Ringing is the little echo that makes a sharp 3D print corner look like it saw a ghost. If your printer leaves ripples after edges but you do not own an accelerometer, you can still find usable resonance numbers today with a printed test, careful lighting, and one calm round of settings changes. In about 15 minutes, you can learn whether your X or Y axis is the noisy violin string, choose a safer input shaping value, and avoid the classic trap: “fixing” vibration by randomly slowing everything until joy leaves the room.
What Input Shaping Actually Does
Input shaping is a motion-control trick that reduces ringing, ghosting, echo lines, or ripples after sharp direction changes. Instead of pretending your printer is perfectly rigid, input shaping accepts the awkward truth: a printer frame, belt, toolhead, bed, and gantry behave like a tiny musical instrument.
Move too suddenly, and the instrument sings. The print surface records the song as faint repeated shadows after corners.
The goal is not to make the printer silent or magical. The goal is to time the motion command so the printer’s own vibration gets damped before it leaves a visible scar. Klipper calls this resonance compensation. Marlin supports input shaping through M593. Prusa Research also ships tuned profiles on newer machines because high-speed printing without vibration control can turn crisp edges into topographic poetry.
I once watched a printer make beautiful calibration cubes at 50 mm/s, then produce a ringing tower at 120 mm/s that looked like a tiny apartment building during an earthquake drill. Nothing was “broken.” The machine simply had an opinion about speed.
Why resonance matters more at high acceleration
Speed gets blamed often, but acceleration is usually the sharper tool. A printer moving at 150 mm/s in a long straight line may look fine. The trouble begins when it has to brake, turn, and launch again. That is where the belts twitch, the frame answers, and the nozzle arrives slightly haunted.
Input shaping targets that launch-and-catch moment. Better settings can let you keep useful speed while reducing visible ripples.
What an accelerometer does, and what you can still do without one
An accelerometer measures vibration directly. That is the tidy laboratory path. Without one, you use the printed part as the witness. You print a test with known speed and acceleration, inspect the distance between ringing marks, and estimate the resonance frequency.
Is it perfect? No. Is it useful? Absolutely. The print surface is a little courtroom. It does not tell every secret, but it tells enough to stop guessing wildly.
- Ringing after corners usually points to axis resonance.
- Acceleration changes the defect more than ordinary print speed.
- A clean test beats ten random setting changes.
Apply in 60 seconds: Pick one problem axis, X or Y, and commit to testing only that axis first.
Safety Before Testing Fast Motion
Input shaping tests often use fast moves, sharp corners, and higher acceleration than your usual print profile. That is not terrifying, but it deserves respect. A 3D printer is a heated machine with moving parts, belts, fans, stepper motors, wiring, and sometimes glass. It is not a scented candle with firmware.
Before testing, stay near the printer for the first few minutes. Keep loose sleeves, hair, tools, and cables away from moving axes. Do not run aggressive acceleration tests on a machine with loose belts, grinding rails, damaged wiring, or a wobbly table that behaves like a trampoline wearing shoes.
OSHA’s general machine-safety mindset is useful here: control hazards before you start, not after something sounds expensive. You do not need an industrial checklist for a desktop printer, but you do need common sense with its boots on.
Pre-test safety checklist
- Belts: Tension should be firm, not guitar-string dramatic.
- Toolhead: No wobble when gently nudged with power off.
- Bed: Clips, magnets, or fasteners should be secure.
- Wires: No snagging during full travel.
- Table: Stable enough that normal travel does not make it sway.
- Temperature: Hotend and bed settings should match the filament, not a memory from last winter.
Decision card: should you test now?
Decision Card: Green Light, Yellow Light, Red Light
| Status | What You See | Next Move |
|---|---|---|
| Green | Printer is mechanically sound and already prints cleanly at moderate speeds. | Run the manual ringing test. |
| Yellow | Minor ringing, slightly loose belt, or table wobble. | Fix the mechanical issue first, then test. |
| Red | Layer shifts, skipped steps, motor overheating, grinding, or wiring snag. | Stop. Repair before running high-acceleration moves. |
If your stepper motors are already running hot, resonance testing may expose the problem. This related guide on stepper motor overheating pairs well with input shaping work because both affect reliability during faster motion.
Who This Is For and Not For
This guide is for the printer owner who sees ringing on vertical walls, wants cleaner corners, and does not want to buy an accelerometer yet. It is also for the practical tinkerer who would rather spend one careful hour printing tests than three chaotic evenings changing jerk, acceleration, belt tension, slicer speed, pressure advance, and desk placement all at once. We have all met that gremlin. Sometimes it is us.
This is for you if...
- You use Klipper, Marlin, RepRapFirmware, or a slicer/profile that supports input shaping or resonance control.
- You see repeated ripples after corners, embossed letters, holes, or sudden direction changes.
- Your printer is mechanically healthy enough to run moderate acceleration tests.
- You can measure small distances on a print using calipers, a ruler, or a well-lit close-up photo.
- You prefer a practical value over a perfect lab graph.
This is not for you if...
- Your printer has obvious mechanical faults, such as skipping, grinding, loose pulleys, or frame movement.
- You need production-grade validation across many machines.
- You are tuning a very fast CoreXY or toolchanger where accelerometer data would save time.
- Your main defect is stringing, blobs, under-extrusion, or moisture, not ringing.
I once blamed resonance for a messy logo print. The culprit was stringing around tiny travel moves. Different dragon, different sword. If your issue appears mostly on logos, read this piece on PETG stringing on logos before rewriting your motion settings.
Visual Guide: Manual Resonance Tuning Flow
Check belts, frame, table, pulleys, and hotend mount.
Run a ringing tower or sharp-corner test at known speed.
Find the spacing between repeating waves after a corner.
Convert spacing and speed into an approximate frequency.
Set conservative input shaping values in firmware.
Print again and compare corners under the same lighting.
The Manual Resonance Method Without an Accelerometer
The manual method is simple: print a part that makes resonance visible, measure the spacing between the repeated ripples, and calculate frequency from movement speed. You are not measuring the printer’s vibration directly. You are measuring the visible trace it leaves behind.
The basic idea is:
Frequency in Hz = print speed in mm/s ÷ ringing spacing in mm
If the printer moves at 100 mm/s and the ripples repeat every 2.5 mm, the estimated resonance is 40 Hz. That is not a divine prophecy engraved on a nozzle. It is a starting value.
The cleanest manual workflow
- Choose one test model with sharp corners and flat vertical walls.
- Use a known external perimeter speed, such as 80, 100, or 120 mm/s.
- Use enough acceleration to reveal ringing, often 3000 to 7000 mm/s² depending on machine type.
- Print with input shaping disabled first.
- Measure the repeated ripple spacing after a corner.
- Calculate the approximate frequency.
- Enable input shaping with a conservative shaper type.
- Print the same test again and compare.
For many bedslingers, manual estimates land somewhere around the 30 to 60 Hz zone. For stiffer CoreXY machines, values may be higher. But do not force your printer into someone else’s number. Borrowing resonance values from the internet is like wearing another person’s glasses while soldering.
Risk scorecard: manual tuning confidence
Risk Scorecard: How Much Should You Trust the Result?
| Signal | Low Risk | Higher Risk |
|---|---|---|
| Ripple visibility | Clear, repeated, evenly spaced marks | Random texture, blobs, seams, or inconsistent waves |
| Test repeatability | Same pattern appears on a second print | Pattern changes wildly each print |
| Mechanical condition | Belts, pulleys, rails, and frame feel stable | Loose hardware, wobble, rubbing, or table shake |
| Measurement quality | Measured several waves, then averaged | Measured one tiny mark under poor lighting |
Print Test Setup That Gives Clean Clues
A resonance test is only useful if it isolates the motion problem. If your filament is wet, nozzle is too hot, seam placement is chaotic, and pressure advance is untuned, the surface becomes a noisy committee meeting. Nobody leaves with a decision.
Start with a plain filament you know well. PLA is easiest for most users because it prints cleanly and shows surface defects without the drama PETG sometimes brings to the table. Use a matte or light-colored filament if possible. Glossy black filament can hide ringing until you tilt it under a lamp and suddenly regret your optimism.
Recommended starting settings
| Setting | Starting Point | Why It Helps |
|---|---|---|
| Filament | PLA or familiar PETG | Reduces material surprises. |
| Layer height | 0.20 mm | Easy to inspect without excessive print time. |
| Perimeters | 2 walls | Keeps walls consistent. |
| External perimeter speed | 80 to 120 mm/s | Fast enough to reveal ringing on many machines. |
| Acceleration | 3000 to 7000 mm/s² | Shows resonance without going straight to chaos mode. |
| Cooling | Normal PLA cooling | Avoids heat-softened corners that mimic motion issues. |
Use one model, one slicer profile, one lighting setup
Print the same test twice: once without input shaping and once with your estimated value. Change only the input shaping setting. This is the boring part, which means it is probably the part that works.
I like taking a phone photo from the same angle before changing settings. The camera becomes a witness with no emotional attachment to your “I think it looks better” moment.
If layer changes add blobs exactly where you are measuring ringing, clean that up first. This guide on blobs at layer change helps separate seam artifacts from real vibration.
- Use known filament and stable temperatures.
- Keep the same test model and speed.
- Photograph before and after under identical light.
Apply in 60 seconds: Save a dedicated slicer profile named “Ringing Test” so future checks do not become archaeology.
How to Read Ringing Marks Like a Mechanic
Ringing usually appears as repeating vertical waves after a sharp corner, hole, letter, or sudden toolpath change. The marks gradually fade as the vibration dies out. That fading pattern is your clue.
Look along the direction of travel after the corner. If the nozzle exits a corner and ripples appear on the following wall, measure from one ripple peak to the next. Better yet, measure across several ripples and divide by the number of spaces. Five waves tell a calmer story than one.
Which axis is ringing?
On a cube or tower, each wall reflects different motion. If the ripple shows on a wall produced mostly by X-axis movement, it may point to X resonance. If it appears on the wall produced mostly by Y-axis movement, it may point to Y resonance. Bedslingers often show stronger Y-axis ringing because the bed moves mass back and forth. CoreXY printers can be more subtle because both motors contribute to gantry motion.
Put a tiny X and Y mark on the test part before you remove it from the bed, or note orientation in a photo. Future you will thank present you with imaginary pastries.
How to measure the spacing
- Find a flat wall after a sharp corner.
- Identify at least 4 to 6 repeated wave peaks.
- Measure from the first peak to the last peak.
- Divide the total distance by the number of gaps.
- Use that average spacing in the frequency formula.
Example: you measure 12 mm across 5 gaps. Average spacing is 2.4 mm. If the wall was printed at 100 mm/s, estimated frequency is 100 ÷ 2.4 = 41.7 Hz.
Short Story: The Cube That Lied Twice
A maker I know printed a fast cube and saw ripples on two sides. He tuned X, then Y, then X again. Each print looked slightly different. After an hour, the bench looked like a museum of small white regret. The real problem was not the calculation. His printer sat on a lightweight folding table beside a washing machine. When the printer accelerated, the table flexed, the washer panel buzzed, and the whole setup joined the performance. He moved the printer to a heavy workbench, tightened one idler screw, and printed the same tower again. The ringing became clean, measurable, and boring. His first good input shaping value appeared in one print. The lesson is unglamorous but golden: do not tune firmware around furniture. Make the machine stable first, then let the test speak.
Mini Calculator: Estimate Resonance From Ringing Spacing
This small calculator gives a practical frequency estimate from print speed and measured ripple spacing. It is meant for manual tuning, not laboratory validation. Use it to choose a starting value, then verify with another print.
Mini Calculator: Manual Resonance Estimate
Formula: frequency = speed ÷ spacing
Estimated resonance will appear here.
How to choose the first value
Round to a sensible number. If the calculator says 41.7 Hz, start around 42 Hz. If your measurement is rough, do not pretend the decimal point is sacred. A manual print method might justify 42 Hz, not 41.738 Hz with a tiny crown on it.
If you get two close estimates from repeated prints, average them. If one estimate is wildly different, inspect the test again for seam blobs, cooling defects, or measurement error.
Eligibility checklist: your result is ready to apply if...
- The same ripple spacing appears on at least two nearby features.
- You measured across multiple waves, not one mark.
- The print speed used in the calculation matches the actual external wall speed.
- Input shaping was disabled during the baseline test.
- The printer did not skip steps, buzz violently, or shift layers.
Firmware Settings for Klipper, Marlin, and Slicers
Different firmware handles input shaping differently. The principle is shared, but the knobs vary. Always check your firmware’s official docs before editing configuration. A typo in a config file can be more exciting than anyone requested.
Klipper: shaper frequency and shaper type
Klipper commonly uses values such as shaper_freq_x and shaper_freq_y, plus a shaper type such as ZV, MZV, EI, or 2HUMP_EI. Manual tuning often starts with a conservative type and the calculated frequency per axis. Klipper’s own documentation explains that input shaping can reduce ringing and vibration, and it also warns that tuning and measurement matter.
A cautious Klipper-style mindset:
- Set X and Y separately if your test shows different frequencies.
- Use conservative acceleration limits after enabling shaping.
- Verify with a repeat print before raising speed.
- Save your old config before edits.
Marlin: M593 input shaping
Marlin uses M593 for ZV input shaping and related parameters depending on build support. Many users test with calibration patterns, then set damping frequency values for X and Y. The important point is the same: do not copy a random number because it made someone else’s Ender clone look heroic in a forum post.
For Marlin users, confirm that your build supports input shaping and that the command is active. Then test with a known pattern, apply one axis value at a time, and save settings only after verification.
Slicer profiles and factory input shaping
Some newer printers ship with pre-tuned input shaping in the firmware or slicer profile. That can be excellent, but it does not erase mechanical reality. Belt tension, added toolhead weight, enclosure panels, worn wheels, table stiffness, and modified hotends can change the response.
If your printer came with factory profiles, do not overwrite them casually. Create a copy. Tune the copy. Keep the known-good profile like a spare house key.
Show me the nerdy details
Manual ringing measurement estimates the vibration frequency from the spatial period left on the wall. Because the print head travels at a known surface speed, the distance between waves maps to time. If waves repeat every 2.5 mm at 100 mm/s, each wave cycle takes 0.025 seconds, which equals 40 cycles per second. Input shaping then modifies command timing to reduce excitation near that frequency. The tradeoff is smoothing: stronger shapers can reduce ringing more broadly but may soften corners or require lower maximum acceleration. This is why the “best” value is not always the most aggressive value.
Common Mistakes That Make Input Shaping Worse
The most common mistake is treating input shaping like seasoning. A little number here, a little acceleration there, some jerk, a dash of belt tension, and soon the profile tastes like confusion. Tune in order. Mechanics first. Baseline print second. Measurement third. Firmware fourth. Verification fifth.
Mistake 1: tuning around loose belts
Loose belts can cause ghosting, dimensional errors, and inconsistent ringing. Over-tight belts can add friction and stress bearings. Both can poison the test. Belt tension should feel firm and even, not heroic.
Factory guidance from printer makers such as Prusa Research often treats belt tension as a quality and reliability issue, not a cosmetic tweak. That is the right attitude.
Mistake 2: confusing pressure advance with input shaping
Pressure advance helps manage extrusion pressure during speed changes. Input shaping helps manage mechanical vibration during motion changes. They often improve different defects. Bulging corners may need pressure advance. Repeated waves after corners may need input shaping.
In my own tuning notes, I write “shape motion, then tune pressure” unless the extrusion is so messy that surface reading becomes impossible. It keeps the dragons in separate cages.
Mistake 3: using the wrong speed in the calculation
The formula uses the actual speed of the wall where ringing is measured. If your slicer says 120 mm/s but minimum layer time, cooling slowdown, or volumetric limits reduce that wall to 60 mm/s, your estimate will be wrong.
Check the sliced preview. If your slicer shows actual speeds by color, use that. If your machine cannot maintain the commanded flow, solve that first. This related article on volumetric flow limit testing is useful when high-speed walls look underfed rather than merely wavy.
Mistake 4: chasing every tiny surface mark
Not every surface flaw is resonance. Zits, seams, glossy bands, moisture bubbles, inconsistent cooling, and extrusion pulses can all pretend to be motion defects. Input shaping cannot dry filament, flatten a warped bed, or convince a slicer seam to behave at a dinner party.
- Do not tune around loose hardware.
- Do not use guessed speed in the frequency formula.
- Do not mistake seams or wet filament for resonance.
Apply in 60 seconds: Before your next test, write down the exact speed, acceleration, filament, and firmware state.
Troubleshooting Map: What the Pattern Means
A print defect is a message, but it rarely speaks in full sentences. Use the pattern, location, and repeatability to decide what to test next.
Comparison table: ringing versus similar defects
| Defect | What It Looks Like | Likely Cause | First Fix |
|---|---|---|---|
| Ringing | Evenly spaced fading waves after corners | Axis resonance | Manual frequency estimate and input shaping |
| Layer shift | Entire layer suddenly offset | Skipped steps, loose pulley, crash, motor issue | Stop high-speed testing and inspect mechanics |
| Seam zits | Dots or blobs aligned at start/end points | Retraction, pressure, wipe, coasting, seam placement | Tune seam and extrusion behavior |
| Gloss bands | Horizontal finish changes | Speed, cooling, temperature, flow variation | Check slicer speed preview and cooling |
| Diagonal under-extrusion | Thin or weak lines on certain angles | Flow limit, acceleration, pressure behavior | Reduce speed or test flow capacity |
If ringing appears only at certain speeds, the machine may be exciting a narrow resonance band. The related article ringing only at certain speeds is a strong companion read because it focuses on speed bands, not just one calibration value.
Coverage tier map: what level of tuning do you need?
Tuning Tier Map
Manual ringing test, one value per axis, conservative acceleration. Best for hobby printing.
Repeat tests at two speeds, compare shaper types, verify with real parts. Best for frequent printing.
Accelerometer testing, full frequency sweep, profile documentation. Best for farms, paid parts, or heavy mods.
When to Seek Help
Manual tuning is practical, but there are moments when help saves time, parts, and nerves. If the printer makes harsh mechanical noises, shifts layers during ordinary prints, overheats motors, or shows inconsistent results after basic checks, stop tuning input shaping and inspect the machine.
A local makerspace, experienced printer technician, or active firmware community can often spot a loose pulley or frame issue faster than another five test towers can. There is honor in asking. The printer will not write a poem about your humility, but it may start printing straight walls.
Seek help before more high-speed tests if...
- Layer shifts happen during the ringing tower.
- The toolhead or bed physically collides with anything.
- A belt frays, jumps teeth, or tracks against a flange.
- Motors become too hot to touch briefly.
- Wiring rubs, snags, or flexes sharply during travel.
- The frame visibly twists or rocks during acceleration.
What to prepare before asking for help
- Printer model and firmware.
- Photos of the test print from two angles.
- Slicer profile speed and acceleration values.
- Whether input shaping was enabled or disabled.
- Measured ripple spacing and calculated frequency.
- Any recent mods, such as direct drive, heavy fan duct, linear rails, enclosure, or new bed surface.
- Layer shifts and harsh noises are repair clues.
- Hot motors can signal electrical or mechanical stress.
- Good help starts with photos, settings, and repeatable symptoms.
Apply in 60 seconds: Take one clear side photo of your test print and one photo of the printer setup before changing anything else.
FAQ
Can you tune input shaping without an accelerometer?
Yes. You can estimate resonance from a ringing test print by measuring the spacing between repeated ripples and dividing print speed by that spacing. It is less precise than accelerometer testing, but it can produce a useful starting value for hobby printers.
What is the formula for finding resonance from ringing?
The practical formula is frequency in Hz equals wall speed in mm/s divided by ringing spacing in mm. For example, 100 mm/s divided by 2.5 mm equals 40 Hz. Use the actual wall speed where the marks appear, not just the headline speed in your slicer profile.
Is input shaping the same as pressure advance?
No. Input shaping reduces mechanical vibration from motion changes. Pressure advance manages extrusion pressure during speed changes. Ringing after corners usually points toward input shaping. Bulging corners or inconsistent extrusion near speed changes may point toward pressure advance.
Why did input shaping make my print look worse?
The frequency may be wrong, the shaper type may be too aggressive, acceleration may be too high, or the real issue may be mechanical. Loose belts, a wobbly table, worn wheels, loose pulleys, or a heavy modified toolhead can make input shaping look guilty when the machine is the actual witness.
Should I tune X and Y separately?
Usually, yes. X and Y axes often have different moving mass and stiffness. A bedslinger may show stronger Y-axis resonance because the bed moves. A CoreXY machine may still need different values due to belt path, toolhead weight, and frame behavior.
What ringing test speed should I use?
Use a speed high enough to reveal ringing but not so high that extrusion fails. Many users start around 80 to 120 mm/s for external walls, with acceleration in a moderate test range for the machine. If the wall under-extrudes or overheats the motion system, reduce the test demand.
Do factory input shaping profiles need retuning?
Sometimes. Factory profiles are useful, but resonance can change when you alter belt tension, add a heavier toolhead, move the printer to a different table, install an enclosure, or change motion parts. Retest after major mechanical changes.
Can input shaping remove all ringing?
Not always. It can reduce ringing, often dramatically, but it cannot fix loose hardware, poor extrusion, wet filament, unstable furniture, or excessive acceleration. The cleanest prints usually come from both sound mechanics and sensible motion settings.
Conclusion: The 15-Minute Resonance Check
The ghost in your print corner is not mysterious once you give it a ruler. Without an accelerometer, you can still run a controlled ringing test, measure wave spacing, estimate resonance, and apply a careful input shaping value. The trick is not bravery. It is patience with a caliper.
Your next step within 15 minutes: print or inspect one sharp-corner test, measure several ripple gaps on the clearest wall, and calculate speed divided by spacing. Save the original profile first. Then make one change, print again, and compare under the same light. That small loop is where better walls begin.
Last reviewed: 2026-07
