Does your printer shake like it’s nervous every time it changes direction? Those ripples on the side of a print, the ones that look like echoes of the model itself, are called ringing. And input shaping calibration is the fix that finally made high-speed printing usable for regular people.
Here’s the thing most guides skip: input shaping isn’t a magic setting you copy from a forum post. It’s a measurement of your specific machine. Belts, frame stiffness, toolhead mass, even how tightly you bolted your accelerometer, all change the answer. Get that wrong and you’ll either kill your print quality or leave speed on the table.
This guide walks through what input shaping really does, how to calibrate it across the major firmwares, and the mistakes that trip up nearly everyone the first time.
What Input Shaping Actually Does
Input shaping is a feedforward vibration-compensation technique. Instead of waiting for the frame to wobble and then correcting it, the firmware filters the motion commands before they reach the motors so the frame never rings in the first place. (Resonance Compensation)
The idea comes from control theory. Researchers N.C. Singer and W.P. Seering at MIT developed the math in the 1980s and 1990s, publishing “Preshaping Command Inputs to Reduce System Vibration” in the ASME Journal of Dynamic Systems, Measurement, and Control in 1990. The core concept: split a single movement command into several smaller ones, timed so their vibrations cancel each other out.
In 3D printing terms, that means the toolhead arrives at the corner slightly differently but the frame stays quiet. You get cleaner walls and the freedom to push acceleration higher without turning your prints into corrugated cardboard.
Why Ringing Happens in the First Place
Every printer frame has resonant frequencies. Hit those frequencies and the whole machine vibrates like a tuning fork. Sharp direction changes at speed excite those resonances, and the nozzle keeps oscillating for a few milliseconds afterward. That oscillation gets baked into the plastic as ripples.
Belt tension, frame rigidity, and toolhead mass all shift those resonant frequencies. This is why a calibration from one machine shouldn’t be copied to another, even if it’s the identical model. It’s also why a “best budget 3d printer” that ships with loose belts often rings far worse than a well-tuned cheaper machine.
The Two Ways to Calibrate
There are basically two paths, and which one you take depends on your firmware.
Accelerometer-Based Measurement
This is the gold standard. You bolt a small accelerometer chip onto the toolhead, run a sweep, and let the firmware measure how the machine actually vibrates. Klipper’s resonance_tester is the best-known implementation, typically using an ADXL345 chip, though LIS2DW and LIS3DH variants are also supported.
The printer runs a series of movements across a frequency range, records the vibration response, and calculates resonance frequencies and damping ratios for each axis. The `SHAPER_CALIBRATE` command automates the sweep and recommends a shaper type and frequency per axis.
Manual Frequency Entry
Marlin added input shaping in its 2.0.x line and lets you enter frequencies manually or measure them with an accelerometer, depending on your build. Prusa firmware ships input shaping with per-printer tuned defaults on several models, so you may never touch it at all. RepRapFirmware on Duet boards also supports input shaping, though the calibration workflow differs from Klipper’s.
The takeaway: any article on this topic has to scope which firmware it covers, because the workflows are genuinely different. Most of what follows centers on Klipper, since it’s the most configurable and most discussed.
Choosing the Right Shaper Type
Klipper offers several shaper types, and they trade smoothing against vibration suppression. Here’s how they compare in practice:
| Shaper | Vibration Suppression | Smoothing Cost | Best For |
|---|---|---|---|
| zv | Moderate | Low | Clean, simple setups |
| mzv | Good | Low-moderate | General purpose |
| ei | Strong | Moderate | Heavier toolheads |
| 2hump_ei | Very strong | High | Stubborn, low-frequency ringing |
| 3hump_ei | Strongest | Highest | Extreme cases only |
The general rule: pick the shaper that gives the best suppression with the least smoothing. More humps means more vibration canceled, but also more rounding of corners. A `3hump_ei` on a well-behaved printer will make your parts look softer than they need to be.
Klipper’s auto-calibration usually picks a sensible default. Trust it first, then experiment if you’re chasing the last few percent of speed.
Step-by-Step Klipper Calibration
Here’s the practical sequence that works. Exact commands shift between Klipper releases, so check the official Klipper docs for your version before running anything.
- Mount the accelerometer securely. A loose sensor measures its own rattling, not your printer’s. Zip ties and hot glue are not good enough. Use a rigid mount.
- Check your belts and hardware first. Loose belts, wobbly carriages, and sloppy eccentric nuts will produce garbage data. Fix the mechanical stuff before calibrating.
- Run the resonance test. The `resonance_tester` sweeps each axis and reports the frequency response. You’re looking for clear peaks, not a noisy mess.
- Run `SHAPER_CALIBRATE`. This automates the sweep and recommends shaper type and frequency per axis.
- Save and restart. Whether results write to `printer.cfg` automatically has changed across versions, so verify your config actually updated.
- Print a ringing test tower. Confirm the improvement with a real print. Accelerometer data tells you what the frame does; a test print tells you what the plastic does.
That last step matters more than people think. Measurement is not a substitute for print tests, and many experienced users insist on both.
A Note on Firmware Versions
Klipper’s command syntax and supported accelerometer chips change between releases. Marlin’s input shaping implementation has evolved too. Prusa’s defaults are baked in per model. If a guide is more than a year old, cross-check the commands against current documentation before you type them into your terminal.
Common Mistakes That Wreck Calibration
Calibrating with loose hardware. This is the big one. If your belts are slack or your frame bolts are finger-tight, the accelerometer measures chaos. Fix mechanics first, always.
Copying someone else’s numbers. A shaper frequency of 48 Hz on one machine might be 62 Hz on yours. Belt tension, toolhead mass, and frame rigidity all shift results. Never paste config values from a forum post without measuring.
Ignoring the Y axis. People calibrate X, see improvement, and stop. Both axes need their own shaper values.
Over-smoothing with aggressive shapers. Slapping `3hump_ei` on everything rounds corners and hides detail. Start with what auto-calibration recommends.
Skipping the test print. Without a physical ringing tower, you have no idea whether the numbers translated into better prints.
Forgetting that speed changes the picture. Input shaping lets you push acceleration higher, but the values you calibrated at one acceleration may not be optimal at a much higher one. Recalibrate after major speed changes.
What Actually Works in Practice
The printers that print cleanest at speed share a few traits. Rigid frames. Properly tensioned belts. Light toolheads. Input shaping amplifies good hardware; it doesn’t rescue bad hardware.
If you’re just getting started and wondering which machine gives you a fighting chance, look for something with a solid frame and a firmware that supports calibration out of the box. A best beginner 3d printer with Klipper preinstalled saves you a weekend of flashing. If you’d rather learn the fundamentals on something simple, a solid best 3d printer for beginners that still allows firmware upgrades is a smarter long-term pick than a locked-down appliance.
Budget matters too. You don’t need to spend a lot to get a frame rigid enough for input shaping to work well, and a best budget 3d printer with a sturdy gantry often calibrates better than a flimsier expensive one. The same logic applies if you’re setting up a machine for a younger maker, where a best 3d printer for kids with sensible defaults and a rigid build plate will ring less out of the box.
The Practical Workflow
In real use, the sequence looks like this: build the printer, tension everything properly, print a calibration cube to check basics, mount the accelerometer, run the resonance test, apply the recommended shaper, print a ringing tower, compare, adjust if needed. Total time: an afternoon. The payoff lasts the life of the machine.
One more practical tip: keep a log. Write down your shaper type, frequencies, and the acceleration you tested at. When you change a belt or swap a hotend, you’ll know exactly what to recheck.
How Much Speed Can You Actually Gain?
Honest answer: it depends on the machine, but the gains are real. Printers that rang badly at 2,000 mm/s² acceleration often run clean at 5,000 mm/s² or higher after proper calibration. That translates to noticeably faster prints without the surface artifacts that used to make high-speed printing a tradeoff.
The catch is that input shaping only addresses vibration. It doesn’t fix extruder flow limits, cooling problems, or part shrinkage. Expect cleaner walls and faster corners, not miracles.
Conclusion
Input shaping calibration is one of those upgrades that feels intimidating until you do it once. The concept is simple: measure how your specific frame vibrates, then filter the motion commands so it stops ringing. The execution takes an afternoon and rewards you with cleaner prints at speeds that used to be off-limits.
Start with your mechanics. Mount the accelerometer rigidly. Run the sweep. Trust the recommendation, then verify with a real test print. Skip the forum copy-paste, and never assume your machine behaves like someone else’s.
The printers that benefit most are the ones with decent bones and a firmware that lets you calibrate. Whether you’re on a budget build or a serious machine, the workflow is the same, and the improvement shows up on the very first ringing tower.
Frequently Asked Questions
What is input shaping in 3D printing?
Input shaping is a firmware feature that filters motion commands before they reach the motors so the printer frame doesn’t vibrate after sharp direction changes. It splits movements into smaller timed commands whose vibrations cancel out, reducing ringing and letting you print faster with cleaner walls.
Do I need an accelerometer to calibrate input shaping?
Not always. Klipper’s best workflow uses an accelerometer like the ADXL345 for automatic measurement. Marlin supports manual frequency entry as well as accelerometer-based measurement depending on your build. Prusa firmware ships with per-printer tuned defaults, so many users never calibrate manually at all.
How long does input shaping calibration take?
The measurement itself takes a few minutes. The full process, including checking belts, mounting the accelerometer, running the sweep, applying settings, and printing a ringing test tower to confirm, usually takes an afternoon. Recalibration after hardware changes is much faster.
Can I copy input shaper values from another printer?
No. Resonant frequencies depend on belt tension, frame rigidity, toolhead mass, and accelerometer mounting. Two identical printer models can need different values. Copying config numbers from a forum post often makes print quality worse rather than better.
Why does my printer still ring after calibration?
Common causes include loose belts, a wobbly carriage, a poorly mounted accelerometer, or calibrating at an acceleration far from what you actually print at. Input shaping compensates for vibration, not mechanical slop. Fix the hardware first, then recalibrate.
Which shaper type should I use?
Start with whatever Klipper’s auto-calibration recommends, usually `mzv` or `ei` for typical setups. More aggressive shapers like `2hump_ei` and `3hump_ei` suppress more vibration but round corners more. Pick the least smoothing that gives clean prints.
