Friday, 3 July 2020

Building a 13 Course Lute, Start to Finish, 18: Making and Applying Varnish

Hi my friends--I hope you're well. It has been a while since I posted, so let's get to it. Today I want to talk about my varnish: what it is, how I make it, and how I apply it to a lute.

A selection of varnish samples on various kinds of woods (honduras rosewood, curly maple, yew, and ash.) I make a varnish sample to go along with every lute that I build.
First, a note of caution. The process of varnish making that I'll be talking about today can be quite hazardous if proper precautions aren't taken. I don't think anyone should attempt to make this varnish without being properly trained. Please don't use this blog post as a 'how-to' for varnish making; it's a description (and not a very complete one), not a set of instructions.

The stuff that I make and use is a terpene resin varnish that's commonly called Fulton Varnish, after the person who formulated it, William Fulton, an aerospace engineer, violinmaker and varnish researcher and experimenter. Fulton published the results of his work in a series of articles in The Strad in the early 1970s; I was introduced to his techniques by Grant Tomlinson in the early 2000s. Grant had been working with Fulton's recipe for years, so by the time I came along he had worked out pretty much all the kinks (and there are a number of kinks that needed working out.)

You may have heard of modern violin makers' search for the 'Secrets of Stradivarius,' the lost methods of the old masters that, if only we could recover them, would allow us to build violins with the magical soul of the old instruments. High on the list of 'secrets' is the oil varnish used by the old Cremonese masters. Of course, varnish alone won't make a great violin (or a great lute), but if it's a good varnish, it will do a few things: be beautiful, in finish, lustre and colour; be stable over a long period of time; and be reasonably simple to make and apply. Fulton's varnish fulfills all of these requirements.

The old makers' varnish is not exactly a 'secret,' in the sense that it's somehow being purposely hoarded or withheld. It's simply a recipe, or more accurately a technique, that's been lost over the centuries, supplanted by the products of the industrial age. Mass-produced finishes took the place of the old and locally-made, even shop-made, varnishes. By the time instrument makers realized that these newer products might not be--ahem--all they were cracked up to be, the old techniques of making oil varnish were mainly lost. Human cultural memory is short.

Many old recipes for making oil varnish survive, but Fulton took seriously the idea that the ingredients named in them might not be quite the same stuff as their modern equivalents. For instance, one of the main ingredients in some of these old recipes is 'turpentine.' When I say that word, you may think of the clear, distilled liquid thinner for paint and varnish. But Fulton realized that the turpentine of 300 or 400 years ago would have been very different: more contaminates and oxygen in the mix, gummier, thicker and browner, more like the sap of the coniferous trees from which it had been derived.

So Fulton decided to take the modern distilled spirit of gum turpentine, and turn it back into the turpentine of 3 or 400 years ago, by oxygenating it (bubbling air through it with an aquarium bubbler) and exposing it to ultraviolet light (in the form of sunlight) over a number of weeks or months. The result of this back-engineering is polymerized terpene resin. Here's what it looks like in a gallon jar:  dark brown in colour, and about as viscous as blackstrap molasses.

Some of the turpentine evaporates during the polymerization process. The piece of masking tape shows the original volume of turps in this batch, so you can see that I'm left with less than half the original volume.
This particular batch of polymerized turpentine was made not by me but by Grant Tomlinson, and had sat high up on a shelf in his workshop for many years. Over the time I've known and worked with Grant he's shown me many techniques of the lute maker's craft, but few as valuable as varnish making. I first assisted him making a batch of varnish in 2005; then in 2010, Grant assisted me in my own first attempt. Last September, when the picture above was taken, I attempted to make Fulton varnish flying solo for the first time.

Now I said above that Fulton's varnish is reasonably simple to make, but that doesn't quite mean that it's easy to make. For one thing, the cooking process creates some pretty noxious fumes, so it must be done outdoors. This, in turn, creates some logistical problems, such as where to cook and when. When I made varnish with Grant in 2005 and 2010, we made it in the parking lot of the building next to our shops. This worked fine until a couple of summers ago when Grant was showing our colleague Wilma Van Berkel how to make varnish. That Sunday, I recall, was a beautiful day, with a slight breeze, which cleared away the fumes quite nicely; but unfortunately they were blown pretty directly toward an apartment building nearby, and partway through the cooking process an alarmed resident came over to see what in the world was going on, and whether she should call the fire department. I figured that this time, I should try to find somewhere more remote to do my cooking.

As for when to cook, since it is done outdoors, a clear day (or at least one without rain) is a necessity. In Vancouver, the end of September pretty much marks the end of decent summer weather; temperatures turn colder after that, and soon the rain--the torrential, unending months of it--begins. The window of time for me to do this cooking was running out fast. If I didn't do it soon, I would have to wait until spring.

I called my friend Bob, who has an acreage in Langley, about an hour out of town, and asked if he would host me. I warned him about the fumes, and he said it would be no problem, he wouldn't mind a bit.

So one Saturday morning in September I woke up, checked the weather forecast, and decided it was time to go. I drove to my workshop, loaded up my equipment, and headed out to Bob's. Here are some highlights of my varnish making day.

The full scene at Bob's beautiful acreage. I'm set up quite far away from a couple of houses on his property. You can see a yellow extension cord running to my work site from the outdoor kitchen a few feet away.

Here is the basic set-up. A milk crate provides a handy box for carrying equipment, and provides me with my seat. On the plywood, there are a hotplate; a cast-iron pot for cooking the resin; a pot lid for checking varnish hardness and colour; a pair of neoprene gloves; a heat diffuser for the hotplate; a lab thermometer (that reads to 400° C); and the jar of polymerized gum turps. Closer to the camera are wet towels, in case of emergency; and a tub of water, the presence of which I'll explain in a second.
On this day, my varnish making is divided into two parts. First, I must heat the polymerized gum turps to a certain temperature in the cast iron pot to make a brittle resin. I begin by pouring the gum turps into the pot, and heating slowly, stirring with the thermometer and keeping an eye on temperature. Around 125°C, a reaction takes place in which some volatile substances in the turps cook off, in the process creating a considerable amount of heat. This exothermic reaction can cause a rapid (and potentially dangerous) rise in temperature, as well as the creation of some heavy noxious gases. The tub of water in the photo above is available in case the reaction gets out of hand: I can lift the pot from the hotplate and dip it into the water to lower the temperature quickly.

In the event, there was very little exothermic reaction created by this particular batch of gum turps. I think one of the reasons why is that the batch was made long ago by Grant Tomlinson (he figured it might have been made twenty years ago), and it had simply mellowed, the volatile substances having dissipated over the years. Another possible reason why there was not much of a reaction is that the batch was not made using a certain chemical--manganese napthenate, to be specific--which Fulton recommended using as a catalyst in the original polymerization process.

Once past that rather tense phase of cooking, I could pretty confidently take the temperature of the gum turps slowly higher, ultimately somewhere above 300° C, the temperature needed for making a beautiful amber-brown shade of resin. (Lower terminal temperatures are used to create lighter-coloured resins and varnishes.)

Here I am, your humble narrator, enjoying the cooking process on a lovely late summer day.

This photo is courtesy of Bob, my host, who took a break from driving his tractor and cutting the grass while I made varnish. 

Here is the result of the cooking the resin, after removing from heat and allowing to cool. Although it looks like a liquid, it is a solid--I've converted the gum turps to a brittle resin.

Now I just need to remove it from the pot, which I do by breaking it up with a hammer and screwdriver. I will use some of this brittle resin for the second stage of my cooking day, where I actually cook varnish. The rest of this brittle resin I will put into a jar and keep until I need to make varnish next--which hopefully won't be for a few years. (The brittle resin will keep indefinitely.)

The varnish-making part of the day can now begin. I crush a certain amount of the brittle resin in a mortar, and mix it in another pot with stand oil. Then I heat this mixture steadily, over the course of about an hour, to around 300°C.
I don't have any photos of this part of the process, because I was busy doing it and paying pretty careful attention, and Bob had left the area. But here's what happened. I sat beside the hot plate, stirring constantly with the thermometer, taking temperature readings every minute and jotting them down in my notebook. Even small variations in the rate of heating, as well as the amount of time the varnish is held at the terminal temperature, can make differences in the quality of the varnish. I keep careful records of the cooking process, in the hope of tracking down the origins of these subtle differences.

Fulton's recipe calls for holding the varnish at the terminal temperature 'until a firm pill stage is reached.' I'm not exactly sure what he means by that, but here's what Grant taught me to do: take the thermometer from the pot and let a drip of varnish fall onto the cold pot lid. Let the drip cool for a few seconds, then tap it with your fingertip and lift away. The varnish will stick, and you'll pull away strands of beautiful amber varnish that billow and drift in the breeze. The longer you hold at the terminal temperature, the finer and longer these gossamer strands will become.

The Sorcerer's Pot-Lid. Who can tell the varnishes it has known!
And then, when you judge that the varnish can become no finer, you take it off the heat and let it cool. You can't keep heating indefinitely; your varnish will gel in the pot. That's about it. When the varnish has cooled to below 150°C, you can add some distilled spirit of gum turps as a thinner, and then filter into jars. Here is the final product.

It's lovely stuff. Can't wait to make a lute, and slap on a coat or two.

Which leads us to the next part of today's post: applying the varnish.

For this part of the process I decided to make a series of short videos on my phone. This marks a first for The Lute's Progress: I've never done a video demo before. I don't want to get your hopes up or anything--they aren't the highest of fidelities, I'm not used to talking to a camera, and my framing of the shots may leave a little to be desired. (For instance, in none of the videos can you see my face clearly. That may be either a glitch or a bonus feature; I'll leave it to you to decide.)

But I couldn't think of a better way to illustrate what I was doing in flattening off the previous varnish coat, and then brushing on the next one. It's not a process that lends itself to taking a bunch of photos, as I usually do, then writing descriptions. So, with all their faults, I humbly present these videos and hope that you can get a reasonably clear idea of what I'm up to when I apply a coat of oil varnish.

And that will be that for this time. Next time, I'll tell you about making the peg box. Cheers!

Video #1: Removing dust particles

Video #2: Flattening off the varnish

Video #3: Brushing varnish on the first few ribs

Video #4: Brushing varnish on the last ribs and capping strip

Sunday, 3 May 2020

Building a 13 Course Lute, Start to Finish, 17: Finishing out, Prep for Varnish

Hi again, and welcome back to the workshop. How are you doing during the Great Hush of 2020?  I'm still here making my lutes; and I hope you're keeping well, wherever you are.

Let's get to it. I've got a bunch of things to do to prep this 13 course lute-in-progress for varnish, including some preparatory seal- and ground-coats to the body and belly. First, however, I need to finish out the instrument.

But what does 'finishing out' really mean? To me, it means going over the entire lute, very methodically, to give it a complete and final physical shape. I've got a checklist of tasks that I go through from top to bottom, which bring the lute to a certain point of completion. This checklist ensures two things: first, that I don't miss out on any details in shaping the lute; and second, that the process doesn't go on forever. I have a task; I do it; I cross it off the list; I go on to the next. Magically, it seems, the lute gets done.

A small item to begin. Remember back in episode 3 of this series, when I bent and glued the ebony veneer onto the neck core? I used two pins to locate the veneer when I glued it, one near the nut end, and one in the middle of the neck. Those two holes are still in the veneer. The one at the nut end will eventually be obliterated by the peg box rebate, so I don't need to worry about that one. The one in the middle, however, needs to be filled.

With a jeweller's pin vise, I drill the hole to a slightly larger diameter.

Then, with files and a scraper, I make a dowel from a scrap of ebony that fits the hole. I glue it in, and when the glue's dry, I nip it off, and file it flush.
Next up: rectify the body-neck joint. My result was pretty good when I veneered the neck and glued it to the body, but you can see  here a small area where the match isn't quite exact. I'll make it exact with files; then I'll go on with the files to give a good shape to the back, starting with the top block area adjacent to the body-neck joint.

Then I move on to the rest of the back, and give the whole thing a final shape. This part of the process probably takes the longest of all the steps I'm describing in this post. I work with files and some small, curved scrapers, using strong side lighting to see how tight the rib lines are. You can see in about the centre of this photo that the rib above sits a little higher than the one below; I will file and scrape in that area until the facets of the ribs, and the line between them, are rectified. I then do the same thing with all other problem areas, all over the back of the lute.

I was going to say "I will file and scrape in that area until the facets of the ribs, and line between them, are perfect," but that's not exactly correct. Perfection is an ideal state, and therefore unattainable; I merely work until I have achieved a satisfactory result. I look at the lute in all sorts of different lighting situations; I set it aside and do some other work for a while, so that I may come back to it with fresh eyes. Each time I come back to it, I find a new spot to work on. Until I don't!--and then it's time to check this task off the list, and move onto the next one.

Here's a good shot showing how I brace the lute when I'm scraping the back--between my knees and the well-padded edge of the workbench.
I have made a few scrapers for myself with different curvatures. I  match the scraper I use to the width and curvature of the rib I'm working with.
From there, I move onto finishing out the neck. I mask off the top of the body (don't want ebony dust to get into the freshly-scraped back), and work with a series of files.
When I get a good overall look with my finest file, I move onto burnishing with shave grass. In this photo, you can see, at bottom, a segment of shave grass, and in the middle, a segment of shave grass that I've slit open, flattened, and backed with a piece of masking tape. I can then use this flattened piece with a cork sanding block (seen at the top), to get a very nice burnished surface on the neck veneer. 
Just a word about the scraping and burnishing process. You may have noticed that I have not mentioned the word "sandpaper" here, and that is because in finishing out this lute (and the other two in this group), I resolved not to use that material. In the past, I have finished out the veneers and backs of my lutes with a series of sandpapers of ever-finer grits, but I decided with this group--and, I hope, from now on in my lute making practice as a whole--not to use it to finish out necks and backs. (I have never used sandpaper on the bellies of my lutes.) I decided to do this for two reasons. First, over the years I have developed sensitivities to wood dust, especially ebony and rosewood. Sandpaper creates a lot of dust, so by eliminating it from the finishing-out routine, I hope to significantly reduce my allergic/asthmatic sensitivities. Second, I have long believed that while sandpaper offers a certain convenience in finishing out, it also tends to give an instrument a kind of 'rounded over' look overall, diminishing the crispness in texture that I would prefer to see in surfaces and edges. I'm not talking about creating sharp edges here; quite the contrary, in the next few photos I'll describe how I methodically eliminate any sharp edges that might interfere with the comfortable playing of this instrument. In this context, I am referring only to the visual aspect of the lute, how it catches the light, and the eye of the beholder, to create a bold visual interest.

So for the back and neck veneers of this lute, the finishing out process is as follows: file, scrape, and burnish with shave grass. That's it.

The biggest and sharpest edge on the lute, at the moment, is the edge of the fingerboard. I want to ease the fingerboard edges and create a radius in an orderly way not only so that it's comfortable to handle, but also so that gut frets will tie more easily and lie flat against the fingerboard. (A too-sharp fingerboard edge will make the frets sit up off the fingerboard, which will interfere with the sound of the first one or two courses.)

I use a file to chamfer a 45˚ surface about 1-2mm wide on the edge of the fingerboard. Then I use the file to relieve these two new edges I've created, until the fingerboard edge is smooth and well-rounded.

I then carry on to the belly edge-binding. Again, I use a fine file to make a 45˚ chamfer that goes about halfway down through the thickness of the binding.

Then I round those two newly-created edges over, until I've got a very friendly-feeling edge. Remember, the bottom of this belly edge will be resting on the player's leg, and the player's arm will rest against the upper bottom corner of the belly. The "softness" of these edges is, in my opinion, a crucial part of making this a very playable instrument.
Where these two chamfered edges meet, I ease off the rounding over, and try to match them simply and crisply.

Finishing out the belly is fairly straightforward: I scrape it carefully with a small, sharp, flat scraper along the grain, in whatever direction works best...

And then I burnish with shave grass. This stuff works best on the spruce if it's rubbed across, but held slightly askew, the grain.
At the end of the day, I wet the belly with distilled water, to raise the grain. I let this dry overnight, and next morning scrape and burnish the belly one more time.

I then seal the rose with a brushed-on coat of Tried and True oil. Then I hang the lute in the UV light box for a couple of days, to let it tan up a bit, before I begin applying ground coats before varnishing.
Two days later....

I bring the group of lutes out of the light box, and make ready to apply the first seal coats to the belly and back. The first thing I need to do is mask of the neck and fingerboard carefully.

Then I mix up my sealer: a dilute mixture of casein glue. I use about a tablespoon of low-fat dry curd cottage cheese, a small amount of artist's slaked lime, and distilled water.

Stirring the slaked lime to prepare for use. 

Dry curd cottage cheese.

I grind the cottage cheese first to minimize lumps, then mix in the slaked lime. Finally, I dilute with distilled water.

Then I begin padding it on the front of the lute with a soft cotton cloth, starting out above the bridge, and working around the rose to the top of the belly, and around and down the belly below the bridge.

Around the bridge, I apply the size with a small brush. I'm a little sparing in my application of the size around this area.

I keep applying the size for a few minutes until I have a nice, even coat. Then I take a fresh cotton cloth and wipe the belly down, along the grain, to remove any excess specks of cheese that may not have been ground down well in the mortar.
Next step is to apply the casein size to the back of the lute. This time I apply it with a wide, soft brush, and brush along the length of the rib. (Notice also in this picture that I have swapped out my green cloth bench pad for a large piece of card stock. During this stage, and during varnishing, I handle the lute by holding onto the neck with my left hand, and allowing the edge of the lute to rest on the card.)

At this time I also apply my size/seal coat to a sample of the same material as the back of the lute. From now on, whatever I apply to the lute, I apply to this varnish sample.

When I've applied an even coat, I wipe down the back with a fresh cloth to remove any large bits of unmixed material. 
And--back in the light box for another couple of days, to let the seal coat harden up.

You might be asking: why apply a seal coat at all? Well, there are a number of reasons. First, a coat of dilute glue such as this will consolidate the wood fibres of the belly and back, and create a flatter surface on which to apply layers of varnish. Second, the seal coat will help with adhesion of varnish coats. And third, and most important, the seal coat will actually create a barrier between the wood and the varnish coat, which in the case of rosewood (which this back is made from) and yew wood (which another lute in this batch is made from), is essential. Both rosewood and yew have oils in them that, unless the woods are sealed off somehow, will interfere with the bonding and curing of varnish. For years I used a coat of shellac as a sealer for rosewood and yew, but that stuff is problematic for a couple of reasons (for one, it's a bit anachronistic to be using in an 'early' instrument such as this.) Grant Tomlinson introduced me to casein, and I've used it for a few years now. It works well, and it is easy to apply.

Okay--one more coat, this time a ground coat, the purposes of which are to penetrate and strengthen the wood of the belly and back, to give added depth to the finish, and to promote adhesion of the coming varnish layers. The three ingredients: marienglas, Tried and True oil, and my own oil varnish.

The three ingredients are measured out on a glass plate.

I mix them together with a muller.

I dilute the mixture with spirit of gum turpentine...

And then apply it in a thin layer to the belly with a cotton pad, in much the same way that I applied the casein size.

I apply this coat to the bridge...

As well as to the rose (I use a very dry brush in doing so.)

As a final touch, to even out the coat, I brush gently but quickly across the grain with a broad, soft brush. (By the way: this is the final coat of finish I put on the belly of my lutes.)
I apply a coat to the back of the lute, using a cotton pad.

I love the way the lute starts to look at this stage--I can already begin to sense the glow of the wood below just these preparatory layers. I'm quite excited to put a few coats of varnish on this lute!

There is my batch: from left to right, the 13 course in rosewood, a 7 course in yew, and an 8 course in rosewood.

Next time, I will tell you about my varnish, and I will show you how I varnish a lute back. There might even be moving pictures....

Stay well, and I'll see you next time.

Saturday, 18 April 2020

Building a 13 Course Lute, Start to Finish, 16: Shaping the Fingerboard and Setting the Action

Hi friends, and welcome back to the series. Hope you're all well, and taking advantage of the downtime caused by the Covid-19 quarantine to make the art you've always wanted but never had the time to do before. Or tidy up the house. Or do your taxes. Or read a good post-apocalyptic novel.

Here in Vancouver, it has never been a better time to ride a bicycle. The air is clear, birds are singing and the streets are snowed under with cherry blossoms. It is a strange--and in some ways, strangely good--time to be alive.

But I want to go back today to a former time, a time before the Great Hush of 2020. If you recall, I had just glued a curved fingerboard on the 13 course lute I'm building.

I've taken off the clamps, and now--a small problem. I've glued the fingerboard to the neck--but I've also managed to glue the caul to the fingerboard!

I'm pretty sure that the caul is stuck only along the edges, so as I use strips of paper towel to soften the squeezed out glue along the joint, I also try to soften it along the edges of the caul.
Eventually, I get a little impatient and gently tap some wooden wedges in from the sides. The caul comes up with a bit of damage, but it's only to the cork lining, which I can easily replace next time I use it. (Apparently, I placed a gluey thumb down on the fingerboard before placing the caul.) Note to self: cover the caul with plastic wrap!
Okay, the fingerboard edge is cleared of glue. You can see the overhang I have left to remove.
Here's a close-up of the joint. Not bad!
I make a start shaping the fingerboard down at the body-neck joint. I need to plane (very carefully) and then scrape to get the fingerboard close to flush with the belly tongue.

Getting close...
There we are. I don't use any sandpaper around this area, only scrapers from now on.
 I begin to trim the overhanging edges of the fingerboard with a standard block plane.  
The idea here is to get the edge pretty flat along its length, and pretty close to perpendicular to the mid-section of the neck (that is, to the glued line between the mahogany neck core and spruce fingerboard spacer.) Notice how I've got my left thumb hooked over the edge of the neck down by the body, to guard against accidentally whacking it on the recoil stroke of the plane.

Down by the body-neck joint, I set aside the block plane and work carefully back toward the body with a chisel (a skew chisel works well for this job.)
For the opposite side of the neck, drawing the plane toward me is the way to go.
The plane and chisel get me close, but eventually I need to start working with files.
A slight boo-boo with the file: I've chipped out a little bit of the fingerboard edge. This particular piece of ebony has quite swirly, interlocking grain, which is very tough, but also tends to chip and tear out. I need to fix this little problem before I can go on.
I add one drop of cyanoacrylate glue to a little pile of ebony dust that I've made with a file and a scrap of ebony.
I apply it like putty with a palette knife, and wait for it to dry thoroughly.
Then I can go ahead and shape the area with the file. 
Using a straightedge with a strong back light shows me how close to flat the edge is (I like to get it as flat as I can.)
On smaller areas, the file works best...
But over larger areas, a sanding block is the way to go.
That's my edge. I don't know if it's visible here, but I leave the fingerboard just the tiniest bit wider than the edge of the neck. I'll be rounding over the fingerboard edge very soon, and will use that extra half-millimetre or so to blend the underside of the fingerboard into the neck.
Right now, I try to keep the joints and edges square and tight. When I finish out the lute to get it ready for varnishing, I will take an overall approach to the whole lute, rounding over the fingerboard edges, half-binding edges, and so on, hopefully in a uniform and coherent way.
First, though, I have to finalize the action. I've already scraped flush the body-neck area, so I now need to focus on the fingerboard from about the end of the belly tongue to the end of the neck. I'm using a Japanese scraper-plane here, pulling it away from the belly.
This swirly ebony tears out so easily that at a certain point I needed to put away edge tools, and just use sanding blocks. I try to avoid using sandpaper in lute making for both aesthetic and health reasons; but sometimes, with difficult material like this, I just have to surrender to the modern woodworking world.

As usual, to check the action I have pieces of fishing line tied to each course on the bridge; I pull each one tight over a spacer at the first fret position, and then read the height at the eighth fret by slipping wooden spacer pieces under the string. The spacer used at the first fret, by the way, is 1.6mm, which represents the nominal height of the first tied fret (1.1mm) plus the necessary clearance for the string not to buzz (0.5mm).
One other thing: in the photo above, note that I have masked off the upper part of the belly with a couple of pieces of masking tape, to keep ebony dust from contaminating it.

I feel I must pause here in order to try to convey to you the exquisite boredom of the work that I'm about to describe. I don't want to rush over this part, and give the mistaken impression that it's a quick job, Bob's your uncle, and we're off to the next stage. No. This job, shaping the baroque lute fingerboard to set the action before varnishing, is, to me, the slowest, most tedious, most excruciatingly boring bit of business in all of lute making. Here is the sequence:

1. Scrape, or shape with sanding block, all, or some, or small, strategic areas of the fingerboard, in order to lower the action at the nut end and/or the body end of the fingerboard, as appropriate.
2. Carefully vacuum all ebony shavings and dust from the work area. Then wipe the fingerboard (from the body toward the nut end) with a piece of paper towel, to remove the finest dust particles.
3. Re-draw the trajectories of each of the courses as they cross the fingerboard from the body to the nut. Re-draw the positions of the 1st and 8th fret.
4. Pull each piece of fishing line for each course tightly over the 1.6 spacer at the first fret, and slip spacers of various thicknesses under the fishing line at the 8th fret, to gauge the specific string heights of each of the courses. Write these numbers down beside the 8th fret mark.
5. Take the straight edge in hand, and with a strong backing light, hold it on the fingerboard exactly on the marked position of each of the courses and sight underneath it to see where the fingerboard is high, or low. Mark the high spots.
6. Take a moment to consider a strategy in the next bout of fingerboard shaping. How far do I need to go, and in what areas? Is there a discernible pattern--for instance, is the whole central area of the fingerboard high (or low?) Can I work in a larger area, or must I restrict myself to the line of only one course?
7. Begin again at 1. Continue repeating this sequence as necessary, until the action is set, and the fingerboard well-shaped (that is, smoothly curved in cross-section, but very flat in long section.)

Sound boring enough? I don't think so. I'd like to make it more boring, but I don't think I can. You'll just have to trust me on this (or, better yet, try it yourself.)

Here's a bit of an illustration, from rather late in the process. The squiggly lines are the high areas; the numbers are, of course, the string heights at the 8th fret.
Getting the final action is a job that can take an entire day, or more, depending on how well I am concentrating. If I can, I like to bite the bullet and get it done quickly. Ebony dust is bad for my health--I have asthma, and it's become a bad trigger. I wear a face mask, long sleeved clothing, and a cloth tied over my (short) hair. Periodically I need to go to the washroom and wash my face and hands. At the end of the day, I remove all work clothing, put it in a bag, change into street clothes in an outer room, and leave, pausing only to wash up carefully once more before getting on my bike and riding home.

These are my final action numbers for a baroque lute measured at the eighth fret. 
Just a word on that--these are 'final' action numbers before varnishing only. I fully expect to have to re-shape the fingerboard at least one more time before the lute is done. Why? Well, over about the next 3 to 4 weeks, this lute is going to spend quite a bit of time in the UV light box, a warm environment; and even though I'll be running a humidifier in the vicinity, the neck, and fingerboard, will continue to subtly change shape. Not too, much, I hope... But I think I've left myself enough leeway in setting the action thus far, that I will have the flexibility to make some small corrections later on.

By the way, the distance in time between the earlier photo and this one--showing string height differences of 0.1mm--is probably a couple of hours. As I said, very slow work, for very small gains.

Okay. Enough boredom for one blog post. I'll see you all next time, for the finishing out-stage--and perhaps some foundation coats to prepare for varnishing. Stay well, and I hope to see you back here again soon.