How to Laser Engrave Photos on Metal: LightBurn Dithering & Z12 60W MOPA Guide

Can a fiber laser reproduce the shadows, skin tones, highlights, and small details of a photograph on metal? Often, yes—but a good image on your monitor does not automatically become a readable metal engraving. Surface finish, focus, dot spacing, the image-processing algorithm, and the laser’s interaction with the metal all affect the result.

This guide builds a repeatable fiber laser photo engraving on metal workflow around the Carverall Z12 fiber laser. We focus on the Z12 60W MOPA, while explaining when a 20W/30W standard fiber laser or the 50W 3D relief model may be more appropriate. You will learn to prepare a photograph, compare LightBurn image modes, test line interval and laser settings systematically, and recognize a poor result before wasting customer material.

Evidence and settings note: Carverall’s published Z12 lineup confirms 20W/30W standard fiber, 50W 3D, and 60W MOPA variants and LightBurn-compatible workflows. This article does not present newly measured Carverall photo-engraving recipes, tested speeds, pulse widths, darkness measurements, or cycle times. Every proposed test below is a recommended experimental starting workflow, not a Carverall-validated parameter preset. You must validate actual settings on your own machine, lens, metal grade, and surface finish.

What does “photo engraving on metal” really mean?

A portrait contains continuous tones. A metal-marking laser produces a physical surface response: a dark mark, a light mark, a change in texture, removal of a coating, or a shallow engraving. To suggest intermediate shades, software can either distribute tiny on/off marks (dithering) or, where hardware and materials permit, vary delivered energy across an image (grayscale processing). The two approaches are not interchangeable.

On uncoated stainless steel, successful photo reproduction depends on obtaining sufficient contrast and keeping dots from merging. On anodized aluminum, a lighter revealed substrate may require inverted image logic compared with dark marks on bare metal. On polished metal, reflections can make an otherwise sharp engraving difficult to photograph or read. A clear result is defined by its intended viewing conditions, not only by a zoomed-in software preview.

For the underlying Image Mode behavior, see LightBurn’s official Image Mode documentation.

Choose the appropriate Carverall Z12 variant

Z12 configuration	Photo-engraving role	Practical limitation to test
20W / 30W standard fiber	Flat logos, text, and contrast-based image marks on compatible metal surfaces.	Usable tone range and marking time depend on material, source, optics, and focus; do not assume MOPA-specific Q-pulse width control.
50W 3D relief	Projects combining a photo-like image with separate depth or relief work, when the design calls for sculpted detail.	A relief map is not the same thing as a photographic grayscale image. Depth strategies must be tested separately.
60W MOPA	Best fit in this guide for studying contrast and texture with additional frequency and pulse-width flexibility.	Extra control expands the test space; it does not guarantee more photo detail or a preset that transfers between alloys.

For a model-by-model overview, read 60W MOPA vs 50W 3D vs Standard Fiber. If you are pursuing actual sculpted depth instead of a photograph, use the separate deep metal engraving and relief guide.

Before engraving: pick a measurable test surface and work safely

1. Record the metal. Identify its grade, thickness, polished/brushed state, and any coating or lacquer. Two pieces that look identical may not react identically.
2. Use sacrificial coupons. Label each coupon with its source lot and a unique ID. Reserve customer parts until the process passes your acceptance checks.
3. Control surface preparation. Remove approved surface contaminants using a material-compatible cleaning procedure. Keep the method consistent across test pieces.
4. Secure and focus the part. Confirm the lens, usable field, machine calibration, flatness, and correct focal position for your actual Z12 configuration. A polished plate that rocks under the scanner is not a controlled test.
5. Apply laser safety controls. Use the required enclosure, wavelength-appropriate protection, trained operating procedures, and effective local exhaust. Do not process unknown coatings or alloys until their emissions have been evaluated. Refer to OSHA’s laser-hazard guidance.

A suitable first test is a small, reasonably flat sample of known, uncoated metal. Choosing an easy-to-label coupon matters more than picking an impressive final object on the first attempt.

Step 1: prepare the photograph before importing it into LightBurn

Start with one original, legally usable image and create a fixed evaluation crop containing skin or midtone gradients, dark details, bright highlights, and a hard edge. Keep the same crop for every software-mode comparison; changing the photo between runs makes the comparison meaningless.

1. Crop to the final physical aspect ratio. Avoid stretching a portrait just to fill a tag or plate.
2. Remove unnecessary background distractions, but preserve hair strands, facial details, and midtone transitions used to judge quality.
3. Inspect the image at its intended engraved size. Detail invisible at normal viewing distance may only add noise to the laser file.
4. Prepare a controlled grayscale copy and retain the original. Adjust exposure and contrast cautiously; heavy sharpening can turn hair or skin texture into harsh speckles.
5. Save the input image with a version name, such as portrait_A_crop_v01, so you can reproduce the exact test.

AI-assisted production tip: AI tools can help catalog image versions, suggest non-destructive crops, flag poorly exposed backgrounds, or produce a test checklist. They should not invent facial details or silently alter the customer’s identity in an engraving proof. Keep the original source and require human approval of retouching.

Step 2: compare LightBurn Image Modes, not just laser power

Import the same crop into LightBurn and select an image layer. On compatible setups, compare these image-processing approaches while holding the physical test size and laser baseline constant:

• Jarvis or Stucki: useful candidates for photographic gradients because they distribute on/off dots to simulate tonal variation. Compare shadow texture and facial midtones rather than selecting one algorithm by reputation alone.
• Dither: another error-diffusion option to compare on the same coupon stock.
• Ordered: useful for a visibly regular dot texture or blocks of flatter tone; it may produce an intentional pattern.
• Threshold: suitable for art that is already truly two-color, not usually a good default for a full-tone portrait.
• Grayscale: varies output power where supported, so its physical behavior can differ substantially from on/off dithering. Do not assume that a grayscale preview on screen predicts accurate metal tone or 3D depth.

LightBurn’s Image Mode reference explains these modes and their different strengths. Confirm that your LightBurn edition, controller, and current device configuration expose and support the selected options; galvo-specific settings can differ from those for diode or CO₂ machines.

Suggested experiment: arrange the same small crop in three or four labeled cells (for example, Jarvis, Stucki, Dither, and Ordered). This is an experimental comparison design, not a preset with verified energy values. If the material behaves differently across the bed, move the test to a more uniform central region and repeat.

Step 3: establish a safe baseline before the photo test

Before committing a photo, use a simple target with text, fine lines, and solid patches. The goal is to establish a mark that is legible and does not produce unwanted edge buildup or excessive thermal discoloration. Keep the coupon, focus, and cleaning method fixed. On compatible LightBurn setups, use Laser Tools → Material Test to generate a controlled grid and record the combination you actually select.

Do not copy generic speed, power, frequency, or pulse-width values from another machine. Source capabilities, lens selection, duty cycle, field correction, workpiece alloy, and desired appearance change the outcome. Start within the certified operating range for your exact laser source and machine, and follow Carverall’s configuration-specific guidance when supplied.

Once you have a stable contrast baseline, freeze those laser variables for the first Image Mode comparison. Otherwise you will not know whether a better photograph came from the dithering algorithm or a different energy dose.

Step 4: test line interval and image detail systematically

Line interval is the distance between adjacent scan lines. LightBurn also displays a related image density in DPI, using DPI = 25.4 / line interval (mm). This is a unit conversion, not a recommendation for any particular machine setting.

A smaller interval means denser rows, which may resolve fine features but also increase overlap, heat accumulation, and runtime. If marks blend into a dark patch, simply increasing DPI can make the photo worse. If you see visible unmarked gaps or bands, the interval may be too large for that mark width and surface response.

Choose an operator-approved reference interval, then test one less-dense and one more-dense candidate within safe limits. Keep the image file, dithering mode, energy baseline, and sample preparation unchanged. Evaluate each from the same viewing distance and lighting angle. Record both the detailed appearance and the actual job time.

For further context on scan interval and hatch geometry, see the LightBurn Image Mode settings and our Z12 MOPA Material Test workflow.

Step 5: add MOPA frequency and pulse-width testing only after the image workflow is stable

On the Z12 60W MOPA, frequency and Q-pulse width provide additional ways to alter the interaction between the pulsed laser and the surface. Their useful ranges are source- and controller-specific. They are not two independent “contrast sliders” that can be maximized for the darkest possible photo.

1. Keep the approved photo crop and an already selected image algorithm fixed.
2. Select a limited, source-supported frequency range and hold pulse width constant. Run a labeled grid that changes frequency only.
3. Compare actual texture, shadow separation, edge crispness, and visible heat effects after identical cleaning.
4. Take only the promising and safe frequency candidates into a separate pulse-width grid, holding all other variables fixed.
5. Repeat the candidate on a new coupon to check whether the result is consistent rather than an artifact of local surface finish.

For a deeper explanation of the controls, read our 60W MOPA pulse width and frequency guide. No exact frequency/pulse-width recipe is claimed here as a Carverall-verified photo-engraving setting.

How to judge a successful photo versus a failed one

The following is a diagnostic comparison framework, not a report of a Carverall test result.

Evaluation criterion	Acceptable candidate	Reject or troubleshoot
Face and highlight detail	Eyes, hair edge, and highlight shapes remain distinguishable at final size.	Midtones collapse into one dark area or highlights disappear.
Dot and hatch texture	Pattern appears intentional and does not distract at normal viewing distance.	Dots merge, scan bands show, or the image has patchy density.
Surface and finish	Appearance remains acceptable after the defined cleaning process.	Loose residue gives false contrast; unexpected discoloration or roughness appears.
Repeatability	A second coupon from the same lot meets the same visual standard.	The apparent winner cannot be reproduced under identical documented conditions.
Production suitability	Actual time, handling steps, and rejection risk fit the planned product.	The image looks good only in a macro photograph or requires impractical rework.

Use a fixed inspection setup: same camera, distance, light direction, exposure, and cleaning stage. A photo engraved on a reflective plate can look dramatically different when the lighting moves. Compare an engraved sample beside the original cropped source image, but avoid judging engraved metal with screen-brightness expectations.

Troubleshooting: why does a fiber laser photo look muddy or striped?

Symptom	Likely test hypothesis	Next controlled action
Photo is too dark and details vanish	Dense dots, excessive surface reaction, or an overly aggressive source image curve.	Hold the image constant and test less dense intervals or a milder validated energy baseline, one factor at a time.
Portrait is pale or hard to recognize	Insufficient material contrast, unsuitable surface finish, or inappropriate negative/positive mapping.	Try a test coupon with a different prepared finish; inspect the negative-image logic before increasing energy.
Horizontal bands or visible scan lines	Interval mismatch, focus/flatness error, or galvo-specific timing/calibration issues.	Verify focus and lens calibration; run a simple uniform fill target before blaming the image file.
Speckled skin and noisy shadows	Image oversharpening or a dithering algorithm that accentuates source noise.	Compare the identical unsharpened crop using Jarvis, Stucki, and Dither under fixed laser settings.
Edges appear wider than the source	Overlapping spots, a poor interval choice, or an unsuitable marking baseline.	Test interval and dot/line width compensation where supported; evaluate a high-contrast edge target.
One coupon works, the next fails	Different alloy, coating, cleanliness, or focus height.	Document lot, finish, cleaning, focus, and placement; retest with matched coupons.

If the issue is specifically color or black marking on stainless steel—not tonal reproduction of a photograph—use our Z12 stainless steel black and color marking guide to avoid mixing separate goals.

Use AI to reduce wasted tests, not to fabricate results

Small engraving businesses often spend more time comparing samples and finding old files than running the laser. A simple structured log can help. For each cell, record:

• Job: coupon ID, source photo version, final dimensions, substrate alloy and finish.
• Machine: Z12 variant, laser source/controller, field lens, software version, and focus method.
• Parameters: image mode, line interval/DPI, scan angle, speed, power, frequency, Q-pulse width where available, and passes.
• Results: inspection photograph, actual cycle time, cleaning method, reject reason, and the operator’s visual score.

AI can summarize observations across the log, cluster common failure reasons, and suggest a next test hypothesis such as “compare lower dot density on the same brushed finish.” It cannot validate a laser setting by itself or replace an operator who inspects the metal. Never let a model fill missing machine settings with plausible-looking numbers; mark unrecorded fields as unknown.

For a reproducible production trial, keep one approved reference coupon and recheck the process after changes in material batch, lens, software profile, or focus setup. This is how photo engraving becomes a business workflow rather than a collection of one-off lucky samples.

Questions operators ask before taking a photo-engraving order

Can a normal fiber laser engrave photos on metal?

It may produce recognizable tonal images on compatible surfaces if the available contrast and dot size support the design. A 20W/30W standard fiber is worth testing for flat photo-like markings, but it does not provide the same pulse-width flexibility as the Z12 60W MOPA.

Is Jarvis always better than Stucki for stainless steel?

No. They are alternative image-processing methods, not machine-specific guarantees. Compare the same image crop on the same metal under controlled conditions and judge the finished material at actual viewing size.

What DPI should I use for a fiber laser photograph?

There is no universally correct DPI. Use the line-interval/spot-response test described above. Excessively dense lines can merge or add heat without preserving new visual information. The lens, laser source, material, and surface finish all matter.

Does grayscale image mode create a 3D relief?

Not automatically. Grayscale image processing maps image tones to available power behavior; a controlled 3D relief process uses different goals and often different preparation, passes, and depth verification. Do not treat the 50W 3D Z12 relief capability as proof that every imported grayscale photograph will become a reliable sculpted relief.

Can the same recipe work on stainless steel and anodized aluminum?

Do not assume so. Bare metal and a treated surface can produce different mark polarity, color, residue, and thermal response. For coated or anodized aluminum details, use the Z12 aluminum engraving guide and qualify the actual finish separately.

From test coupon to a sellable metal portrait

Begin with a repeatable output specification: final size, accepted contrast, visible features, finish after cleaning, and maximum practical job time. Compare Image Modes first, validate line interval second, and only then use the Z12 60W MOPA’s extra parameters to explore texture and contrast. Record what the metal actually did—never only what the LightBurn preview suggested.

When you are ready to move beyond a test coupon, explore the Carverall Z12 lineup and select the standard fiber, 50W 3D, or 60W MOPA configuration that matches your work. For more application-specific testing, browse the Z12 Tutorials library.