.)

But what is the significance of this in a lute rose? Well, the heraldic figure is one that is used on a variety of coats of arms, both of individuals and families, and geographical locations such as towns, cities, and countries. Could the motif point to a possible origin of the rose--either a particular person, or a particular place? I can't say for sure, though a couple of intriguing possibilities present themselves. A quick survey of the coats of arms with the motif listed on the Wikipedia page shows many localities in France and Switzerland, both of which seem unlikely candidates. There are, however, two in the Czech Republic: Loukov, a village and municipality (obec) in Kroměříž District in the Zlín Region, 249 km east of Prague; and Věžky, a village and municipality in Přerov District in the Olomouc Region, 230 km (143 mi) east of Prague. (And by the way, I turned up no associations of this design with Padua or regions near it.)

The coat of arms of Loukov, By Pkotrla at Czech Wikipedia
The coat of arms of Věžky, By Pkotrla at Czech Wikipedia 

Of course this is all speculation (and definitely not an exhaustive survey), and probably of very little consequence for our appreciation of the lute itself. However, to me, as a lute maker and student of the work of the old masters, it is interesting to contemplate the idea that, on balance, it's more likely that the rose's origin is closer to Prague than to Padua--in other words, that the rose might be one that Edlinger had on hand, rather than the original rose in the bass lute.

On to the last example, NMM 10213, the smaller of the two Edlinger lutes. There is a label inside reading "Magno dieffopruchar a venetia" with no date; Edlinger's label alongside it shows that he converted the lute to 13 courses in 1728. This lute, I must say, is a bit of an oddity, though a fascinating one.

Photo courtesy NMM
Photo courtesy NMM

Here's the inset rose:

Photo courtesy NMM
Again, it's a beautifully and crisply carved old rose, complete with the central leaf motif that (I imagine) originally accompanied it. The borders around each of the three circles are most noteworthy to me because they appear to be round-bottomed channels. I've seen this feature in other old lute roses, but no matter what tool I've tried, I've not been able to find one to make this kind of rounded channel in a straight line, let alone in a perfectly-executed circle such you see here. (If any lute makers or carvers reading this have figured it out, please let me know.)

I had actually been interested in this lute as a possible model for a small theorbo, but when I got to the museum and examined it, I saw pretty quickly that the idea wasn't going to work out. Simply put,  the shape is just a little too wonky--the outline is wobbly, and the bowl is kind of lumpy and irregular. I felt that if I used this lute as a model, I would end up having to change the cross sections, longitudinal section, and outline so much that I could barely point to the original lute as the origin for the shape.

But even though I didn't end up using the information I took from this lute to make a new instrument, the time I spent with it in the museum was, I feel, incredibly important to my evolving sense of what a lute actually is or can be. Before then, my idea of "lute" was shaped mainly by my contact with the work of modern makers--lutes built on solid wood molds that have been carefully shaped to iron out all irregularities in body shape, and where every detail of workmanship is executed to the nth degree. Modern lutes, by and large, are pretty much perfect artifacts, and in this way they are very unlike most, if not all, of the old lutes that one sees in museum collections.

This is true of both the Edlinger lutes in South Dakota, but especially true of this one. Besides the wobbly old body and the inset rose, it's pretty clear that Edlinger also re-used an old belly when he put this lute together--as you can see in this photo, he's added a strip of spruce at the very top to fill in the gap where the belly wasn't quite long enough. There also appear to be old "ghost" marks of some previous body frets (you can see one just above the 12th fret); and, though it's hidden behind the third course bass-side string, there appears to be a ghost mark of the very tip of an old fingerboard point which has been filled in with a spruce patch (you'll just have to take my word for it.)

Photo courtesy NMM
What else? Ghost marks of what appear to be the ends of an old capping strip on each side of the bowl; a strip of maple that's been added to the edge rib on the treble side, to add depth to the body; a "new", two-piece capping strip that's been stuck on the bottom of the bowl, almost, it seems, as an afterthought.

Photo courtesy NMM
If it was anybody besides Edlinger who had done this, you'd probably say that some of it is pretty rough work. Indeed, you can say it of him, and if you want to be brutally honest, you can say that he must have used up a lot of spare parts he had kicking around his workshop to put this lute together.

And yet, I don't think in this case that to say those things is an insult. What Edlinger did was, very simply, put together a lute for a certain client who had a certain musical purpose or necessity in mind.  He did what can be called, somewhat generously, a workmanlike job in doing so. Quite obviously, the lute served that purpose well--the marks of it having been played vigorously, and for a long time, are clearly seen.

It must have been a wonderful instrument. And though in its present state it is far from being technically perfect, it is nevertheless very mysterious, and very beautiful.