Advanced Carving: Venice Room (Part 8) Workshops 86-97

The flute of a gouge points somewhere on a clock face, and where you aim it decides whether the wall of a cut stands up straight or leans back underneath. Straight up is twelve o'clock and leaves a vertical wall. Rotate that same tool clockwise and it begins to work under the edge. One idea, and it governs about half of what happens in workshops 86 through 97.

Have you ever finished a leaf, stood back, and found the edge looking soft where you meant it to look sharp? Or the opposite, an edge that reads hard and mechanical when it should have folded? Neither of those is a sharpening problem. Both are the tool pointing at the wrong hour.

The rest of this stretch is stranger, and I like it more. It is about carving things you cannot see while you carve them, because the light in my shop is not the light this piece will ever live under.

What decides whether an edge reads hard or soft

• Twelve o'clock, wall vertical, nothing undercut by accident.
• At two, the very same tool folds an edge back on itself.
• Ten or nine takes it the other way, for the mirror of the same shape.
• A hairline nobody notices under a work lamp can be the thing that carries the piece at night.
• Two lamps, warm and daylight, both dimmable. Move them until the shadow tells you something.
• A design that measured fine on a screen can come out too small in wood. Fix it and carry on.

Where does twelve o'clock come from?

Look down at a gouge with the hollow facing up. That hollow is the flute, and it is the middle of the tool. Held with the flute aimed straight upward you are at twelve o'clock, near enough ninety degrees, and a push there leaves a wall standing vertical out of the ground.

That is what I want wherever a shape must stay crisp. Running a number 11 hard up against a rising edge, I hold twelve and the neighbouring surface survives. Lean it even slightly and the tool eats under the thing I am trying to protect.

Undercutting is not a sin. It has its own moment, and a whole stretch of this room was about it. Doing it by accident, on an edge you were about to need, is the sin.

What happens when you rotate to two o'clock?

The edge starts folding. Take the flute round to two and the same push travels in under the line instead of down beside it, and what your eye reads afterward is a leaf tip curling back rather than a flat shape sitting on a ground.

I do it carefully, because two o'clock removes wood you cannot put back. But it is the difference between an ornament that looks printed on and one that looks like it is coming off the board at you.

And rotating the other way?

Eleven, ten, sometimes nine. Same tool, mirrored, nothing else changes about it. Starting a concave I go one hour left of vertical, feel the corner of the edge take hold, and let it run wherever the wood lets it.

If you are in Europe, nine o'clock is twenty one o'clock. Or if you have been in the military, you already know exactly what I mean. Whichever numbers you grew up with, the point is the same: the flute is a dial, and you are choosing a setting rather than wrestling the tool.

What is a macaroni tool for?

Getting closer to a design than a gouge can. A number 23, the macaroni, is really three chisels in one iron. Three sides, three flat edges, and it drops a trench with square walls in a single push.

Mine goes down to four millimeters, which is the one that lives in the tight separations along a volute. There is a rounder cousin called the fluteroni, and both of them are worth owning if you carve ornament rather than figures.

Neither belongs in a beginner's first roll. Buy sweeps first, learn what a 3, a 7, a 9 and an 11 actually do, and the tools course covers that order properly.

What is a grounding tool, and why is mine stamped 38?

A grounding tool flattens the background around a raised ornament, which is what makes a ground read as a ground instead of a field of tool marks. Mine are Ashley Iles. The small one measures a sixteenth of an inch, 1.5 millimeters, for corners nothing else will enter. The bigger runs six millimeters, near a quarter inch, and it goes under an edge where a little undercutting does no harm.

While we are on numbers, here is the one that trips people up when they buy secondhand. A back bent gouge with a number 8 sweep comes stamped 38 by Ashley Iles: the 3 says back bent, the 8 says the sweep. Pfeil stamps the identical tool 28. Two makers, one shape, two numbers, and neither of them is wrong.

The two millimeter back bend in that sweep rounds the inside of a tight scroll better than anything else I own, and after a few of those it wants stropping. Mirror Paste on a piece of leather, and I use the edge of the leather to get inside the hollow rather than the flat of it.

Why carve a ridge nobody can see?

In my shop, it is invisible. In the room where this piece ends up, it will do the work.

I raised a ridge here so small that under a floodlight it disappears. Play the lamp across it at a low angle and there it is, a hairline of highlight with a thread of dark under it. At night, under a warm lamp, that thread is going to make a huge difference. Not a small one.

Then I did it again inside the next separation, and again in the one after. A micropipe in every separation, is how I think about them. Each is a tiny version of the raised tube that runs down a lobe, and together they turn a flat looking hollow into something with an edge to catch on.

What light should you carve under?

Two lamps, and they matter more than most people's third gouge. Mine each carry two color temperatures, so I can put the bench under a warm evening light or a flat daylight, and dim either one down to almost nothing while still being able to see.

They come at the work from two different positions and both move. When I want to know what a hollow is really doing, I dim one, drop it lower, and watch the shadow lengthen. Nobody has to buy the same lamps. Anything adjustable is a good investment for a carving bench, and it will teach you more about depth than another chisel will.

The habit runs right through this stretch. Rough the shape, then move the light and look again, because shadow is what the viewer is actually reading, not the wood.

What about daylight from a window?

Different problem, same question. At night the lamp is above and the shadow drops. During the day the light comes sideways out of a window, strikes the face of the ornament straight on, and the shadows go shallow.

So a hollow tuned only for the evening lamp goes half flat by lunchtime. Knowing that, I take some areas deeper than the design asks, half an inch, twelve millimeters, so there is still something left to read when the sun is doing the lighting.

Does it matter? Every little bit matters. That is not a slogan and it is not modesty, it is arithmetic: a room full of ornament is a room full of small decisions, and the ones you skip are exactly the ones a visitor feels without being able to name them.

Why leave wood behind for the cleaning?

You are going to need it. On any of these shapes I come very close to the final line and then stop, and what I leave is for the cleaning operation. Always a good idea to leave a little. Six millimeters keeps coming up, about a quarter of an inch, and that is the gap between where the roughing ends and where the surface finally sits.

It is not caution for its own sake. A coarse tool drags. My number 5 was pulling the material along instead of slicing it, and nothing you do with your hands mid stroke will fix that. What fixes it is having wood still there to take off with a number 3 afterward.

Cleaning is also most of what happens in these twelve workshops. Count the words in these recordings and clean beats every other craft word I use, more than two hundred times. Clean the sides, clean around it, clean it a little nicer. Nobody films that for a highlight reel, and it is the difference between a rough shape and a carving.

When does a scraper beat a gouge?

Anywhere a flat must stay flat. A skew laid almost level takes a shaving like a plane and leaves a hard clean corner behind it, and after that a small card scraper tells me the truth about whether the surface really is level.

The other trick in this stretch has nothing to do with steel. When a section stops going well I leave it and start somewhere else on the piece. You get tired of one area, your eye stops seeing it, and an hour later you come back and the mistake is obvious.

Why did the design fail once it was in wood?

It measured fine on a screen. In basswood, at full size, those small elements inside the panel areas were simply too small, and there is no argument to have with that. The wood wins.

I could have redrawn it by hand on the spot. Instead I used what was in my pocket. A paper cutout, photographed with my phone, emailed to myself, opened on the iPad, and then pinched bigger until it looked right against the actual carving. Read the new size off the screen, mark it, carry on.

Originally that area was going to take four pieces. Some of them would have been very hard to reach with any tool I own, so the number changed. A design that cannot be carved is not a design yet.

What happens where two planes meet?

Everything gets harder at once. A section that belongs to two surfaces has grain running in different directions on each of them, so a stroke that slices beautifully on one face tears on the other two inches away.

The proportion rule does not change. Middle line first, set by the golden ratio, exactly as in the twelve workshops before the corbel. On a curved face I set it with three leg proportion calipers instead of by eye, because eyeballing a ratio across two planes is asking a lot.

Then five main divisions, and five more inside each of those. On a piece this big the small ones are what stop it reading as a slab with decoration stuck to it.

Which tools do this work?

Ordinary sweeps, plus three odd ones.

Tool	Size	What it does	Where
Number 11 veiner	Small	Outlines and separates, held at twelve so nothing gets undercut by accident	Every separation on the piece
Number 5	Medium	Fits a shallow profile, though it drags if the edge is tired	Broad curves and shallow ground
Number 7	Half inch	Blends	Between a concave and whatever sits above it
Number 3	Medium	Cleans the sides after a coarser tool has been through	Everywhere
Macaroni, number 23	From 4 mm upward	Drops a trench with three flat walls and gets nearer a line than a gouge	Tight separations
Fluteroni	Small	The rounded version	Where a square trench would look harsh
Skew	Wide	Planes a flat, hard corner along the edge, nothing rounded	Panel faces
Card scraper	Small	Proves a surface is actually flat	After the skew

None of it works blunt. A tired bevel drags where a fresh one slices. Sharpening is the hardest of the three skills, it takes years, and it has a course to itself.

Who is this stretch of the room for?

This is the eighth block of one Venetian project, and the corbel that arrived in workshops 74 to 85 is still on the bench. It started back in workshops 1 through 12 and ran through 13 to 25 and 26 to 37 before that.

If the clock idea is new to you, that is the useful thing to take away and it costs nothing to try this afternoon. Take any scrap, run three grooves with one gouge at twelve, at two and at ten, and look at the three walls under a low light. You will not need me to explain the difference.

Do you want to know why the old acanthus work looks alive in a dim room while a modern copy looks flat? Would you rather understand the light than buy another chisel? Then the whole school is one door, and this room is one course behind it. Shall we carve?

Want a look around first, that is fine. The map of every lesson is open to anybody, the story of the school sits on one page, and the classes in Maine are there if you would rather stand at a bench next to mine. Start here if you have never held a gouge, and the design course is where I send anybody who wants to draw their own ornament first.

One more thing, and it has nothing to do with the clock. On the camera one of these dark hollows is just a black blur and you genuinely cannot see what is happening in there. I kept carving anyway and told you it looked good. It did.