Advanced Carving: Venice Room (Part 7) Workshops 74-85

A corbel is a carved bracket that stands out from a wall and carries whatever sits above it, and carving one means shaping a solid block on every side instead of working into a flat board. Mine runs 42 inches long and nine inches thick. There are two of them. Everything on this page exists because of those numbers.

Workshops 74 through 85 are where the Venice Room stops being a panel. Up to here I had one face to worry about and a bench that held it flat. Now the shape carries on round the back, the blank is heavy, and half the job turns out to be measuring, holding and sawing rather than carving at all.

Have you ever been stopped by the practical side of a piece rather than by the carving? Not knowing how to clamp it, or how deep to go, or what to do after something snaps off in your hand? Those questions get answered here, and I am going to answer them the way they actually happen, mistakes included.

What decides whether a piece this size ever gets finished

• A template printed full size and glued to 5 mm plywood beats any amount of measuring by eye.
• A pointing machine reaches where calipers cannot, reading off the flat bench top.
• Bore to the depth you want, then carve down until you meet the bottom of your own hole and stop there.
• The blank rides on a three inch plate and long screws. The plate is the whole holding system.
• Stock two inches thick goes through a scroll saw on a very fine blade.
• Thin CA glue soaks right in and leaves a mend harder than the wood beside it.

What makes a corbel harder than a panel?

On a relief panel every decision happens on one plane. You look down, you see the whole composition, and the back of the board is just a back. The corbel gives you no such comfort. The drop rolls over the front, keeps going, and dies somewhere on the rear face where the camera never goes.

So I have to build a three dimensional form, and that is the challenge of the whole stretch. Two of them, matching, on a piece too big for the bandsaw. I marked the waste on front and back, then on both sides, and only then picked up a gouge.

The proportion rules from the last twelve workshops still apply. What changes is that a line now has to look right from three directions at once, and moving it to please one of them can ruin the other two.

How does a full size shape get onto a nine inch block?

On paper, one to one. I print the profile at actual size, glue it to plywood about 3/16 of an inch thick, five millimeters, and cut it out. That gives a stiff template I can stand against the blank all day without it curling or tearing.

For the second corbel I print more copies of the same sheet and lay them straight onto the wood to mark from. Cheap, fast, and the two pieces come out as a pair rather than as cousins.

Then calipers, over and over. I glued the template for the opposite corbel by mistake, which does not really matter, because I can reference either one and measure across. Accuracy here is not fussiness. Get a measurement wrong and the wood is gone before you notice.

When do calipers stop working?

When the point you need sits inside a hollow, or behind something, or on the far face. Calipers want two jaws and clear air between them, and a corbel stops giving you that fairly early.

That is where the pointing machine earns its place. A sculptor's tool, a needle on an adjustable arm, and I like it a great deal. Once I have the depth of a point taken from the back, the piece lies flat on the bench and I measure to that point from the bench surface itself.

The needle also swings, so I can come at one spot from several angles instead of forcing the arm into a place it will not go. Set it flat, take the reading, move it, take another.

Why bore a hole before carving the waste off?

The hole tells you when to stop, and nothing else on a piece this deep will. I measure the depth I want, wrap tape round the drill bit at that mark, and bore straight down. After that I chop out the waste until I meet the bottom of my own hole.

An inch and a half in one spot. Five eighths more in another, about 16 millimeters. The tape is only a rough guide, since my reference line sits lower than the high ground beside it, so I stop early and check rather than trusting the mark.

On one of those holes I came within a whisker of going too far. Almost messed up, and almost does not count. That is the whole reason for drilling short and checking twice: a hole an eighth too deep is a hole you cannot un-drill.

Holding something this heavy while you carve it

A three inch plate, screwed to the back of the blank with two and a half inch screws, dropped into a Veritas carver's vice. That plate is the whole system. I keep several of them, two inch and three inch, already mounted on different pieces, so swapping work takes a minute rather than an afternoon.

The vice itself sits on a block of wood, and that block is clamped to a sheet of plywood, and the plywood is fastened to what I have to admit is not a real bench. Behind me it is drawers. The whole tower wobbles a bit and I do not care, because I can spin the piece to any angle without walking round it, and the result is what gets judged.

There is a proper Veritas carver's bench for this. Out of stock everywhere, about $1,000 for the bench alone, and it would take floor space I have not got. My shop is not a shop any more. It is a studio now, and I have to be careful about what goes in it.

Inside the small clamp I glued thick leather, using the clear Gorilla Glue rather than the ordinary one, since the ordinary one foams and the clear one does not. Now a finished surface can go in the jaws and come out unmarked.

Will a scroll saw go through two inch stock?

Yes. Somebody asked me exactly that, and this stretch is where I prove it. Eight quarter material, two inches thick, on a tiny skip tooth blade, and the saw never complained.

A bandsaw would be quicker if you own one. My scroll saw is an Excalibur General International, an older model made in Canada rather than the ones they build now, and I love that machine. Powerful, and no vibration at all.

It changes how much gouge work there is. Cheeks either side of a middle leaf can be chopped with a chisel, or the piece can go flat on the saw table and the waste falls off in seconds. I switched to a spiral blade for one curve, which I normally avoid, and it did what I needed.

Why leave extra wood at both ends?

Waste at each end becomes a handle. Screw one, clamp the other, flip the whole thing over and keep going without ever touching the carved face. When both are gone the piece is finished and there is nothing left to grip.

I try to keep metal out of harm's way too. A tool that slips into a clamp jaw is a ruined edge, so I set the work lower than the hardware, and gluing small blocks of wood over the exposed metal would be smarter still.

Did Gibbons really build his work in layers?

I could have carved this whole area out of the solid, the way the rest of the room was done. I left it for the school on purpose, because I wanted to show the method Grinling Gibbons used: separate layers, carved apart, then fitted and glued into place.

Here is the honest part. How do I know that is what he did? I do not know. I am reading books, and the people who wrote those books can only guess as well. Nobody watched him work. What I can tell you is that the method holds up under a chisel today, and that a second layer gives depth no single block will give you.

The stem I carve here would stand on its own as a project. The wood here is basswood, though the same piece would sit happily on an oak ground. No flowers on it, just leafage, and the flowers elsewhere on the body go in as inserts dropped into a carved socket, the same trick as the Hashemi flowers.

What do you do when a piece snaps off?

Glue it back and carry on. I broke a sliver here under a sixteenth of an inch thick, knocked off while I was unscrewing something, and it went straight back on.

The glue matters more than people expect. Mine is Starbond, a CA glue thin as water, and it does not sit on the surface. It runs into the fibers and stiffens them from inside. Carve up to a repaired break and the wood there behaves almost like plastic under the edge, clean and dense, while the surrounding basswood slices normally.

I read that people doing marquetry use it to fill gaps between veneers, and those are not people who would trust a glue that fails. Honestly I do not love CA glue in general. It changed my mind enough that I now think about stiffening weak areas before they break at all.

Two things make that decision easy on this job. The break was tiny, and the corbel is paint grade, so the repair will disappear under the finish. On a piece meant to stay bare I would think much harder.

Which tools and kit do this job?

Mostly ordinary gouges, and then a short list of things that have nothing to do with carving.

Kit	Size	What it does	Where it turns up on this job
Number 11 veiner	7 mm and 15 mm	Outlines a shape and frees the waste beside it	Nearly every workshop in this stretch
Number 9	6 mm, driven	Bores a plug hole by hand where a drill will not reach	Sinkings
Number 7	14 mm	Brings the shape to within a millimeter of the line	The drop, first of all
Number 3	Medium	Cleans	After every roughing pass, over the whole piece
Pointing machine	Needle on an arm	Takes a depth where calipers will not fit	Back face
Veritas plate	3 in plate, 2.5 in screws	Carries the weight of the blank and lets it turn	All twelve of the workshops
Scroll saw	A narrow blade	Cuts eight quarter material and takes off waste a gouge would spend an hour on	Layers, cheeks
Starbond CA glue	Thin	Mends breaks and stiffens weak grain	Any time something snaps

Why does a tool handle work loose?

Wood moves. A ferrule that gripped in winter can spin in summer, because the handle shrinks and swells with the moisture in the air, and that is not a defect in the tool.

Some of the older ones I own never did this. The brand changed hands, and now a few of them will pull clean away from the ferrule if you tug. CA glue fixes it, and so does epoxy, which is closer to a resin and stays put. The steel is still good steel. Only the handle let go.

None of it helps if the edge is dull, which is a whole course of its own. Sharpening is the hardest of the three skills and it takes years.

Who is this stretch of the room for?

This is the seventh block of one project. It opened in workshops 1 through 12, carried on through 13 to 25, 26 to 37 and 38 to 49, and by 50 to 61 the shadows were doing the work. Nobody arrives here on their first week.

If you are still fighting a flat panel, go and fight that, and draw before you carve. A pencil and paper cost nothing and I would rather you spent a month there than ruined a nine inch blank. You have to be real about which stage you are on.

But if you have carved in relief and you want to know how the same hand makes something that reads from every side, this is the stretch. Do you want to understand the holding and the measuring as well as the carving? Would a bracket that looks like it grew straight from the wall interest you more than another flat ornament? 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 tools course answers what to buy first.

One last thing. The flower inserts on the main body got carved off camera, and not because they are any kind of secret. It is exactly the same operation I already filmed on another course, so putting it on tape again would just be redundant.