Print Speed vs Quality: Where Input Shaping Changes the Math

For most of FDM’s history, the speed-versus-quality trade was simple and unforgiving: print faster, get worse results. Push the speed up and flat surfaces sprouted wavy ripples, corners blurred, and dimensional accuracy drifted. Then a pair of motion-control techniques — input shaping and pressure advance — quietly rewrote that math. A well-tuned modern printer can run two to five times faster than a machine from a few years ago at comparable quality. This guide explains what’s actually happening, so you know when speed costs quality and when it doesn’t.

What Actually Limits Print Speed

Three different physical limits cap how fast a printer can move and still print well, and they bind at different points:

1. Vibration (ringing). As the toolhead accelerates and changes direction, the printer’s frame and gantry flex and oscillate. Those vibrations get stamped into the print surface as ripples — called ringing or ghosting — most visible just after sharp corners. This is the limit input shaping addresses.
2. Pressure lag in the nozzle. Molten plastic doesn’t respond instantly to extruder commands. When the printhead speeds up, pressure in the nozzle lags behind; when it slows for a corner, residual pressure keeps oozing. The result is over-extruded corners and under-extruded line starts. This is what pressure advance addresses.
3. Melt rate (volumetric flow). The hotend can only melt plastic so fast. Past that rate, the extruder can’t keep up and you get under-extrusion regardless of how good the motion control is. This is a hard physical ceiling set by the hotend.

Without electronic compensation, ringing and pressure lag force you to keep speeds low to preserve quality. Input shaping and pressure advance attack the first two, leaving melt rate as the remaining ceiling.

Ringing: The Classic Speed Defect

Ringing (also called ghosting) is the wavy, repeating ripple that appears on flat surfaces near corners and sharp features. It comes from mechanical resonance: every printer has natural frequencies at which it vibrates, and when the toolhead’s accelerations excite those frequencies, small inputs produce disproportionately large oscillations. The print surface records the wobble.

The traditional fixes were all about reducing the excitation: lower acceleration, lower speed, stiffen the frame, tighten the belts, reduce moving mass. Those still help — a loose belt or a wobbly frame defeats any software — but they cap your speed.

Input Shaping: Canceling the Vibration

Input shaping is a software technique that pre-filters the motion commands so the printer never excites its own resonant frequencies in the first place. Rather than sending one sharp acceleration command, the controller splits it into a precisely timed sequence whose vibrations cancel each other out — open-loop control tuned to the machine’s measured resonance.

The practical effect is large: input shaping lets printers hold high accelerations and speeds — commonly 200-500 mm/s on capable hardware — without the ringing those speeds would otherwise produce. It’s why a modern printer can run at 250 mm/s and still show clean flat surfaces, while an older machine ripples badly at half that.

Tuning input shaping requires measuring the printer’s resonance. The usual method uses an accelerometer to record the frequencies and amplitudes during a test routine, then picks a shaper and frequency that cancel them. On Klipper-based machines this is a documented calibration; many commercial printers now run their own resonance compensation automatically, which is why high-speed printing has become a baseline expectation rather than an exotic mod.

Pressure Advance: Sharp Corners at Speed

Pressure advance (the term in Klipper; “linear advance” in Marlin) compensates for the lag between extruder commands and actual plastic flow. It anticipates speed changes — easing pressure off slightly before a corner so the nozzle doesn’t keep oozing into the turn, and building pressure back up before a straight run so lines don’t start thin.

The visible payoff is consistent line width through corners and at the start and end of every extrusion, instead of bulging corners and gaps. Like input shaping, it’s tuned with a calibration print and is largely what makes fast printing dimensionally accurate rather than just fast.

What Still Costs Quality, Even With Compensation

Compensation isn’t magic. Some quality costs of speed remain no matter how well the printer is tuned:

• Layer adhesion at the melt-rate limit. Push volumetric flow past what the hotend can melt and you under-extrude — weak, gappy layers. High-flow hotends raise this ceiling, but it’s always there.
• Overhangs and bridges. Fast cooling and fast motion both hurt unsupported geometry. Slow these features down regardless of overall speed.
• Outer-wall finish. The visible surface benefits from slower, steadier motion. A common, effective strategy is to print outer walls slowly (30-50 mm/s) while letting infill and inner walls run at full speed — the eye only sees the outer wall.
• The first layer. Always slow, always. Speed gains here are never worth a failed print.

A Sensible Speed Strategy

You don’t have to choose one global speed. The best results come from spending speed where it doesn’t show and saving quality where it does:

1. Calibrate input shaping and pressure advance first if your printer supports them. This raises the whole quality-vs-speed curve — you get more of both.
2. Run infill and inner walls fast. Nobody sees them.
3. Slow the outer wall to 30-50 mm/s for clean visible surfaces.
4. Slow overhangs, bridges, and the first layer for their own reasons.
5. Respect the melt-rate ceiling. If layers look starved, you’ve outrun the hotend, not the motion system — back off flow or upgrade the hotend.

The Bottom Line

The old rule that speed always costs quality is now only half true. On a printer with input shaping and pressure advance properly tuned, a large fraction of the speed-quality trade has been engineered away — the printer moves fast without ripples and turns corners without blobbing. What remains is real but narrow: melt rate, overhangs, and the visible outer surface. Tune the compensation, then print the parts that don’t show fast and the parts that show carefully. For where these settings live in a real slicer, fdmdesk’s OrcaSlicer settings walkthrough ↗ maps the relevant toggles, and the slicer ↗ you choose determines how its calibration tools expose them.

For more context, Bambu Lab printer reviews ↗ covers related topics in depth.