Estimate separate X, Y and Z corrections using your own calibration measurements. For each axis, enter the nominal calibration dimension, its measured printed dimension, and the desired finished dimension of your next model. The calculator returns a scale multiplier, a scale percentage and the corrected model dimension.
Enter all lengths in metres: divide millimetres by 1,000 before entering them. For example, 20 mm = 0.020 m and 19.8 mm = 0.0198 m. Multiply corrected dimensions in metres by 1,000 to read them in millimetres. Each input accepts 0.000001–10 m, with a 0.000001 m step. These are calculator entry bounds, not printer capability or measurement-accuracy claims.
The compensation formula is derived here from a proportional-response assumption, independently for each axis: predicted printed length = corrected model length × (measured calibration length / nominal calibration length). Setting predicted printed length equal to the desired dimension and rearranging gives C = T × N / M. This is a calibration-ratio model, not a published universal shrinkage value for any material.
For the scaling operation, Prusa documents independent axis scaling by percentage and setting explicit model dimensions. This supports how to apply the result; it does not establish that a printer’s dimensional error is proportional. Prusa: scaling and model dimensions.
A result of 101.0101% means a multiplier of approximately 1.010101, or a 1.0101% increase. It does not mean adding 101.0101%. A result below 100% reduces that axis. Equal nominal and measured dimensions give a multiplier of 1 and a scale of 100%.
| Axis | Nominal calibration (mm) | Measured calibration (mm) | Desired finished dimension (mm) | Correction multiplier | Scale (%) | Corrected model dimension (mm) |
|---|---|---|---|---|---|---|
| X | 20 | 19.8 | 100 | 1.010101 | 101.0101 | 101.010101 |
| Y | 20 | 19.6 | 50 | 1.020408 | 102.0408 | 51.020408 |
| Z | 20 | 20.2 | 30 | 0.990099 | 99.0099 | 29.702970 |
| Unchanged axis | 20 | 20 | 40 | 1.000000 | 100.0000 | 40.000000 |
- Print a calibration model at 100% object scale with additional shrinkage compensation disabled. Keep the material, orientation and process settings representative of the intended part.
- Measure each axis after the same cooling or post-processing stage intended for the finished part. Repeat measurements to check consistency, and enter a representative measurement for each axis.
- Start with a model whose dimensions equal your desired finished dimensions. Apply each axis’s percentage independently, or set its dimensions to the corrected dimensions. Use one of these methods once; applying both would double the correction.
- Preserve the mapping between the calibration axes and the printer axes. Print and measure a verification part before relying on the adjustment.
The calculation assumes a constant multiplicative error per axis. It does not model warping, local feature errors, hole offsets, first-layer distortion, measurement uncertainty or size-dependent error. Test more than one calibration size if you need to check that assumption. Large corrections warrant checking units, axis labels and the printing setup first. The corrected dimensions must still fit your printer’s usable build volume.
Zero, negative and out-of-range dimensions are invalid. If a result is unavailable, correct the inputs; unavailable is not a zero correction. Displayed decimal places describe numerical rounding, not achievable print accuracy.
Draft production flag: no registered thumbnail asset was supplied. Create and register a 400 × 400 pixel WebP thumbnail, then reference its real asset ID in the first image block. Use the established cheerful yellow five-point star mascot floating with hands and no separate legs or feet, a white background, and playful navy, blue and teal calibration-cube and X/Y/Z measurement graphics. Inspect neighboring tool thumbnails for style consistency before creation.