MOPA Color Picker
Pick any colour → ① verified swatches burned on Hank's ComMarker B4, ② a predicted B4 setting from a model fitted to them (with confidence), and ③ reference settings from ~1,900 cells of another machine's photographed grids.
Read first: Colors measured from photos of another user's test grids (credit: Povilas, nothing.lt gallery); results vary by machine, lens, focus and material — run a small test first. Hank's machine: ComMarker B4 60W MOPA.
Source machine: JPT M7 60W MOPA with an F290 200×200 mm lens. Hatches are bi-directional; when passes > 1 the hatch angle increments 90° each pass (crosshatch). Starting hatch angle and overscan are not shown in the photos.
On the B4: its field lens is smaller (~126×126 mm field, likely ~F160), so the spot is smaller and energy density is higher at the same power and speed. Treat these as starting points — expect to scale them (usually less power and/or more speed) and dial in with a small grid around the suggested values.
Verified colours by material
Each colour belongs to exactly one material — never mix. Verified first, vivid before black/grey. Plates · API /api/materials
1. Choose a color
Or tap any sampled color below. Downloads: color_settings.json · color_settings.csv · color_chart.png (phone chart) · b4_measured.json (B4 measured)
2. Settings
Swatches actually burned on Hank's ComMarker B4 and measured — angled phone photos (primary, most saturated angle) + flatbed scan (secondary; a flatbed's direct light under-reports interference colours like blue/purple). Hank's by-eye checks rank highest. Ranked first everywhere.
Loading…
A continuous model (settings → colour) fitted to every B4 measurement, with the JPT M7 photos as a weak prior, is inverted for your colour — including in-between colours nobody has burned yet. Confidence is honest: far from B4 data = needs testing.
Loading model…
Reference only: photographed test grids from a JPT M7 60W + F290 lens (different machine). Starting points — scale for the B4.
Loading data…
All sampled colors (grouped by family, light → dark)
B4 test-grid rounds (Hank's ComMarker B4 — iterative colour optimizer)
Loading…
Each round is a test card on an 85×55 stainless business card (82×52 mm test area). After a burn + flatbed scan, the measured colours land in the B4 dataset above and the next card's settings step toward each target colour.
How this was made
Each cell's color is the median RGB of the cell centre in the original photo. On plain plates the color is white-balanced against the bare steel next to the grid (gain clamped 0.8–1.25; raw value kept as hex_raw); mirror-finish dog tags have no neutral reference and are uncorrected (confidence med). Settings are read from the labels engraved on each grid — nothing was invented; unreadable values are null. Matching uses CIE76 ΔE in Lab (sRGB, D65): ΔE < 3 is close to indistinguishable, > 10 is a clearly different color. Photographed laser colors are angle- and light-dependent, so they look duller than in person.
Calibrating your own machine? Use the MOPA Colour Calibration Tool (download a LightBurn grid kit, upload a scan).