Calibration
Input Shaping Calibration
Input Shaping Calibration treats motion defects as evidence: collision, pulley, belt, cable drag, acceleration, or CoreXY path. Repeat a small motion test before buying parts.
Independent third-party notes. Verify firmware, heater, electrical, and vendor-specific work against official documentation for your exact printer.
Quick Readout
Input Shaping Calibration treats motion defects as evidence: collision, pulley, belt, cable drag, acceleration, or CoreXY path. Repeat a small motion test before buying parts.
Pick what you see
Pick the Input Shaping Calibration branch
Choose the visible evidence or log clue that matches first. The card below keeps the next move to one test and one variable.
If you see
The shift happens near a curled edge, infill hit, or knocked support.
- Likely cause
- Nozzle collision caused skipped steps.
- First test
- Inspect collision marks and repeat a small motion cube.
- Change only this
- Change only clearance/cooling/support/collision cause.
- Parameter range
- Reduce 15-25% for proof
- Stop when
- Shift disappears without belt changes.
- Verify with
- Motion cube plus failed part height.
Download test STL Whether a side jump repeats under the same motion load or height.
Open matching branch Submit tested case Pick the exact path
Most failed fixes go wrong when they start from the wrong branch.
Klipper quick proof
Use this when the failure appears on Klipper or the closest matching setup.
- First test
- Inspect collision marks and repeat a small motion cube.
- Change only this
- Change only clearance/cooling/support/collision cause.
- Stop when
- The repeat test clearly improves or points to a different branch.
After a recent change
Use this if the symptom started after a nozzle, spool, plate, slicer, firmware, or maintenance change.
- First test
- Restore the last known-good context or isolate only the recent change with one small repeat test.
- Change only this
- Undo or isolate the recent change; do not retune the whole profile.
- Stop when
- The repeat test clearly improves or points to a different branch.
When the result does not change
Use this when the first proof test looks the same after one safe variable change.
- First test
- Repeat the same test once to rule out a bad slice or one-off print.
- Change only this
- Switch branch instead of stacking another setting.
- Stop when
- The repeat test clearly improves or points to a different branch.
Visual diagnosis
Match the visible pattern before changing settings.
Synthetic diagnostic reference or structured visual guide; confirm with the page test before treating it as proof.
Looks like this
- Input Shaping Calibration repeats on the same printer, material, or print condition.
- The visible pattern changes when one branch variable changes.
- The symptom can be reproduced with a small test instead of a full model.
Not this
- Fixing a dirty plate, clogged nozzle, slipping belt, or wet spool with calibration numbers.
- Using benchmark values without a verification print.
- Changing multiple calibration variables in the same run.
Common look-alikes
- Model geometry step
- Layer split from weak bonding
- Warped corner hit that caused the shift
Inspect in the photo
- Where the defect starts and whether it repeats at the same location.
- Whether the texture is smooth, rough, lifted, thin, blobby, or shifted.
- What changed recently: material, nozzle, plate, firmware, slicer, or printer maintenance.
Photo cannot prove
- The exact slicer value that caused it.
- Whether the spool is dry, the nozzle is worn, or the config is correct.
- That a purchase is needed before the same small test is repeated.
Original visual references
Synthetic examples for fast pattern matching.
These are Print Fixes synthetic diagnostic references, not user-submitted photos. Use them to compare shape and location, then confirm with the test or log evidence on this page.
Download a quick test
Layer-shift motion cube
Use when motion load, collision, belt path, cable drag, or acceleration needs a small repeatable test.
- File
- STL
- Typical time
- 15-22 min
- Material
- PLA or the material that shifted
- Dimensions
- 30 x 30 x 35 mm
- Footprint
- 30 x 30 mm
- Height
- 35 mm
Quick print settings
Download STL - Layer height
- 0.20 mm unless the page says first-layer only
- Infill
- 10-15%
- Walls
- 2
- Supports
- Off
- Speed
- Use current profile for baseline, then change only the proven variable
What it testsWhether a side jump repeats under the same motion load or height.
When to use itUse after checking obvious collisions, pulley screws, cable drag, and belt path.
Keep unchanged
- Material and spool
- Nozzle size
- Bed surface
- Every slicer value except the one variable being tested
Expected good resultThe same geometry repeats without a side jump or rubbed collision mark.
Failure result meaningA repeat shift points to motion, cable, pulley, belt, acceleration, or motor-current branch.
Slicer notes
- Keep acceleration unchanged for baseline
- Supports off
- Repeat after one mechanical change only
Good result meansThe same geometry repeats without a side jump or rubbed collision mark.
If it does not changeA repeat shift points to motion, cable, pulley, belt, acceleration, or motor-current branch.
If it gets worseRestore the last known-good value and switch to the next branch instead of stacking more changes.
Recommended Checks
0/4 done Start with the first check. Keep this page open while you test. The checklist saves on this browser so you can come back after the print finishes.
Verification
- Repeat the same test model or the same problem area after the change.
- Compare before and after photos, print time, surface quality, and failure location.
- Keep the previous profile until the new value passes at least two similar prints.
- For firmware or heater-related issues, confirm logs stay clean after a safe heat or motion test.
Field guide
Follow the branch that matches your print
The shift happens near a curled edge, infill hit, or knocked support.
- Likely cause
- Nozzle collision caused skipped steps.
- First test
- Inspect collision marks and repeat a small motion cube.
- Change only this
- Change only clearance/cooling/support/collision cause.
- Verify with
- Motion cube plus failed part height.
- Stop when
- Shift disappears without belt changes.
The print shifts on one axis, often at load changes.
- Likely cause
- Loose pulley, belt path, or belt tension.
- First test
- Mark the pulley and check grub screws before reprint.
- Change only this
- Change only pulley/belt item.
- Verify with
- Same cube repeat.
- Stop when
- Shift no longer repeats on that axis.
The shift repeats when the toolhead or bed reaches one cable position.
- Likely cause
- Cable drag, chain snag, or connector strain.
- First test
- Move the axis by hand through that height/path.
- Change only this
- Change only cable routing/strain relief.
- Verify with
- Same motion cube height.
- Stop when
- Shift no longer follows cable position.
Slow prints pass but fast jobs shift or ring badly.
- Likely cause
- Acceleration, jerk/square corner velocity, or motor current margin.
- First test
- Repeat with acceleration reduced 15-25%.
- Change only this
- Change only acceleration/motion load.
- Verify with
- Same small high-speed section.
- Stop when
- Shift stops without hiding a loose pulley.
A CoreXY/Voron shift has diagonal or belt-path clues.
- Likely cause
- Belt path, idler alignment, or gantry motion issue.
- First test
- Inspect belt path and run a small motion cube.
- Change only this
- Change only one belt/path item.
- Verify with
- Same motion section.
- Stop when
- Movement repeats square and clean.
Concrete Parameter Range
| Setting | Start | Range | Change when | Stop when | Too far looks like |
|---|---|---|---|---|---|
| Acceleration | Current profile | Reduce 15-25% for proof | Shift happens under fast motion/load | Repeat test stays aligned | Over-slowing hides loose pulleys or collisions |
| Belt tension | Current mechanical state | Small equal adjustments, then repeat same path | Belt feels loose/tight or axis jumps | Axis repeats without side shift | Binding, noise, or uneven CoreXY path |
| Pulley screws | Current marks | Inspect/retighten one axis before repeat | Shift follows one axis or load change | No new shift at same height | Changing belts before confirming pulley slip |
| Cable path | Current route | Move/strain-relief one suspect point | Shift repeats at one height or toolhead position | Same motion path passes | Creating new drag or connector stress |
Material / Machine Differences
Bambu / enclosed ecosystemUse printer-specific calibration and plate guidance first; do not copy Ender/Voron values blindly.
Ender / Bowden-style printersSeparate mechanical path and Bowden friction before treating the symptom as slicer-only.
Klipper / custom printersRecord firmware, config, motion, and log context so the next branch is evidence-based.
Wrong Turns
Changing multiple settings in one printThe improvement becomes impossible to attribute and the next branch gets weaker.
Buying a part before a proof testA free cleaning, Z, temperature, or config fix may be missed.
Using a different model for verificationGeometry changes can hide whether the original symptom is fixed.
Stop tuning when
Do not keep chasing perfection after the signal is clear.
- The same small test improves after one documented change.
- The symptom turns into a different failure family; switch branches instead of stacking edits.
- A safety, heater, wiring, or firmware warning appears; stop printing and use the safe diagnostic path.
Common setups
Jump to the branch that matches your machine or material
Klipper quick proof Use this when the failure appears on Klipper or the closest matching setup. Inspect collision marks and repeat a small motion cube. Change only clearance/cooling/support/collision cause. After a recent change Use this if the symptom started after a nozzle, spool, plate, slicer, firmware, or maintenance change. Restore the last known-good context or isolate only the recent change with one small repeat test. Undo or isolate the recent change; do not retune the whole profile. When the result does not change Use this when the first proof test looks the same after one safe variable change. Repeat the same test once to rule out a bad slice or one-off print. Switch branch instead of stacking another setting.
Copy before changing more settings
Input Shaping Calibration diagnostic brief
Fill this out after the first test so the next branch is based on evidence, not memory. The useful case is the one where only one variable changed.
Page: Input Shaping Calibration
Printer:
Slicer:
Firmware:
Material / brand / color:
Nozzle size / material:
Bed surface:
Exact symptom or error text:
Recent change:
First test run:
One variable changed:
Result:
Next branch:Still not matching?
Jump to the next likely diagnosis
Problem Pattern
Input Shaping Calibration is useful after the printer and filament are basically healthy. It keeps one calibration variable isolated so you can trust the before-and-after result.
Likely Causes
- The printer or filament changed after the last calibration value was saved.
- A profile value was copied across nozzle sizes, materials, or printers.
- Mechanical, extrusion, or drying problems are being mistaken for calibration errors.
- The calibration coupon improved, but the real part was never verified.
Print Context
- Page type
- slicer calibration
- Best first move
- Reproduce the issue on a small test, then change one variable.
After the test
Use the result, do not keep changing random settings.
If one check clearly changes the print, repeat that exact test once before moving on. If nothing changes, switch diagnosis instead of stacking more slicer edits.
Only after the evidence points here
Parts and supplies for the proven branch
Filament dryer or dry box
Before you compare
Print the same small stringing or surface test before and after a controlled dry cycle, without changing slicer values.
- Buy signal
- Popping, steam marks, rough surface, weak layers, or fine hairing improves on the same spool after drying.
- Skip if
- The spool prints clean after a simple temperature step or seam move.
- Save evidence
- Before/after photo, material, drying temperature/time, room humidity if known, and unchanged slicer settings.
Drying is a purchase only when moisture signs survive one controlled slicer change.
Filter for
- Adjustable temperature
- Fan circulation
- Spool clearance for the material you use
- Print-while-drying path if TPU/PETG stays loaded
Avoid buying
Compare after test - Passive storage box for a spool that is already wet
- A dryer purchase when a 5 C temperature step fixed the stringing
Nozzle and cleaning kit
Before you compare
Run a hot extrusion or cold-pull check, then print a small flow wall with the same filament and temperature.
- Buy signal
- Extrusion curls, skips, or stays inconsistent after cleaning, or a brass nozzle has seen abrasive filament.
- Skip if
- The problem is only first-layer Z, bed mesh, or wet filament.
- Save evidence
- Free-air extrusion photo, cold-pull result, nozzle size/material, filament type, and whether flow changed after cleaning.
Replace the nozzle only after the extrusion path test makes the blockage or wear visible.
Filter for
- Correct nozzle thread and length
- Brass for normal PLA/PETG
- Hardened steel or similar only for abrasive filaments
- Cleaning needles sized for the nozzle
Avoid buying
Compare after test - Hardened nozzles as a first-layer fix
- Random nozzle packs that do not match the hotend
Print Fixes may earn from qualifying purchases when commerce links are configured. Diagnostic steps stay independent: buy only when the failure evidence points to the part.
Warnings
- Calibration values are not universal across materials, nozzles, or hotends.
- Do not tune pressure, flow, temperature, and speed in one pass.
- A good calibration coupon can still fail on the actual part geometry.
Useful when
- Changing Input Shaping calibration with a measurable test instead of trial and error.
- You are saving calibration values by filament, nozzle, and printer.
Skip if
- Fixing a dirty plate, clogged nozzle, slipping belt, or wet spool with calibration numbers.
- Using benchmark values without a verification print.
More traps to avoid
- Changing several slicer settings at once and losing the actual cause.
- Ignoring filament condition or bed cleanliness while tuning advanced values.
- Keeping one global profile for different materials, brands, colors, and nozzle sizes.
Bench Note
Page: Input Shaping Calibration
Printer / firmware:
Slicer profile:
Filament brand and material:
Nozzle size:
Bed surface:
Recent changes:
Result to compare next:FAQ
When should I run Input Shaping calibration?
Run it after the printer is mechanically sound and the filament is in reasonable condition, otherwise calibration hides another problem.
How many settings should I change at once?
One. Save the old profile, change one value, and verify on the same test so the result means something.
Where should I record the value?
Store it with printer, filament brand/color, nozzle size, build plate, slicer version, and date.