Which dithering mode should you use to laser engrave a photo?
A square on a test grid tells you nothing about how a face will look. Photos go through a dithering mode first, and laser engraving dithering decides whether you get a crisp portrait or a brown smudge. Pick it with a test, not a hunch.
Applies to
- Process
- Diode, CO₂ & fiber lasers engraving photos (bitmaps)
- Software
- xTool Creative Space, LightBurn image modes
- Test image
- One photo with pure white, deep shadow and fine detail
Why it matters
A photo isn’t a vector. Before the laser can burn it, the software converts it into dots. Most modes — Jarvis, Bayer and the rest — switch the laser fully on or off and build the tones out of tiny dots. Grayscale is the odd one out: it varies the laser’s power instead.
Each mode lands differently. Bayer is the most obvious, close to a halftone. The others differ in subtle ways you only see on the wood.
So hold everything else still. Same speed, same power, same lines per cm. Then the difference you see is the mode, not your settings.
What good looks like
- The test photo has a bright white area, a dark area and fine detail (hair, fabric, eyes)
- Every mode ran at the same speed, power and lines per cm
- You picked the winner by looking at the real burn, in more than one light
- The winning mode is saved with the material’s settings
How to apply it
- Pick a test photo with strong whites, deep darks and fine detail. Make or Break Shop uses Baby Yoda for exactly that reason.
- Engrave it in Jarvis or grayscale at a few speed and power combos, and pick the best one.
- Lock that speed and power. Make or Break Shop landed on 40% and 250 mm/s on 3 mm ply with a 40 W xTool S1.
- Engrave the same image once in every dithering mode your software offers.
- Compare them side by side, close up and at arm’s length, and keep the winner.
Test it on your setup
- Zoom into the software preview before you burn. You’ll see the on/off dots in the dithered modes and the varying shade in grayscale.
- Put the Jarvis and grayscale versions side by side first. Those two cover most jobs.
Exceptions & tradeoffs
The winner is material-specific. The mode that wins on birch ply may lose on slate or anodized aluminum, so re-run it for a new material.
Mind you, lines per cm changes the look as much as the mode does. Lock it while you test modes, then test it on its own.
Common mistakes
- Choosing a dither mode from someone’s screenshotBurn the modes on your own material and compare
- Nudging speed or power between mode testsLock speed, power and lines per cm; vary only the mode
- Setting photo settings off a plain squareTest with a real photo that has highlights, shadows and detail
Frequently asked questions
- What is the best dithering for laser engraving photos?There’s no single winner. Jarvis and grayscale are good first picks. Then run the same photo in every mode at one setting and choose by eye on your material.
- What is the difference between Jarvis and grayscale laser engraving?Jarvis turns the laser fully on or off and builds tones from dot patterns. Grayscale varies the laser’s power to shade each spot.
- What does Bayer dithering look like on a laser engraving?It’s the most pronounced pattern, close to a printed halftone. You can see the dot grid.
- Why does my laser photo engraving look muddy?Usually the speed and power are too hot for the material, or the mode or lines per cm doesn’t suit it. Find speed and power first, then test modes and resolution separately.
Using a real photo with whites, darks and detail to set bitmap engrave settings — How to Find the Best Laser Settings ↗ at 14:04.
Running one image through every processing method at the same speed and power (40% / 250 mm/s) — How to Find the Best Laser Settings ↗ at 15:25.
Grayscale varies power while the other modes switch on/off; Bayer is the most pronounced — How to Find the Best Laser Settings ↗ at 16:06.
The method — find a setting in Jarvis or grayscale, then run one image through every mode at that setting — is Make or Break Shop’s, and so are the 40% / 250 mm/s numbers from his S1. His ready-made image tests are a paid xTool file. You can build the same thing free by duplicating one photo and changing only the mode.