Bulging corners and uneven lines: pressure advance

This guide is for prints whose straight walls look fine but whose corners bulge, round off or show a thin patch just after the turn. It explains what pressure advance does, how to calibrate it in Bambu Studio, PrusaSlicer and Prusa firmware, OrcaSlicer, Cura and Creality Print, and the other corner causes makers document. Calibration options are kept to the printer models each maker names. Make sure you're in the right place:

What does pressure advance do?

Bambu Lab's explanation: "When filament is extruded, it takes time for the pressure in the nozzle to build up to a level where plastic flows consistently." During acceleration, "this lag can cause underextrusion, resulting in gaps or thin lines." When decelerating, "the residual pressure causes excess filament to ooze out, leading to blobs or stringing." Bambu adds that "these issues worsen with faster print speeds."

The feature counters this by pushing a little extra filament as the head speeds up, and reducing or reversing flow as it slows. Bambu says the result is "cleaner corners, sharper details, and more consistent extrusion." Bambu and Prusa both call the number that sets its strength the K value.

Bambu's under-extrusion page ties corners to this directly: "If under-extrusion only appears at the model's corners," the pressure advance value "is likely set incorrectly." It says a wrong value "may cause too much or too little extrusion at the moment the print head changes speed."

How do you read a calibration test?

Every slicer's test prints lines or shapes at a range of values. You pick the best one by eye.

• Bambu Lab (pattern test): pick the value with "the cleanest printed corner," which "should be as sharp as possible, without any signs of under-extrusion." In Bambu's example, a low K value of 0.006 is "too low, as the corner is bulging and it's round," while from 0.024 up "we obtain underextrusion in the corner."
• Prusa (Linear Advance): "If sharp corners have blobs, increase the K value. If you see missing filament, decrease the K value."
• OrcaSlicer (tower test): "Examine each corner and seams of the print and mark the height that yields the best overall result."

Creality's Creality Print 4.3.8 tutorial puts it the other way round: "A larger PA value will cause overflow, and a smaller PA value will cause insufficient extrusion of consumables." That doesn't match what Bambu and Prusa describe for corners. Whichever you read, the test works the same way: choose the line or height that looks cleanest.

Before you calibrate

Bambu lists conditions that make its calibration results unreliable:

• Damp filament, "which will render the calibration results unsuitable for use with fresh filament." See our wet filament guide.
• A worn nozzle or one with an internal blockage.
• Soft materials such as TPU, which Bambu says have "a high probability of calibration failure."
• A build plate that isn't sticky, so the test lines don't adhere. Bambu says this affects the lidar scan on X1 series printers.

Bambu also says to clean the extruder, because "poor extrusion often comes from dirty gears, not calibration." OrcaSlicer's calibration guide puts pressure advance after temperature and max volumetric speed, and before flow and retraction.

Bambu Studio and Bambu Lab printers

Bambu says to run Flow Dynamics Calibration:

• When you use "a new filament from a different brand or model."
• When the nozzle is worn, or when you replace it.
• When "the maximum volumetric speed or print temperature is changed in the filament settings."

If you don't calibrate and haven't picked a profile, Bambu says the printer uses default compensation, and "in most cases, it is possible to achieve good printing results with the default parameters."

Automatic or manual: which printers do what

Bambu says "H2D, X1 and A1 series printers support auto calibration," the A1 series only "from firmware version 01.04.00.00," and that "P1 supports only manual calibration." On the X1 series, the printer prints lines and scans them with the lidar. On the A1 and H2D, it purges at the wiper and measures extrusion force with the eddy current sensor in the toolhead.

To run it automatically, connect the printer, click Calibration, Flow Dynamics, then Auto-Calibration. Pick the nozzle, plate type and filament, and press Calibrate. Bambu shows the recommended K value at the end and saves it to the printer. You can view and edit it under Manage Result.

Bambu's notes on the automatic result:

• On A1 series and H2D printers, the automatic K value "may sometimes be significantly higher than the manual calibration value, which is normal." Bambu says this is meant "to avoid filament pile-up problems at the corners of the model."
• With a 0.2 mm hotend, on any series, Bambu says automatic calibration often gives inaccurate results or fails, and recommends the manual method.
• A third-party hotend may give inaccurate results, which Bambu says is "especially relevant for the A1 series and H2D printers."

Manual calibration

1. Clean the plate. Bambu recommends a smooth plate such as the Smooth PEI Plate, though the Textured PEI Plate also works.
2. Set the filament type and color on the Device tab.
3. Go to the Calibration tab, select Flow Dynamics, then Manual Calibration.
4. Click Sync, then choose the nozzle diameter, plate type and filament.
5. Choose the Pattern method. Bambu says straight-line tests "are less accurate because they don't clearly show extrusion issues like corner patterns do."
6. Set the interval. Bambu recommends a step of 0.002.
7. Print, find the cleanest corner, enter that value under Factor K, name it and click Finish.

Calibrating before each print

If you tick "Flow Dynamics Calibration" when sending a print, the printer calibrates during preparation and uses that K value instead of the manual one. Bambu says P1 series printers don't have this option.

• X1 series: only the first filament in the job is calibrated. Others use the K value you set manually, or the default.
• A series: each filament is calibrated the first time it's used in the job (supported since firmware 01.03.00.00).
• H2D: the choice is Automatic (recalibrates only if the filament, nozzle or hotend changed), Open (every print) or Off (uses your manual K value, or the default).

Bambu Studio also has a "Pressure advance" setting. In Bambu Studio 2.08.02.61 its tooltip reads "Pressure advance(Klipper) AKA Linear advance factor(Marlin). Useless for Bambu Printer," and its default is 0.02.

Prusa printers and PrusaSlicer

MK3 and MK2 families: Linear Advance

Prusa calls this "an advanced calibration for MK3-family and MK2-family printers." Its article names the MK2.5, MK2.5S, MK3, MK3S and MMU2S. Firmware 3.9.0 introduced Linear Advance 1.5 for the MK3S, MK3, MK2.5S and MK2.5.

• Stock profiles: "These values are part of our PrusaSlicer configurations and filament presets. No need to adjust if you are using these."
• Calibration: Prusa offers K-factor calibration G-code files for a 0.4 mm brass nozzle, for PLA, PETG, ASA, PC, nylon, ABS and HIPS. Each prints lines that go slow, fast, then slow. Look at the transitions and pick the most consistent line.
• Where to set it: in PrusaSlicer, "the K-value is set in the custom G-code section in the Filament Settings tab, NOT under the printer-specific custom G-code."
• Other slicers: Prusa says Simplify3D and Cura users "just need to add 'M900 Kxx' into the starting G-code script," and must change it by hand for each material.

Prusa's tested values for a 0.4 mm nozzle include PLA 0.05, PETG 0.08 and ASA/ABS 0.04. It says different brands and colors may need a slightly different value "when printing at extreme speeds," and warns that "changing them is recommended only for experienced users."

MK4/S, MK3.9/S, XL and MINI/+: Pressure Advance

Prusa says Pressure Advance replaces Linear Advance on these printers "from firmware version 5.0.0 and above." It's set with the G-code M572, and PrusaSlicer's default start G-code includes an M572 command whose value changes with nozzle diameter. Prusa says older M900 commands are "automatically converted 1:1" to a Pressure Advance value.

Prusa's advice is: "We recommend keeping the default S parameter, optimized for most applications on our printers." Its article doesn't describe a calibration print for these printers.

OrcaSlicer

OrcaSlicer's calibration tests are in the Calibration section of the interface. For pressure advance it offers three methods, each "with two versions: one for a direct-drive extruder and one for a Bowden extruder":

• Tower: slower, but "does not rely on the quality of the first layer." The value rises one step per millimetre of height. Find the best height and calculate: start value plus step times height. Orca's example: 0 + (0.002 × 8) = 0.016.
• Pattern: adapted from Andrew Ellis' pattern generator. It uses your outer wall and first layer speeds.
• Line: quick, but "its accuracy depends heavily on the quality of your first layer."

Orca's warnings: on Marlin printers, "Linear Advance must be enabled in firmware (M900)," and on Bambu Lab printers, don't select the "Flow calibration" option. Orca says the results "should be saved to the material profile," where the Pressure Advance setting lives.

Orca also has adaptive pressure advance, which picks a value based on flow and acceleration. Its wiki says it "has been tested with Klipper-based printers" and is untested on Marlin and Bambu Lab printers. It says that with one layer height, speed and acceleration, "there is no need to enable this feature."

Cura

Cura 5.13.0's base printer definition includes a "Pressure advance factor" setting, described as a "Tuning factor for pressure advance, which is meant to synchronize extrusion with motion," with a default of 0.05. It's switched off from view in that base definition, so whether you see it depends on your printer's profile. We didn't find a pressure advance calibration test documented by UltiMaker for Cura. For Prusa MK3 and MK2 family printers, Prusa's method is the M900 line in the start G-code, above.

Creality Print

Creality's calibration tutorial for Creality Print 4.3.8 includes a pressure advance step with two methods, "PA tower" and "PA line." For the tower, you enter a start value, end value and step. After printing, find the height with the best result and calculate PA as the start value plus height times step. Then enter the value in the consumables (filament) parameters.

Creality says the best value "is related to the type of consumables, so it is highly recommended to re-measure the most appropriate PA value when changing the type of consumables." The tutorial describes Klipper firmware but doesn't name printer models. Check that your printer and Creality Print version support it.

What else causes bad corners?

If calibration doesn't fix it, these are the other corner causes the makers document.

• Cornering speed (jerk, junction deviation, square corner velocity). OrcaSlicer says cornering settings can "reduce artifacts such as ringing, ghosting, and overshooting." For jerk, it says lower values give "smoother corners, better quality," while higher values give "faster cornering, potential artifacts." Orca's Cornering test finds the highest value before "corners start losing sharpness." Cura 5.13.0 says raising jerk "can reduce printing time at the cost of print quality."
• Outer wall acceleration. OrcaSlicer says outer wall acceleration "is usually set to a lower value than normal printing to ensure better quality." Cura 5.13.0 says the same of acceleration as a whole: raising it can "reduce printing time at the cost of print quality."
• Outer wall speed. Cura 5.13.0 says printing the outer wall slower "improves the final skin quality," but that "a large difference between the inner wall speed and the outer wall speed will affect quality in a negative way." In the base definition, Outer Wall Speed defaults to 30 mm/s.
• One value for every speed. OrcaSlicer says the effective value "typically decreases" as speed and acceleration rise, so a single value is a compromise between "bulging on features with lower flow speed and accelerations" and "gaps on faster features." Its adaptive pressure advance is meant for this.
• Too much acceleration with input shaping. Orca says not to calibrate adaptive pressure advance above the acceleration Klipper's input shaper recommends, because beyond it "the perimeter separations that appear like PA issues are caused by the input shaper smoothing the shape of the corner." Our ghosting and ringing guide covers input shaping.
• Flow changes after bridges (Bambu Lab). When the outer wall goes from low to high flow, "abnormal bulges or under-extrusion dents may occur." Bambu suggests Flow Dynamics Calibration, a lower outer wall speed and acceleration, and "Smooth Speed for Discontinuity Area" in Bambu Studio. Our Z banding guide has more.
• A lifted corner (Bambu Lab). If a corner warps up, "the nozzle-to-part gap shrinks, causing extra filament to squeeze out sideways, creating continuous bulges." Fix the adhesion. See our warping guide.

If corners still won't come out sharp and every wall is too wide, check overall flow in our over-extrusion guide. If you're measuring parts to check fit, our dimensional accuracy guide covers where to measure and why corners can mislead.

A sensible order

1. Dry the filament and clean the nozzle and extruder gears.
2. Confirm you're on the maker's profile for that filament and nozzle.
3. Run your slicer's pressure advance or Flow Dynamics test, using the pattern or tower method.
4. Save the value in the filament profile, or the printer's K profile on Bambu Lab printers.
5. If corners still blur or ring, lower outer wall acceleration and try the cornering test.
6. Recalibrate after changing filament brand, nozzle or print temperature.

Checklist

• Corners bulge or round off, or lines thin out just after corners.
• Filament dry, nozzle and extruder clean.
• Test printed on a clean, sticky plate.
• Value chosen from the sharpest, fullest corner.
• Value saved to the filament or K profile you actually print with.
• On Prusa MK3 or MK2 family printers with Cura or Simplify3D, M900 K added to the start G-code.
• Outer wall acceleration lower than normal printing, as OrcaSlicer describes.
• Recalibrated after a filament, nozzle or temperature change.

How this guide was checked

This is a research-based guide. We haven't printed these calibration tests for it. Every explanation, menu path, default and printer model comes from the Bambu Lab, Prusa Research, OrcaSlicer, UltiMaker Cura and Creality pages listed in the sources, retrieved on 2026-10-12. Defaults are from Bambu Studio 2.08.02.61 and Cura 5.13.0 source files. Calibration options are kept to the models each maker names. Read more in our editorial standards.

Print problems: what to check before you buy ↗