Input Shaping and Vibration Tuning
Input shaping reduces ringing/ghosting by pre-adjusting motion commands so the printer’s resonances aren’t excited during fast direction changes. Get the best result by fixing mechanical looseness first, running a repeatable corner-based ringing test, tuning shaper type and frequency per axis, then validating on a small real print before increasing acceleration for long jobs.
TL;DR
If you see fading ripples after sharp corners, tune input shaping (shaper type + frequency, often per axis) using a ringing test printed at a fixed speed/accel. Don’t chase “perfect” by cranking frequency; if corners get rounded, reduce acceleration or change shaper type.
What input shaping fixes (and what it doesn’t)
Ringing (ghosting) is a repeating ripple pattern that starts right after a corner, embossed letter, or any sudden direction change. It comes from the belts/frame/toolhead flexing and oscillating after an acceleration event. Input shaping modifies the commanded acceleration so that the printer’s natural resonance is excited less, reducing the ripple amplitude and making it die out faster. Input shaping will not fix ripples that are not tied to corners (those are more often extrusion inconsistency, Z wobble, or cooling changes). It also can’t compensate for major mechanical slop; loose parts will still move unpredictably.
Before you tune (mechanics first)
- Belts: correct tension, no missing teeth, no belt rubbing on flanges.
- Pulleys/gears: set screws tight on motor shafts; no pulley wobble.
- Toolhead/bed: no play in bearings/wheels/rails; nothing rattles when you wiggle it by hand (power off).
- Frame and feet: printer sits solidly; avoid a springy table that can add its own resonance.
- Baseline print profile: temperature, flow, and cooling are already reasonable so you’re not mistaking other artifacts for ringing.
- Know what your firmware/slicer uses: acceleration plus jerk/junction deviation (names differ, but they all control how hard corners are taken).
Ringing test setup (make it repeatable)
Use a small ringing/acceleration test with sharp corners and flat walls so the ripple pattern is easy to see. Keep variables fixed: same filament, nozzle temp, fan, layer height, and number of walls (one or two walls makes ringing easiest to read). Print slow enough that the nozzle can melt reliably, then use acceleration (not speed) as your main “stress knob” to reveal ringing. Record your settings each run: acceleration, jerk/junction deviation, shaper type, and shaper frequency (or the firmware’s measured values). If your tool supports separate X and Y shaping, plan to tune each axis because resonances are often different.
Tuning sequence (one change at a time)
- Baseline: print the same ringing test with your current settings. Confirm ringing is clearly visible after corners; if it isn’t, increase acceleration until it is (don’t change multiple things).
- Enable shaping: turn on input shaping using a reasonable default shaper and frequency, keep speed/accel the same, and reprint. You should see ripples shrink and decay faster.
- Tune frequency: adjust only shaper frequency (or run the firmware’s measurement routine if you have an accelerometer) and reprint until corner-starting ripples are minimized.
- Tune per axis: compare X-facing vs Y-facing walls. If one side still rings, tune that axis separately (or address mechanical asymmetry like belt tension and moving mass).
- Choose shaper type: if ringing is low but corners look rounded or small features look “mushy,” try a different shaper type or reduce the shaping aggressiveness. Don’t push frequency far past the best ringing result to compensate.
- Raise acceleration carefully: once ringing is controlled, increase acceleration in steps and reprint. Stop when you reach your acceptable tradeoff between sharp corners, noise, and print stress.
- Validate on a real part: print a small functional piece with corners/text/holes. Some models reveal artifacts that a test tower hides (like corner bulging from pressure/flow limits).
Common outcomes and what to do next
Ringing barely changes after enabling input shaping
Likely cause: Mechanical looseness dominates; the printer/table is adding resonance; shaper settings not actually applied; acceleration too low to reveal ringing
Fix: Verify belts/pulleys/toolhead are tight and the printer is on a solid surface; confirm the shaper is enabled/active; retest at higher acceleration so ringing is measurable.
Ringing improves but corners look rounded or fine features soften
Likely cause: Shaper is too aggressive for the desired sharpness; shaper type not ideal; acceleration/jerk too high for your detail goals
Fix: Try a different shaper type or lower shaping aggressiveness; then reduce acceleration slightly and rerun the same test.
X wall looks clean but Y wall still rings (or vice versa)
Likely cause: Different resonant frequency per axis; belt tension mismatch; axis has more moving mass or a looser structure
Fix: Tune X and Y separately if supported; otherwise balance belt tension, check for play on the noisy axis, then retest focusing on that axis.
Waves appear that do not start at corners
Likely cause: Not classic ringing: extrusion inconsistency, Z wobble, or cooling/fan fluctuations
Fix: Print a corner-heavy ringing test to isolate motion ringing; then check extrusion calibration/temperature stability and Z-axis mechanics if the pattern persists.
At higher acceleration, you hear sharp thumps and see layer shifts
Likely cause: Skipped steps from excessive acceleration/jerk, too-low motor current, or mechanical binding
Fix: Lower acceleration/jerk first; then check for binding, verify belt path is smooth, and confirm motor current/driver settings are appropriate before retrying.