Resin Print Exposure Calculator
Cure depth is logarithmic, so exposure barely changes depth but moves every dimension sideways at the same rate — which is why holes shrink.
What more exposure buys — and costs
Layer height against exposure and print time
Cure depth is logarithmic, so exposure is a weak lever — except sideways
Light decays exponentially into resin, so cure depth follows Cd = Dp·ln(E/Ec). Doubling the exposure adds only Dp·ln2 ≈ 0.069 mm of depth on standard resin. Each further doubling adds the same small increment, never more — which is not what the settings screen leads anyone to expect.
So doubling the layer height needs only about 1.65× the exposure, not twice. Adhesion needs the new layer to cure into the one below by a roughly fixed bond depth — it's a weld, not a proportion — so the required cure is the layer plus a constant and the ratio is exp(Δh/Dp). The rival model, where overcure scales with layer height, lands above 2× whatever ratio you assume, while this one lands below 2× whatever you assume. The two sit on opposite sides of a doubling, and shipped printer profiles are on the sublinear side. That's how you can tell the bond is a fixed depth.
But overexposure moves XY at the full rate while barely moving depth, and that asymmetry is the whole problem. The same logarithm applies sideways, so doubling the exposure adds 0.069 mm of cure to every surface: a hole shrinks 0.139 mm across and a peg grows the same. Adhesion is nowhere near failing at that point. Overexposure ruins fit long before it ruins the print, which is why "add a second to be safe" quietly destroys tolerances.
And the working window is narrow for a computable reason. Bounded below by adhesion and above by tolerance, its ratio is exp(tol/2Dp): a ±0.05 mm tolerance leaves a 1.28× window — under a second wide at typical settings — and ±0.02 mm leaves 1.11×. Doubling the tolerance squares the window ratio. That's why resin exposure feels so fiddly: the window genuinely is that tight.
- Halving the layer height doesn't double the print time — it's about 1.66× on a typical setup, because the fixed lift and retract dominate. Most of each layer is the platform moving, not light.
- Test at the temperature you print at. Resin reactivity moves with temperature, so a profile dialled in at 28 °C under-cures in a cold garage. Prints tearing off supports in winter at summer settings is the classic sign — a heater beats more exposure.
- High-detail resins buy detail with shallow penetration. Less light reaches sideways, so bleed is small — at the cost of a tighter absolute window and real sensitivity to LCD uniformity.
- Print a test rather than reasoning about it. A small part is well under a pound of resin, so an exposure test costs pennies and settles what no amount of arithmetic quite can.
How to use
- Set the exposure at the bottom of the window, not the middle.
- Treat any extra exposure as dimensional error, because that is what it is.
- Retest when the room temperature changes, not just when the resin does.
- Print a small test part rather than reasoning about the settings.
Frequently asked questions
Why are my resin print holes too small?
Overexposure, almost always. Cure depth follows a logarithm of the dose, and the same logarithm applies sideways — so doubling the exposure adds about 0.069 mm of cure to every surface on standard resin, taking 0.139 mm out of every hole across. Adhesion is nowhere near failing at that point, so the print looks perfect and nothing fits.
Does doubling the layer height mean doubling the exposure?
No — about 1.65 times, not twice. Adhesion needs the new layer to cure into the one below by a roughly fixed bond depth rather than a fixed proportion, so the required cure is the layer plus a constant and the time ratio is the exponential of the height change over the penetration depth. Going from 0.05 to 0.10 mm works out at exp(0.5), which is 1.65.
How do I know the bond depth is fixed rather than proportional?
Because the two models land on opposite sides of a doubling, whatever numbers you assume. A fixed bond always predicts less than twice the exposure for twice the layer height; overcure proportional to layer height always predicts more than twice. Printer profiles as shipped are on the sublinear side — roughly 2 to 3 seconds at 0.05 mm going to 4 or 5 at 0.10 — which settles it.
What is the Jacobs equation in resin printing?
Cure depth equals the penetration depth times the natural log of the dose over the critical dose. It comes from Beer-Lambert absorption: light falls off exponentially as it travels into the resin, so the depth at which it drops below the curing threshold moves only logarithmically as you add more. It is the reason exposure is a weak lever on depth and a strong one on everything else.
Why does adding a second of exposure ruin my tolerances?
Because the depth was already in surplus and the tolerance was not. Extra exposure adds the same small amount of cure in every direction — but downward it is padding an adhesion margin you already had, and sideways it is spending a tolerance you did not. The asymmetry is the whole problem, and it is why the safe-feeling adjustment is the damaging one.
How wide is the usable exposure window?
Narrower than people expect, and computably so. It runs from the minimum that bonds to the maximum your tolerance allows, with a ratio of the exponential of half the tolerance over the penetration depth. A plus or minus 0.05 mm tolerance leaves about a 1.28 times window, under a second wide at typical settings. At 0.02 mm it is 1.11 times.
Why is my resin exposure so fiddly to get right?
Because the window genuinely is that tight, not because you are unlucky. The exposure band narrows exponentially as the tolerance tightens — doubling the tolerance squares the window ratio — so a part needing 0.02 mm accuracy has barely a tenth of a second of usable range. Anyone printing display models rather than fitted parts has a far wider window and finds the whole thing easy.
Does halving the layer height double the print time?
No, about 1.66 times on a typical setup. There are twice as many layers, but each one needs less exposure, and the fixed lift and retract dominates both. With six seconds of mechanical movement against two or three of light, most of every layer is the platform moving rather than the screen on — which is also why a faster resin buys less than people hope.
Why do my prints fail in winter at settings that worked in summer?
Temperature. Resin reactivity and viscosity both move with it, so a profile dialled in at 28 degrees under-cures in a cold garage — and the usual symptom is parts tearing off supports during a lift rather than anything obviously exposure-related. Warming the room or the vat is a better fix than adding exposure, which costs you dimensional accuracy to solve a thermal problem.
Which resin is easiest to dial in?
Standard 405 nm resin, because its moderate penetration depth gives a forgiving window. High-detail and dental resins deliberately use shallow penetration — that is exactly what buys the detail, since less light reaches sideways — but the price is a narrow absolute window and real sensitivity to how evenly the screen lights the plate. Dial a new printer in on standard resin first.
What exposure should I start with?
The bottom of the computed window rather than the middle, because every second above the minimum is dimensional error you are choosing to accept. If the print delaminates, step up in small increments and stop at the first setting that holds. Starting high and working down wastes prints and teaches you nothing about where the boundary is.
Does more exposure make a resin print stronger?
Barely, and it makes it more brittle. Once the layer has bonded to the one below, extra dose is curing material that was already solid — it raises crosslink density a little, which trades toughness for hardness, and it keeps going during post-curing. Parts that snap rather than bend are usually over-cured rather than under-exposed, which is the opposite of most people first guess.
🔒 This tool runs entirely in your browser. Nothing you enter is uploaded, logged, or stored.