Give the test cube one job: show whether a small flow adjustment improves this filament profile under repeatable conditions. Begin with a working printer; a jam or loose drive gear needs its own repair. Once extrusion is consistent, the multiplier is a useful fine adjustment. Keep the thin-wall result beside a normal top-surface print so one delicate measurement does not have to settle the whole question.
Know which value you are changing
Prusa distinguishes the slicer's extrusion multiplier from the printer's flow setting. Both affect extrusion, and their effects multiply. A slicer multiplier of 0.95 with firmware flow at 100% requests 95% of the otherwise calculated amount. Changing one setting does not change the displayed value of the other. See Prusa's calibration procedure[1].
Record both current values before starting, then choose the one you intend to adjust. Keep nozzle, material, temperature, layer height and speed constant during the comparison. These arithmetic examples show why the two controls belong in the same note; they are not suggested filament settings:
| Slicer multiplier | Firmware flow | Combined requested flow |
|---|---|---|
| 0.95 | 100% | 95% of the otherwise calculated amount |
| 0.95 | 90% | 85.5% of the otherwise calculated amount |
In the second row, 0.95 × 0.90 = 0.855. Moving only the slicer value back to 1.00 would still leave firmware flow at 90%. Decide on a recorded baseline before printing, so the next comparison has one change to explain.
Prepare the wall test
Prusa’s procedure names a Precision Method Cube[2]. Slice that test in Spiral vase mode with a known extrusion width. The documented example uses a 0.4 mm nozzle and 0.45 mm extrusion width at 0.1–0.2 mm layer height. Those values describe the example’s scope; do not force a different nozzle or layer-height setup to match them merely to reuse its numbers.
Preview the sliced wall before printing. Confirm Spiral vase is enabled and use the actual configured wall width in the calculation. After the print cools, take measurements on the middle of the flat walls, clear of the base and corners. That gives each reading the same job: checking the intended wall rather than a different part of the shape.
Measure more than once
Following Prusa’s method, take at least three measurements in the middle of each wall and calculate the overall average. Use light, repeatable contact and keep the individual readings as well as the result. Prusa cautions that cheap digital calipers may not measure a single perimeter reliably; extra digits on the display do not settle that limitation.
Look at the readings before trusting their average. If repeated contact changes the result noticeably, or one face consistently differs from the others, repeat the measurement and investigate the variation before choosing a tiny correction. Averaging can describe the readings you took; it cannot tell you that the instrument, wall and extrusion were consistent.
For an illustrative current multiplier of 1.00, target width 0.45 mm and measured average 0.48 mm, the ratio suggests 1.00 × 0.45 / 0.48 = 0.9375. Use that as a starting value for the next print, then measure again to see whether the adjustment produced the intended wall.
Cross-check on a solid top surface
Use Prusa’s Visual Method Cube[3] for a separate top-surface check. Its documented visual procedure uses the default Prusa PLA presets at 0.20 mm layer height. Keep that scope in mind: a PLA result does not validate a different filament profile. To compare the two tests on your chosen filament, keep the material and compatible profile conditions consistent rather than switching to PLA simply to follow the example. The result to look for is an even top without excess material gathering near the perimeters or visible gaps between the lines; that gives you something concrete to compare with the wall test.
In that procedure, buildup calls for reducing the multiplier, while visible line gaps call for increasing it. Prusa makes an exception for microscopic gaps near the perimeters, so do not chase every tiny edge opening with more flow. Make a small recorded change and print again rather than assuming the computed value has finished the calibration. If the wall result and top surface disagree, revisit the sampling and profile conditions before making another adjustment.
Save the successful setting in the filament profile with the conditions used. Then verify a representative functional part separately, checking the holes, external dimensions and loaded joints that matter to its job. That final check lets you judge whether the calibrated flow also works for the part you intended to print.
