Cutting a compound miter without swearing

Compound miters earn their reputation before you've cut a single piece, because two saw settings have to be right simultaneously, both numbers came from a formula nobody memorizes, and the failure mode isn't "slightly off" — it's a joint that visibly gaps on one face while looking fine on the other. The actual math is not the hard part. The hard part is that almost every printed compound-angle chart in circulation defines its angles slightly differently, and if you don't know which convention you're reading, two correct tables can look like they disagree with each other.

Flat miters first — no bevel involved

Before adding any slope, a flat N-sided frame (a picture frame, a simple box, anything with vertical sides and no lean) needs each corner's two mating boards cut at half the angle between adjacent sides. That's just 180 divided by the number of sides:

• 3 sides: 60° per end
• 4 sides: 45° per end (the one everyone already knows)
• 5 sides: 36° per end
• 6 sides: 30° per end
• 8 sides: 22.5° per end
• 12 sides: 15° per end

No bevel is needed for any of these — the blade stays at 90°, only the miter gauge or sled angle changes. The full method and the rest of the N values live on miter angles for common polygons; if you just need one number, miter angle for a 5 sided box and its sibling pages are the fast lookup.

Where it stops being flat

The moment the sides of a box, frame, or vessel also lean — flare outward as they rise, like a basket, a lamp shade, or an octagonal planter with sloped sides — the flat-frame formula stops being sufficient on its own. Now the saw needs two settings cut simultaneously: a miter angle (the rotation in the horizontal plane, same idea as before but modified by the slope) and a bevel angle (how far the blade tilts off vertical to match the lean of the sides). Get only one of the two right and the joint still won't close, because the two mating faces are no longer meeting at a simple flat angle — they're meeting on a tilted plane, and both saw settings have to reproduce that plane's geometry exactly.

The convention that causes most of the arguments

Read this before cross-checking any published table against a calculator, this site's included: the slope input here is measured from horizontal — a slope of 0° describes a flat, unslanted frame, and the number increases as the sides lean further off flat. Plenty of printed compound-miter charts instead measure the identical physical lean from vertical (0° meaning a perfectly upright, straight-sided box). These are not two different formulas producing two different answers for the same cut — they're the same cut, described against two different zero points, and a slope input of, say, 20° in one convention is a 70° input in the other. If a number here doesn't match a chart from a book or another site, check which axis that chart is measuring from before assuming either source made an error. It's the single most common reason a compound-miter calculation gets blamed for being "wrong" when it isn't.

Three worked examples

All three below were run through this site's own angle math, not read off a printed table, specifically so the miter and bevel settings can be trusted together as a pair for the same cut.

A 4-sided box, sides sloped 30° from horizontal (a moderately flared basket or planter): miter 40.893°, bevel 20.705°.

A 6-sided vessel, sides sloped 45° (a steeply flared hexagonal bowl or shade): miter 22.208°, bevel 20.705° — worth noticing that this bevel setting matches the 4-sided example above exactly, at a different slope and a different side count. That's not a coincidence in the data; it's a reminder that miter and bevel don't move together in any simple proportional way, and neither should be estimated from the other.

An 8-sided frame, sides sloped only 15° (a gently flared octagonal box): miter 21.806°, bevel 5.684° — here the miter angle is close to the flat 8-sided value (22.5°) because the slope is shallow, while the bevel is correspondingly small, which is the general pattern: a shallow slope nudges the flat-frame miter angle only slightly and asks for a small bevel, while a steep slope moves both settings much further from their flat-frame starting point.

Run your own side count and slope through the angle calculator rather than interpolating between these three — the relationship between slope and either output angle is not linear, so guessing at the number for a slope between two known cases will be wrong in a way that only shows up once the joint is cut.

Cutting order and the test that actually catches mistakes

Set the bevel first and leave it alone — it's the setting most saws make more awkward to adjust, and it's shared by every corner of a given piece as long as the slope doesn't change partway around. Set the miter per corner as you go, and always cut a set of test scraps at both settings and dry-fit them into a full corner — a mitered corner cut from only two pieces can look plausible sitting flat on a bench and still be visibly wrong once a third and fourth piece close the loop, because small errors in either angle compound as they go around the shape. If the joint gaps on the outside face but closes on the inside, the bevel is off; if it gaps unevenly across the width of the cut, the miter angle is off — the two failure modes look different once you know to look for them separately instead of just re-cutting both settings at once and hoping.

A practical note on the saw itself

Most compact miter saws bevel in only one direction without flipping the workpiece, which means cutting a full set of matching corners for a flared box often means alternating between bevel-left and bevel-right cuts, or flipping every other piece face-down and mirroring the miter setting instead. Decide which approach a given saw supports before laying out which face of each piece is "up," because getting that backwards is the single easiest way to end up with four correctly-angled pieces that still won't form a closed shape — the angles are right, but two of them are mirrored the wrong way. A stop block clamped to the fence, set once per corner angle, is worth the extra setup time on anything with more than four repeated corners; resetting a miter gauge by eye for every single cut is where small, compounding errors creep in fastest on a compound-angle project specifically, because unlike a flat frame there's no bevel to visually double-check against a square.

What this doesn't cover

Segmented turning on a lathe is a related but distinct problem — the pieces are typically flat-sawn rings glued up from angled segments rather than tilted panels meeting at a corner, and while the underlying trigonometry rhymes, the setup and the practical failure modes are different enough to deserve their own treatment; see segmented turning angle basics for that case specifically. And miter vs. bevel angle, explained is worth a read if the difference between the two settings themselves — not just their numeric values — still isn't intuitive; a compound cut is genuinely two separate rotations happening on two different axes of the saw, not one angle split two ways.