Matching Inset Cabinetry to Original Millwork

The first visit to a period house is spent with a tape, a profile gauge, and a notebook. The dimensions worth recording are the ones the eye measures without knowing it: the width of the stiles and rails on the original doors and on any surviving built-in, the height of the base and the plinth beneath it, the width and the projection of the window and door casings, the height of the wainscot cap and the chair rail if there is one, the depth of any panel raise or the depth of the reveal where a flat panel sits back from its frame, and the size of any bead run on a face frame or a door edge. Each of these is a proportion the house has already decided, and a new run of cabinetry that uses different ones will look slightly foreign even to someone who could not say why.

Profiles are recorded by tracing or by sample. A contour gauge pressed into a casing gives a working outline; a short length of the original moulding, if a piece can be found loose in a basement or behind a later addition, is better, because a knife can be ground to the actual section rather than a drawing of it. Where the only original is still on the wall, a tracing is made on card and checked against the moulding from two angles, since the depth of a quirk or the fullness of an ogee is easy to misread from the front alone.

Inset construction puts the door inside the face-frame opening rather than over it, so the gap around the door is visible on all four sides and has to be the same on all four sides, on every door, across the whole run. The Architectural Woodwork Institute’s Architectural Wood Casework standard sets gap and flushness tolerances by grade, and its figure list includes a drawing titled “Gaps, Inset Face Frame”; the specific figures are in the standard itself rather than on the public page, so this guide does not quote them. What the standard establishes, and what a period house makes obvious, is that the reveal is a specification and not an accident.

The reveal exists because wood moves. The Forest Products Laboratory’s Wood Handbook gives white oak a radial shrinkage of 5.6 percent and a tangential shrinkage of 10.5 percent from green to oven-dry, and the pattern — tangential movement roughly twice radial — holds across the species used for cabinetry. Expressed as a coefficient, the same source puts white oak at about 0.00365 of the dimension per one percent change in moisture content in the tangential direction and 0.00180 in the radial. A solid inset door in a frame-and-panel construction is built so that the panel floats in its groove and the frame carries the dimension; the reveal is then sized for the frame’s seasonal movement and not the panel’s, which is a much smaller number. Quartersawn stock, which presents the radial face, roughly halves the width change of flat-sawn stock and is the reason it is the first choice for inset doors in a dry-summer California house.

Holding the reveal in practice is a matter of sequence. The face frame is assembled and squared first; the doors are fitted to each opening individually, with the hinge side planed to the frame and the latch side fitted last; the shop’s moisture content is kept close to what the house will see in service, which in a Sacramento Valley or Bay Area interior means drier than the lumberyard. Installation then has its own tolerances, which the installation-tolerances guide covers: a frame that is racked by an out-of-level floor will show it first in the reveals, because the openings are no longer square and the doors still are. A period house with settled floors and plaster walls that are nowhere plumb is where scribing the cabinetry to the room matters most.

The commonest mistake in finish matching is to try to do it with colour alone. A new piece of flat-sawn red oak will never match an old quartersawn white oak built-in no matter what is put on it, because the ray fleck, the pore structure, and the way the surface reflects light are different in kind. The survey settles the species and the cut, and the match begins by building the new work in the same wood, sawn the same way. Only then is colour addressed, and it is addressed in layers — a dye or stain for base colour, a toner in the clear coat to shift hue, a glaze in the recesses to put back the shadow that a century of handling and dust has left in the original.

The original finish is almost never what it was. Shellac and oil-and-varnish finishes of the period have oxidised and darkened; the wood under them has changed colour with light; the surface has been waxed or recoated or both. The new finish has to match that aged surface rather than a fresh one, which means matching a warmth and a depth as well as a colour. Sample boards are made in the same species and cut, finished with the candidate schedule, and held against the original in the room — in daylight and under the kitchen’s lamps, since the two can disagree — until one is accepted. The sample that matches on the bench and not in the room is the one that was tested in the wrong place.

Painted work shifts the problem from colour to sheen. A colour chip can be matched by a paint store; what makes painted new cabinetry sit beside original painted trim is that the sheen agrees, the film thickness looks similar, and the edges of the profile are neither sharper nor rounder than the original. Modern catalysed finishes are harder and more resistant than the shellac, oil, and early varnishes on the original woodwork; the Architectural Woodwork Institute’s factory-finishing standard, ANSI/AWI 0400, is concerned with the performance and appearance of factory-applied finishes and explicitly excludes jobsite finishing from its scope, which is one reason shop-finished cabinetry arrives in a period house already matched rather than being finished to match on site. The trade-off is honest: a durable new finish will age differently from the original over the next decades, and the sample that matches today is a match for now.