Advanced Carving: Venice Room (Part 14) Workshops 156-163

Undercutting is the one thing on a carving you have to ration. Every extra millimetre of shadow is labour the design never asked for, and on gilded work the gold cannot reach in there anyway. Run the separation uphill instead, and the leaf reads as though it were undercut far deeper than it is.

You have been told that deep undercutting is what separates good carving from ordinary carving. So you spend an hour going in under every edge, and the piece looks about the same as it did before, only now you are behind.

Do you know how much of your undercut anybody will ever see? Have you looked at your own work from the chair it will actually be seen from? I love undercutting. On this room I kept talking myself out of it.

These are the last eight workshops of the Venice Room, numbers 156 through 163. The ornament wraps the molding, the shell closes, and somewhere in the middle of it I was a sixteenth out and had to say so out loud.

What decides how far in under an edge you go

• Point of view. An undercut nobody can stand in front of is labour spent on nothing.
• Gilding. Gold will not lie down inside a deep hollow, which is why the old work is shallower than you expect.
• A separation that climbs uphill under the next part, doing the same job for free.
• The bench. Go too far and your tool is into your own workbench.
• What the wood behind the edge still has to give, because an edge with nothing behind it has no support.

Why would a carver undercut less on purpose?

Labour comes first, and gold comes second. As much as I love undercutting, going deeper than the design needs is a big mistake, because all it buys is labour the room will never show. On a room this size that adds up. The second reason is older and better, and it comes from the way this ornament was meant to be finished.

Point of view settles most of it. On these last movements the panel gets looked at from one side of a room, so I made every part pass under its neighbour on that side. My undercut is then in the place nobody stands, and if it is not perfect, no human eye is going to catch it. Clean where it shows. Adequate where it does not.

How do you make a shallow undercut read deep?

Run the separation uphill so it disappears under the part above it. Instead of driving a tool horizontally in under an edge, you carve the leaf so the separation climbs as it goes, and the piece above swallows the end of it. The shadow arrives without the labour.

One of the old techniques, and they kept it because it looks beautiful and it simplifies the job of undercutting at the same time. Two birds. The old carvers did not worry about undercutting nearly as much as a modern carver does, and once you know what they were finishing the work with, you understand why.

Where did this leaf come from?

Greek culture first. The Romans took the idea of the acanthus leaf from the Greeks and made it their own, and the Corinthian capital is where most people meet it. Then the Roman Empire divided, and the eastern half, Byzantium, was a Christian country. They took the leaf from Rome.

The Greek leaf, the Roman one and the Byzantine one each have their own courses in the school, and they are three genuinely different leaves rather than three names for one. Carve them in that order and the borrowing is visible in your own hands.

Here is the part that changes how you work. Romans did not use gold leaf. Byzantium did.

Why did the Byzantines gild carving at all?

Scripture, the way I read it. They put acanthus leaves in their churches and they wanted the carving gilded, and the idea did not come from Rome, because Rome had not done it. Go a long way further back. Solomon built a temple, and Moses built a tabernacle and the Ark of the Covenant, and all of it was covered with carving, and all of the carving was covered with gold.

So Byzantium took the idea from Solomon and Moses and started gilding carved work. That is my reading of it, and I say so because it matters at the bench: a carver who knows the piece is going to be gilded designs the leaf differently. They kept the undercuts simple so the gold could get in underneath. Really, really smart, what they did. My opinion.

What is actually under the gold?

Red clay. Ordinary earth red clay, not gesso, which is what a modern gilder would reach for. They laid the clay first, then the gold over it. And to get in there and make the ground smooth they could not have gone crazy underneath, which is the whole point about the undercut coming back around again.

Go to modern Turkey and walk into some of the old churches that have been kept, and you can see it for yourself. Gilding still on parts of the carving, and inside the cracks in it, red clay showing through. The technique is still sitting there in the cracks, centuries after the hand that put it on.

The decision	What the gilding wanted from it	How deep	What it costs you to do it the other way
Undercut a leaf	shallow enough for gold to reach in	only a few millimetres	hours of labour on a shadow nobody ever sees
A wood allowance	against the highest point of the molding	about 6 mm left	concaves you can no longer carve into it
Run a separation	uphill, hidden by the part above	1 mm at the top	a straight horizontal reach in with a tool
Lay the ground	red clay laid on under the gold	very thin indeed	gesso, which is a different craft entirely

How accurate does a molding have to be?

A sixteenth of an inch, about one and a half millimetres, is already too much. On ornament that error hides inside a curve and nobody finds it. On a molding it runs in a straight line beside a straight line, and the eye reads the difference instantly. That would be an unforgivable mistake. My words, on camera, about my own work.

Because I made it. I marked a line for the convex on one side, marked its partner on the other, and something looked off to me. Not to the camera. To me. The way I mark that line is with the tool itself, laying the big chisel against the wall of the cut and pressing hard enough to leave a mark I can find later. I checked with the tool, then with calipers, and one side was right on the money and the other was a sixteenth out. The line I placed was no bueno.

Then the luck. I had gone wrong in the direction that left wood on, not off, so the fix was to take a little more away rather than to glue anything back. If it had gone the other way I would have been looking at a repair on a molding that is already low. Which is the argument for checking three times instead of twice, and I still hate making mistakes.

What checks a shape when you have no gauge?

With a finger. Once a convex is close, run a fingertip along it and a finger can sometimes detect a thousandth of an inch, which is far past what your eye will do under workshop light. I had cleaned one of these convexes and it looked right, and my finger found waviness the light had hidden.

For a right angle, any square will do. Hold it against the wall of the cut and look. Ninety degrees on a molding is not a preference, and eyeballing it is the thing that gets people, because a wall leaning in at all reads as a mistake the length of the run. Mine was dead on 90 when I checked, and I checked because I was not sure.

Late on, I thought the molding was finished, then put the pointing machine back on it and found the height was not where it should be. Another millimetre off the top, and the two sides match now.

Which tools are allowed on a period reproduction?

Only the ones that existed. A macaroni tool would have made one of these cuts far easier and quicker, and I stood there wanting to use it, and I did not, because that tool was not available yet back then. Using it would be cheating on a room that is meant to be a 15th century room.

What I used instead was a 1.5 mm veiner, run down each side of the future bead, then a 2 mm to open it out, then a flat back bend to level the bottom. Slower. Three tools where one would have done. The cut looks the same as the macaroni would have made it, which is exactly the point.

My own line is not pure, and I will say that plainly. Rasps went over the shell work on this room, because the whole thing is paint grade and a rasped surface under paint is a surface nobody sees. A fluteroni got used earlier in this project, where the mark would be hidden. The rule I actually follow is that a modern tool can save me time where the result is identical and invisible, and it cannot change the shape of a period profile.

Can you undercut to a knife edge without snapping it?

Work it from the back. A knife edge worked from the front has nothing behind it while you are still pushing on it. So I rough the form from the front, then turn the piece over. Finishing from the reverse face means the wood I have not removed yet is still bracing the cut.

Depth has two floors under it, and the second one surprises people. The first is the design. The second is my workbench, and on the low sections of this relief I was close enough to it that going further would have put the tool into the bench. Deep, but not too deep. I said that out loud several times in these eight workshops and I meant it as arithmetic rather than caution.

Leave yourself a working allowance too. Against the highest point of the molding I kept about a quarter inch, 6 mm, of material because the concaves and convexes still had to be carved into that section, and a shape carved into wood you already removed is a shape you cannot have. One spot opened up a crack while I was in there. I took the cracked wood out and kept going, and the volute beside it never knew.

Where does the waste come off near an edge?

From the outside inward, every time. Working outward toward a free edge lets the tool catch the corner, and then the grain decides what happens rather than you, and a chip below your line is not fixable by taking more wood off. So I come in from the outside and lose nothing.

Across the grain for the bulk of it, with a number 9 at 20 mm, three quarters of an inch. Across is safer and it gives more control than running along the fibres. And stay above the final line, always. I want to be safe rather than sorry, which sounds obvious until you are tired and the deadline is close.

One place on this panel I drilled straight through on purpose and then worked toward the hole, which is far easier than excavating blind from one face. A plug pass from each side with a number 9, then the drill, then out with the plug, then the hole gets opened little by little.

Why does a drawing look wrong until it is carved?

Flat is a lie about a form. Looking at just the drawing, this section did not look appealing at all, and I have had that argument with myself on paper more than once. Establish the form and it starts to make sense, because half of what makes an ornament read is depth the paper cannot show you.

Which is why I redraw constantly rather than drawing once well. When I am drawing I have to keep in mind what the final thing will look like, and the final look is often not the same as the drawing. Design and carving are one job done in two passes, not two subjects.

The pipes finish the same way they always do here. One concave down each side, deep near the volute and fading to nothing, and the wood left standing between them becomes the pipe. Three main divisions in the head of the leaf, and I left the subdivisions off. Sometimes that is overkill and the piece does not need it.

My cameras cost me some of the good passes, and I should admit that too. At one point I worked toward myself, which is not the way to do it, because turning the piece round would have put my back to the lens. My clean side ended up on the right and my rough side on the left, when I wanted the opposite. There were landscapers working outside that day as well, and you can hear them. Somewhere in there I also swapped to a number two for no reason except that I get tired of reaching for the same tool all day.

The flowers on this last section I did off camera, and I said why: the same flowers are already shown, pass by pass, in another project. If flowers are what you came for, watch the Hashemi flowers instead. That is where this room's ornament crosses back into the rest of the school, along with acanthus over moldings, the Corinthian leaf and the tools it all runs on.

Fourteen parts, 163 workshops, and one room copied out of Venice. If you want to see it from the beginning, part one is where the first line went on the wood, and part thirteen is the molding this one wraps. All of it opens with tuition to the school. Never carved a leaf before? Then start much smaller and come back to this room when your hands are ready.

Shall we carve?