How to Fix Layer Lines and Banding on Resin Prints
Layer lines, banding, and layer separation are three different defects with three different causes. Here's how to tell them apart by measuring the spacing, and how to fix each one in the right order.
This guide contains affiliate links: we may earn a small commission at no extra cost to you.
You pull a model off the plate, wash it, cure it, and under the desk lamp there they are: horizontal stripes running across the cloak or the shield where the surface should be smooth. It is the most common complaint about resin prints that otherwise came out fine, and it is also the one most often fixed in the wrong order, because “layer lines” gets used as a catch-all for three genuinely different defects.
The three are worth separating before you touch a setting. Fine, uniform, spaced exactly at your layer height: those are layer lines, and they are an aesthetic problem solved with orientation and layer height. Distinct ridges or tonal stripes spaced whole millimetres apart: that is banding, and it is a mechanical, thermal, or lift-speed problem that a finer layer height will not touch. A gap you can catch a fingernail in: that is layer separation, and it is under-exposure. If you are new to the whole workflow, the complete beginner guide to resin printing miniatures covers the underlying process this guide assumes, and the general troubleshooting guide covers the other nine ways a print goes wrong.
This guide is the deep version of one of those failures. Work top to bottom: diagnosis first, then the free fixes, then the ones that cost print time, then the ones that cost money.
Diagnose It Before You Fix It
Put the model under a bright lamp and rotate it slowly so the light rakes across the surface at a low angle. Raking light is the whole diagnostic; straight-on light hides all three defects. Then answer one question: how far apart are the marks?
| What you see | Spacing | What it is | Where the fix lives |
|---|---|---|---|
| Fine uniform grain, only on shallow surfaces | Matches layer height (0.05mm, 0.03mm) | Layer lines | Orientation, layer height, anti-aliasing |
| Stair-stepping on curves, worse on faces and domes | Matches layer height, follows curvature | Aliasing / stepping | Anti-aliasing, layer height |
| Distinct ridges or tonal stripes | Whole millimetres, regular pitch | Mechanical banding | Z axis, plate, FEP tension |
| Stripes across one height band only | Irregular, confined to a section | Thermal or lift banding | Temperature, lift speed, rest time |
| A step or crack you can feel with a nail | Anywhere, often at one layer | Layer separation | Exposure, FEP condition |
Two things make this measurement easier. A magnifier headset turns “some kind of lines” into a spacing you can actually judge, and looking at the same surface on two models from the same plate tells you whether the cause is the model or the printer. If every model on the plate has the bands at the same height, the printer or the room did it. If one model has them and its neighbours do not, that model’s orientation did it.
Handle prints with nitrile gloves while you are inspecting anything that has not been washed and cured yet. Uncured resin is a sensitiser and the resin safety guide covers handling properly.
Rule Out Layer Separation First
This is the one that matters structurally, so clear it before spending time on cosmetics.
Layer separation is not texture. It is a physical gap where adjacent layers failed to fuse, and the model is weak at that point even if it survived the wash. Run a fingernail across the defect: texture catches nothing, a gap catches. Under magnification, separation often shows as a clean crack following one layer rather than a repeating pattern following many.
The cause is almost always that the resin did not receive enough UV to bond one layer to the next. Two things starve it:
Exposure too low for the resin. Every resin has its own exposure window, and the window shifts with temperature and with how old the bottle is. Do not guess at a number. Run a proper test and read it, which is what the exposure test and calibration guide is for.
A degraded FEP. A film that is fogged, scratched, or scattered with cured resin flakes diffuses UV before it reaches the vat, so a nominally correct exposure arrives weaker than the slicer thinks. Hold the vat up to a light and look through the film. Cloudy is a replacement, and the FEP replacement guide covers the procedure and tensioning. While the vat is out, it is worth cleaning it properly rather than just wiping it.
Only once you are confident every layer is bonded should you start chasing surface quality.
The Free Fix: Orientation
Before any setting changes, look at how the model was placed.
Layer lines are a geometry problem before they are a settings problem. On a surface standing near-vertical, each layer steps sideways by almost nothing and the lines are invisible. On a surface lying at a shallow angle to the plate, each layer steps sideways by a comparatively long distance, and those steps read as visible banding across the whole panel. Same printer, same resin, same layer height, completely different result.
That means the flat surfaces you care about most, such as shields, cloaks, vehicle panels, and tabletops on terrain, should be rotated to sit steeper rather than flatter. Most humanoid miniatures land in a 30 to 45 degree range for exactly this reason: it keeps broad surfaces off the shallow angles and pushes support scars onto hidden faces at the same time. The orientation guide has the full angle-by-angle breakdown, including the trade-off against suction forces and support count.
This costs nothing and no print time, which is why it goes first. It also fixes the specific case people most often try to solve with a finer layer height: a shield that came out striped is usually a shield that was lying down.
Layer Height and Anti-Aliasing
Once orientation is right, layer height sets the floor on how fine the remaining lines can be.
0.05mm is the working default for infantry, rank-and-file, and anything that will be seen at table distance. It is fast enough to batch a full plate overnight and the lines are not visible at arm’s length.
0.03mm is for display pieces and for models dominated by broad curved surfaces, such as busts, faces, and rounded armour, where stepping shows most. The cost is real: dropping layer height by 40 percent adds roughly that much to print time, since the printer is doing more layers of the same work.
0.1mm is not a miniatures setting. It exists for large functional parts. On a 32mm model it produces exactly the striping this guide is about.
Anti-aliasing is a separate control and it addresses a different artifact. It softens the hard pixel edges on the LCD so that curves step less abruptly, which helps most on faces, domes, and gentle slopes. On 4K-class printers a low to moderate setting is generally worthwhile. On 8K and 12K screens the pixel pitch is already fine enough that manufacturers commonly recommend a low setting or none, because heavy anti-aliasing there softens genuine detail without a visible gain. Follow the guidance for your specific printer rather than a number from a forum post about a different machine.
Both settings live in your slicer profile. The ChiTuBox settings guide and the Lychee settings guide cover where each control sits and what else moves with it.
Thermal Banding: Bands That Only Cover Part of the Print
This is the case that confuses people most, because the printer is clearly capable of a clean surface: it produced one on the lower half of the same model.
Resin viscosity is temperature-dependent, and viscosity determines how quickly resin flows back under the model after each peel. Cold resin flows back slowly. If the printer does not wait long enough, the next layer cures against an incompletely refilled gap, and the surface records it. A room that drifts several degrees overnight produces bands at the heights corresponding to the coldest hours, which is why this defect so often appears as a stripe across one section of every model on the plate.
The fix is stability rather than a specific number. Bring the resin and the room to a temperature inside your resin manufacturer’s stated range, commonly somewhere around 25C, and hold it there for the whole print. A stable 22C beats a 25C that fell to 16C at 3am. Printing in a garage, a basement, or an unheated spare room in winter is the classic trigger, and printing resin in cold weather covers enclosure, heating, and resin warming in detail.
Two related causes worth checking at the same time:
Resin level dropping through the print. As the vat empties, the fluid dynamics of the peel change. Topping up the vat before a long print removes that variable.
Settled pigment. Resin that has sat unshaken separates, and pigment-heavy resin at the bottom of the vat cures differently from the thinner resin above it. Shake the bottle thoroughly before pouring and stir the vat gently before a long print.
Mechanical Banding: Lift, Rest, and the Z Axis
If the bands are evenly pitched over the entire height of the model, and they survive both a temperature fix and a layer height change, the printer’s movement is producing them.
Work through these in order, one change per test print. Changing three things at once and getting a clean print teaches you nothing.
Slow the lift speed. A fast lift peels the model off the FEP violently and pulls resin away faster than it flows back. Slower lift is the single most reliable change against mechanical banding. It costs print time proportionally, which is the trade you are making.
Add rest or wait time before exposure. Most slicers expose a delay between the plate reaching position and the LCD firing. A short wait lets resin settle and the plate stop oscillating. On tall models and on resins that run thick, this matters more than the lift speed itself.
Check the build plate is genuinely tight. A plate that flexes or shifts even slightly under peel force writes that movement into the surface. Confirm the clamp is fully tightened and the plate has no play when you push on a corner. The pre-print checklist covers this along with everything else worth confirming before a long run.
Check FEP tension. A slack film stretches further before releasing, and an over-tight one transmits shock. Manufacturers specify tension for a reason; an evenly tensioned film peels consistently and that consistency is what you are chasing.
Clean and lubricate the Z lead screw. Resin dust and old grease make movement inconsistent. Clean the screw and apply the lubricant your printer manufacturer specifies, not a generic one. Bands at a consistent multi-millimetre pitch that survive every other change are the signature of Z movement, because that pitch corresponds to the screw’s own geometry.
Hollowing changes peel forces. A solid model has a large cross-section fighting the FEP on every layer. A hollowed model with proper drain holes peels with far less force, and less force means less of everything described above.
Fixing Lines on a Model You Already Printed
Sometimes the print is otherwise good and you would rather rescue it than spend another four hours.
Wet-sand mild lines. 400 to 600 grit wet/dry paper, used wet, smooths layer structure on broad flat surfaces without destroying the surface. Sand lightly and check often.
Do not sand fine detail. Chainmail, faces, fur, scrollwork, and lettering all lose their definition long before the layer lines go. On those areas, accept the texture and let paint hide it.
Use a filler primer. A slightly heavier-bodied primer sits in fine surface texture and levels it, where a thin coat follows the texture and leaves the lines showing through. Tamiya Fine Surface Primer is the usual choice for this, and brush-on primers such as Vallejo Surface Primer or Stynylrez work for spot treatment on one panel. Build it in thin coats: a thick coat that fills lines also fills detail.
Adjust the paint job. Texture and matte finishes hide fine lines; smooth blends and gloss show them. If a shield still reads as striped, a texture, a freehand device, or battle damage over that panel solves it faster than more sanding. The painting guide covers surface prep in more depth.
Sanding always beats reprinting on time, and reprinting always beats sanding on result. The dividing line is roughly whether you can still feel the lines after one pass of primer.
Confirm the Fix on a Test Piece
Do not validate a settings change on the model you actually want. Print something small with a broad angled flat surface, or one of the purpose-made banding test models circulating on the STL sites, and read the same surface each time under the same raking light.
One variable per print, written down. Lift speed, then rest time, then temperature. What you are building is a settings profile for that specific resin on that specific printer, which is worth far more than a single clean print, because it survives your next bottle and your next model. Resins differ enough that a profile dialled in for a standard grey will not transfer unchanged to an ABS-like or a water-washable resin.
The Short Version
Measure the spacing before you change anything. A gap you can feel is exposure, and it goes first. Fine uniform grain is orientation and layer height. Millimetre-pitch ridges are the machine: temperature, lift speed, rest time, then the Z axis, in that order and one at a time.
The mistake worth avoiding is the expensive reflex: dropping to 0.03mm and reprinting because a surface looked striped. If the cause was a cold room at 3am or a shield lying flat on the plate, that reprint takes 60 percent longer and comes out striped in exactly the same place.