Skip to content

Input Shaping Calculator

The Input Shaping Calculator converts ringing test measurements into resonance frequencies, for printers without an accelerometer.

Free, runs in your browser, no registration.

Input shaping cancels ringing (also called ghosting or echoing) — the repeating ripples that appear on walls just after a sharp corner.

The cause is mechanical. When the toolhead changes direction hard, the frame and belts flex and oscillate at their natural frequency. That oscillation prints into the part. Input shaping deliberately splits each move into timed impulses that cancel the vibration before it starts.

This is not a pressure advance problem. Tell them apart:

  • Ringing — repeating waves that fade with distance after a corner. Input shaping.
  • Corner bulge or gap — a defect at the corner itself, not repeating. Pressure advance.

Tuning the wrong one is the most common wasted evening in Klipper calibration.

If you have an ADXL345, LIS2DW, or a printer with one built in (K2 Plus, SV08, most modern CoreXY), skip this calculator entirely:

SHAPER_CALIBRATE
SAVE_CONFIG

Klipper measures the resonance directly and writes the shaper and frequency into printer.cfg. It takes two minutes and is more accurate than any visual method.

Run it once per printer, and again after any change to mass or belt tension — new toolhead, new hotend, retensioned belts.

That’s what this calculator is for.

  1. Print the Klipper ringing test model at high acceleration, with the test macro:
    SET_VELOCITY_LIMIT ACCEL=7000 ACCEL_TO_DECEL=7000 SQUARE_CORNER_VELOCITY=1
    TUNING_TOWER COMMAND=SET_VELOCITY_LIMIT PARAMETER=ACCEL START=1500 STEP_DELTA=500 STEP_HEIGHT=5
  2. Measure the distance between ringing peaks in millimetres with calipers.
  3. Note the speed the test ran at.
  4. Enter both into the calculator:
    frequency (Hz) = speed (mm/s) / distance_between_peaks (mm)
  5. Do this separately for X and Y — they are almost never the same.
  6. Put the results in printer.cfg:
    [input_shaper]
    shaper_freq_x: 52.4
    shaper_freq_y: 38.1
    shaper_type: mzv
ShaperVibration reductionSmoothing costUse when
ZVLowestLeastFrequencies are high and stable
MZVGoodLowDefault choice for most printers
EIBetterModerateFrequency measurement is uncertain
2HUMP_EIHighHighTwo distinct resonance peaks
3HUMP_EIHighestHighestMessy resonance, low frequencies

There is a real tradeoff: more aggressive shapers remove more ringing but round off fine detail. MZV is the right default. Only move up the table if MZV leaves visible ringing.

If SHAPER_CALIBRATE recommends 3HUMP_EI, that’s usually a hint that the machine has a mechanical problem worth fixing — loose belts, a wobbling gantry, or a bed that isn’t rigid — rather than a shaper choice to accept.

  • Tuning input shaping before fixing mechanical looseness. Shaping masks the symptom; a loose eccentric nut is still loose. Check belt tension and frame rigidity first.
  • Confusing ringing with pressure advance artifacts. See above.
  • Using the same frequency for X and Y. On most printers they differ substantially — moving mass is different on each axis.
  • Not re-running after hardware changes. New toolhead, new hotend, retensioned belts — all shift resonance.
  • Testing at low acceleration. If the test doesn’t provoke ringing, you can’t measure it. The tuning tower above deliberately runs high.

Input shaping is a motion calibration, separate from the extrusion chain. Do it in parallel:

Extrusion: rotation distance → flow → max volumetric speed → pressure advance → extrusion rate smoothing Motion: belt tension → input shaping → acceleration limits

They meet at high acceleration, which is where extrusion rate smoothing becomes relevant.

Measured shaper frequencies and types for my Voron 2.4, Mercury One.1, K2 Plus and SV08 are documented in the Minimal 3DP config packs, including the raw resonance data. Free configs are in the klipper repository.