Halftone Moire in Screen Printing: The Angles and Mesh Counts That Actually Prevent It
Mesh count too close to your LPI, and a halftone angle that matches the mesh grid, causes screen-printing moire. Here is the formula and angles that fix it.
01/ ARTICLE
Introduction
Moire on a screen-printed halftone comes from two settings fighting the mesh at once: a mesh count too close to your LPI, and a screen angle that lines up with the mesh's own grid. Fix both together and the pattern disappears. Fix only one, and it's back the moment you change garment, ink, or press.
This is the mesh-and-fabric version of the problem. Prepping a file for a riso drum instead of a screen? The fix is different, because the interference source is different -- drum resolution, not woven mesh -- and that's covered in the riso version of this problem.
This guide is for the print-maker sending a halftone file to a screen printing press. It covers the mesh-count formula that actually works, the angles that hold up under a squeegee, the dot shape that helps, and three easy-to-skip factors -- emulsion thickness, tension, thread diameter -- that will undo a correct angle-and-mesh choice on their own.
What Mesh Count Do You Actually Need?
Mesh count in screen printing should be roughly four to five times your halftone's LPI. Take the line count you want printed, multiply by a factor between 3 and 5, and that's your target mesh count in threads per inch. A practical working minimum: 4.5x (1.8x if you're specifying in metric threads-per-cm).
Here's the math: a 45 LPI halftone wants at least a 200-count screen. 45 x 4.5 = 202.5, rounding to a 195 or 200 mesh in stock sizes. Go lower and detail breaks down fast -- dots fill in, disappear, or print inconsistently because the mesh openings are too coarse to hold the smaller dot sizes in your highlights.
One detail most printers skip: the ratio should land on a number that isn't a clean whole multiple. A mesh count that's an exact 4.0x or 5.0x multiple of your LPI still shows residual moire, even when it's technically "in range." Nudge the mesh count off that clean multiple -- 195 instead of 200, say -- and you reduce whatever interference is left.
The number printed on the mesh box isn't the number you'll actually print with, either. Mesh elongates 4-8 percent once it's stretched to operating tension on the frame. A 195-count mesh tensioned to a standard 34 N/cm can open up to roughly 183 threads per inch once it's mounted and ready to go. Buy mesh based on the raw formula without accounting for that drop, and you can end up under-meshed for your LPI without touching a single setting. Buy a step higher than the formula suggests if you're tensioning the screen yourself.
There's also a hard floor on how small a dot the screen can physically hold, independent of LPI: minimum printable dot size equals two times the thread diameter plus two times the mesh opening distance. That caps your achievable tonal range before angle or dot shape ever enters the picture.
Which Screen Angle Actually Prevents Moire?
Screen mesh is woven in a strict grid, running vertically and horizontally. Halftone dots placed at 0, 90, 180, or 270 degrees sit directly on that grid, and the two patterns interfere in a way the eye reads as a slow, wavy moire across the print. This happens no matter how well your mesh count matches your LPI -- angle and mesh count are separate levers, and a wrong angle defeats a correct mesh count every time.
22.5 degrees or 45 degrees are the angles that work for single-color printing. Both offset the dot rows enough from the mesh's own grid that the two patterns stop reinforcing each other. 45 degrees is the more common default, partly because it's also the angle least noticeable to the eye at a normal viewing distance. 22.5 degrees is the fallback when 45 would clash with a second ink or a patterned garment.
Multi-color work adds a constraint on top: the angles between different color separations need their own spacing, or the overlapping dot matrices from two inks create a fresh interference pattern on top of whatever mesh interference you've already solved. Screens closer than 30 degrees apart produce an objectionable pattern; keeping at least 30 degrees, ideally 30-45, between colors is the standard fix.
| Job type | Screen angle(s) | Why |
|---|---|---|
| Single color | 22.5° or 45° | Off the mesh's 0/90 grid |
| Two-color (duotone) | Dark ink 45°, light ink 75° | 30° separation; darkest ink placed where it's least visible |
| Three-color (tritone) | Darkest 45°, mid 75°, lightest 15° | Maintains 30° minimum spacing across all three |
| Four-color (rotated set) | ~5°, 20°, 50°, 80° | Rotated off the classic 0/15/45/75 set, which still clashes on some mesh weaves |
What LPI Should You Even Start With?
LPI choice comes before angle and mesh count, since it sets the ceiling for both. Fine LPI -- 55 and up -- holds detail on smooth substrates and tight artwork, but it demands a correspondingly fine mesh, which means less ink deposit and more risk of the mesh becoming the limiting factor. Coarse LPI, 35 to 45, is the safer default for heavier garment blends or textured fabric, where a fine mesh would clog or under-deposit ink no matter how well the angle is dialed in.
Where the halftone actually gets built matters too. Generating the halftone in Photoshop vs a dedicated tool changes how much direct control you have over LPI, angle, and dot shape in one pass. Photoshop's Color Halftone filter locks you into round dots and limited angle control; a dedicated tool exposes all three independently. Comparing options before committing to a workflow? A free halftone generator tested against a real print job covers what the no-cost tools actually get right, and where they fall short on a press.
Dot shape belongs in this decision, not as an afterthought once LPI and angle are locked. Round or elliptical dots outperform square dots because square dots echo the mesh's own square weave and reinforce whatever interference the angle didn't already fix. Elliptical dots also transition tone more gradually, avoiding the sudden jump in value that happens when square dots merge into each other at mid-tones.
The Three Things Angle and Mesh Count Won't Fix
Get mesh count and angle right and you've solved most of the problem. These three will still bring moire back if you ignore them.
Emulsion-over-mesh (EOM) needs to sit between 8 and 12 microns. Below 8, dots form improperly and the mesh threads start eclipsing the edges of small dots. Above 12, the emulsion is thick enough to block ink transfer through capillary action -- weak, inconsistent highlight dots, a different failure mode that can look like moire but isn't caused by angle or mesh count at all.
Mesh tension has to stay even, within about 1-2 N/cm across the frame. A screen tensioned unevenly stretches the mesh grid unevenly too, which creates localized moire in specific areas of the print even when the overall mesh count and angle are both correct on paper.
Thread diameter changes how much of a dot's edge gets eclipsed by the mesh itself. Heavier HD-grade threads produce more moire than thinner S-grade threads at the identical mesh count, because a thicker thread blocks proportionally more of each small dot. Done everything else right and moire is still showing up? Check whether the mesh you're using is HD weight before you start second-guessing the file.
Get mesh count and angle right first. Then, if moire is still on press, work through these three before you assume the file itself is wrong.
Frequently Asked Questions
What mesh count do I need for a 45 LPI halftone in screen printing?
Multiply your LPI by 4.5 (1.8x if you're working in threads/cm) to get the lowest mesh count that will hold the dot. A 45 LPI halftone needs roughly a 200-mesh screen (45 x 4.5 = 202.5). Round up, not down, and remember the box count isn't the count you'll actually print with once the screen is tensioned.
What screen angle should I use for a single-color halftone print?
22.5 degrees or 45 degrees. Both keep the dot rows off the mesh's own 0/90/180/270-degree grid, which is what triggers moire in the first place. 45 degrees is the more common default; 22.5 degrees is the alternate when 45 degrees clashes with a second color or a patterned garment.
Why does my halftone still moire even after I changed the angle?
Angle only fixes one axis of the problem. If your mesh-to-LPI ratio lands on a clean whole number, or your emulsion-over-mesh is outside 8-12 microns, or your mesh elongated more than expected during tensioning, you'll still get interference at a correct angle. Angle and mesh count have to be right together, not one or the other.
Does dot shape affect moire in screen printing?
Yes. Square dots line up with the mesh's own square weave and reinforce the interference pattern. Round or elliptical dots break that alignment and transition tone gradually instead of jumping value where dots touch, which is the second-biggest lever after angle and mesh count.
How much does mesh elongate after tensioning, and does it matter for moire?
Typically 4-8 percent once pulled to standard operating tension. A 195-count mesh can open up to roughly 183 threads per inch on the frame. If you selected your mesh count based on the box number instead of the post-tension number, you can end up under-meshed for your LPI without changing anything else, and moire reappears.
Conclusion
Mesh count at four to five times your LPI, adjusted for the 4-8 percent the mesh will elongate once tensioned. Screen angle at 22.5 or 45 degrees, off the mesh's own 0/90/180/270 grid. Round or elliptical dots instead of square. Get those three right, and halftone moire in screen printing is solved before you even get to exposure or ink.
This is the garment-mesh version of the problem. Headed to a riso drum instead of a press? The interference source and the fix are different.
Test angle and LPI combinations against your own source image before you burn a screen: try the halftone effect in kott's free studio.
02/ OUT
Every setting described above is a real control. Open your own image and sweep it.
All journal entries