Resolution, Accuracy and Precision

Three different questions

Term	The question it answers	How you find it
Resolution	What is the smallest step or feature the machine can make?	The printer's specification, and the layer height you choose
Accuracy	How close is the printed part to the dimensions in the design?	Measure printed parts against the CAD dimensions
Precision	If I print it again, do I get the same result?	Print the same part several times and compare
Tolerance	How far off can the part be and still do its job?	You decide, from what the part has to do

Formlabs defines accuracy as closeness to the true value, where the true value is the dimension you designed in CAD, and precision as the repeatability of measurements. Tolerance is the acceptable variation for your application, and Formlabs notes tighter tolerances generally cost more in manufacturing and checking.

Resolution: Z and XY

Z resolution is layer height: how thick each slice is. It controls how visible the steps are on slopes and curves (see Choosing a Layer Height).

XY resolution is how finely the machine can place material within one layer. Formlabs describes it differently for each technology:

• FDM: the smallest movement the extruder can make within a layer. In practice the nozzle opening and the width of the extruded line limit how small a feature can be.
• Laser resin (SLA): a combination of laser spot size and how finely the beam can be steered. Formlabs says the minimum feature can never be smaller than the laser spot.
• Projector or screen resin (DLP, MSLA): the pixel size, the smallest feature the light source can reproduce within one layer.

Formlabs also cautions that not every printer's published resolution holds true in practice. Treat a resolution figure as an upper limit on detail, not a promise.

Why resolution is not accuracy

Formlabs puts it directly: printer specifications alone do not represent final dimensional accuracy. A machine with very thin layers can still make a part that is noticeably off size. The errors come from elsewhere:

• Material shrinkage. Plastics shrink as they cool or cure. Formlabs notes that resin parts shrink during post-curing, and that ABS warps more than PLA on FDM because it shrinks more as it cools.
• Process effects. Formlabs notes that parts are built layer by layer and each layer can add error.
• Post-processing. Sanding off support marks or layer lines removes material and changes the dimension.
• Calibration. Flow, belt tension, first-layer height and slicer profile all shift the result. Formlabs notes that generic setups can work but often need more calibration and experimentation.

Typical accuracy by process

Formlabs publishes typical tolerances for each technology. They are general figures for the process, not a promise for your machine Rule of thumb:

Process	Typical tolerance (Formlabs) Rule of thumb
FDM	±0.5 percent, with a lower limit of ±0.5 mm
SLA and DLP resin	±0.15 percent for 1 to 30 mm features, ±0.2 percent for 31 to 80 mm, ±0.3 percent for 81 to 150 mm, minimum ±0.02 mm

What this means at the printer

• Want finer surface detail? That is resolution: lower the layer height, or on resin, the printer's pixel or spot size sets the floor.
• Parts the wrong size? That is accuracy. Measure, then calibrate; do not change layer height. See Dimensional Accuracy.
• Every print a little different? That is precision. Look for things that vary between prints: bed temperature, first-layer height, wet filament, or a loose belt.
• Parts must fit together? Decide the tolerance first, then design clearance into the model. Formlabs recommends fit tests such as hole-and-shaft samples to find the clearance your setup really needs Measure it.

Measuring honestly

1. Print a simple test part with known dimensions, using your normal profile.
2. Measure several features with calipers, including an outside width and an inside hole, after the part has cooled.
3. Compare each to the design value. The difference is your accuracy for that feature Measure it.
4. Print the same part again and measure again. The spread between prints is your precision.
5. Change one thing at a time and re-measure.