UV-DTF white and varnish masks are two production separations with different jobs. White backs the colour so it remains visible on a dark, clear, or coloured object. Varnish controls clear ink above the colour. Neither plane is the adhesive. Build the masks from the intended finished stack, convert every pixel control to a physical distance, and verify layer order, polarity, mapping, mirror state, cure, and registration in the actual output queue.
Start with the finished decal, not the channel list
For a common reverse-transfer UV-DTF construction, read the functional stack from the object towards the viewer:
- Object surface: the cleaned, compatible substrate.
- Pressure-sensitive adhesive: supplied by the transfer-film system.
- White ink: an opaque backing under colour where required.
- Colour ink: the visible image.
- Clear varnish: the outer clear-ink region or selective finish.
Film A and film B move and protect that stack during printing, lamination, peeling, and application. They are not W or V planes in the artwork file.
Functional order is not automatically printer deposition order. A reverse-print queue may deposit varnish, then colour, then white against a carrier so that transfer presents the finished order above. Another device or film construction may require a different sequence. Record the exact printer, ink set, film pair, queue, mirror state, and documented layer sequence; do not infer them from filenames such as W1 and V1.
NestSheet’s dedicated UV-DTF mode remains roadmap. The current UV-DTF workflow page describes that boundary: supported TIFF and PSD/PSB export rows can carry white and varnish planes, but a file with those planes is not an end-to-end device recipe.
Give each plane one unambiguous job
| Plane | Source intent | Geometry rule | Typical file-side failure |
|---|---|---|---|
| Colour | Visible RGB/CMYK artwork | The approved visible edge | Wrong composite, conversion, scale, mirror, or crop |
| White | Intended opaque backing | Usually inside or equal to the colour footprint | Missing backing, an even white rim, weak edge support, or unintended white under faint alpha |
| Varnish | Intended clear-ink coverage | Full, selective, or intensity mask defined independently of white | Missing highlights, a clear rim, unwanted flood, or flattened grayscale levels |
| Adhesive | A property of the film/application system, not an image plane | Determined by film construction and transfer | Edge lift or release failure that mask editing alone cannot solve |
Do not clone one grayscale plane into the other merely because their silhouettes look similar. White answers “where does colour need opaque backing?” Varnish answers “where should clear ink exist, and at what encoded level?” A transparent hole may need neither; a white-only mark needs a real white source even though no coloured pixel can generate it; a selective gloss detail may need varnish beyond neither.
The spot-channel guide owns container structure, channel naming, polarity, and import semantics. Here the concern is how W and V geometry work together in the physical stack.
Generate white from intended opacity
When white is alpha-derived, the production mask begins with source opacity, not colour brightness. That distinction prevents a pale yellow pixel from being mistaken for “less white” merely because it looks light.
NestSheet’s current generated-white path exposes concrete controls:
- Alpha threshold: 0–255; the current default is 15. It removes opacity below the chosen inclusion boundary before the base mask is formed.
- Choke: 0–10 px; the current default is 2 px. Choke erodes the white geometry inward.
- White amount: 0–255. It scales mask strength across the generated plane; it does not mechanically realign colour and white.
- Mode/halftone controls: these change distribution inside eligible geometry. They do not make a contaminated edge clean or turn transparent pixels into a named white-only separation.
Use a file-supplied white source when the job contains intentional white-only information. If the source carries only RGBA artwork, a fully transparent pixel provides no visible footprint from which to derive that mark.
Inspect alpha first, then the generated white plane, then an ART/WHITE overlay. An even perimeter reveal follows geometry and may justify a smaller mask. A reveal displaced mainly in one direction points to colour-to-white registration, transport, scale, or origin; increasing choke may hide the symptom while leaving the mechanical error untouched. The canonical white-halo diagnosis covers that branch in depth.
Treat varnish as a finish mask, not glue
A varnish plane can represent a flat clear region or a grayscale intensity map. The receiving RIP and output mode decide whether multiple encoded levels survive as distinct output; a gradient-looking file is not proof of physical relief.
In NestSheet’s current varnish-bearing export row:
- gradient is the default mapping, with source intensity carried into V;
- binary mode uses a 0–255 threshold, currently defaulting to 10;
- the shipped exporter uses the selected, applied, or canonical artwork and its alpha, or an AI-fused derivative; dedicated varnish data detected during upload is not currently wired into this export path;
- spread dilates the mask outward by 0–20 px, with a current default of 1 px; and
- combined supported rows can emit up to four W and four V planes.
Binary threshold does not affect gradient mode. Spread changes mask geometry; it does not change adhesive chemistry, ink cure, surface energy, or film compatibility. Upload normalization can detect and extract dedicated varnish data, but the shipped export path does not consume it today.
Convert pixel controls to a physical allowance
A pixel value is incomplete until the output raster density is known:
millimetres per pixel = 25.4 ÷ output PPI
physical choke or spread = pixels × 25.4 ÷ output PPI| Output raster density | 1 px | 2 px |
|---|---|---|
| 300 PPI | 0.0847 mm | 0.1693 mm |
| 600 PPI | 0.0423 mm | 0.0847 mm |
| 1200 PPI | 0.0212 mm | 0.0423 mm |
A 2 px choke at 300 PPI is four times the physical distance of 2 px at 1200 PPI. Record both pixels and millimetres in the job receipt. Choose the smallest measured allowance that survives the shop’s controlled registration test; NestSheet defaults are starting file values, not universal process limits.
Make the RIP prove the handoff
The current export matrix can write supported TIFF or PSD/PSB structures with separate W and V planes. That is a file claim. The RIP must still prove that it discovered each plane and assigned it to the intended ink role.
| RIP-side check | Pass condition | Why it matters |
|---|---|---|
| Plane inventory | The expected colour, W, and V planes appear once each | Missing or duplicated planes change the stack |
| Names and mapping | Every imported name maps to the intended device ink role | A label is not an ink assignment |
| Polarity | Solid test regions show coverage; declared empty regions show none | Current serialized W and V use opposite byte conventions |
| Dimensions and origin | Colour, W, and V report the same canvas, scale, crop, and origin | Correct masks can still be offset |
| Mirror and sequence | Preview orientation and deposition order match the validated transfer construction | Reverse transfer can invert intuitive visual order |
| Screening and ink limits | The recorded queue settings match the approved test | Extra passes or density changes cannot be inferred from channel count |
| Cure/exposure | Every layer uses the tested device-specific condition | A correct mask cannot compensate for under- or over-cure |
Multiple W or V channels are duplicated output planes for a validated mapping, not automatic extra opacity or texture. Likewise, W and V names are editable defaults rather than universal queue commands. Use the RIP handoff hub to record the exact environment, then compare the imported preview with known expected regions.
The queued PSD/PSB-versus-TIFF guide owns the container decision. Once a validated container is chosen, this layer check remains the same: inspect roles, coverage, alignment, sequence, and physical output.
Diagnose failures by their shape and interface
| Symptom | First branch to inspect | What not to change first |
|---|---|---|
| White rim follows the entire colour perimeter | Alpha edge, white threshold, physical choke, then bidirectional registration | Varnish density |
| White reveal is mostly on one side | Colour-to-white registration, transport direction, shared origin, scale, mirror | A global alpha threshold |
| Clear rim extends past the intended finish | V source, spread in physical units, V registration | White amount |
| Selective varnish is missing or floods the decal | Source mask, gradient/binary choice, name, polarity, RIP mapping | Adhesive or surface preparation |
| Colour loses opacity on a dark or clear object | White source, W coverage, polarity, mapping, ink limit, cure | Varnish spread |
| Decal lifts cleanly from the object | Substrate cleanliness, surface compatibility, adhesive film, application pressure, dwell | W/V mask geometry |
| Ink stack separates internally | Ink/film compatibility, layer sequence, inter-layer cure or exposure | Alpha cleanup alone |
| Surface stays tacky, clouds, cracks, or abrades | V load, cure/exposure, film handling, bend and material compatibility | Channel name alone |
This ordering prevents a common category error: correcting a physical interface with an image operation. A mask edit is justified when the defect is already visible in that plane or follows its geometry. If source, exported planes, and RIP preview agree but the printed layers move or separate, continue in the device, material, and application branch.
Use a controlled three-mask proof
Before a production sheet, make a compact diagnostic file with:
- a solid colour shape with matching white and no varnish;
- the same shape with full varnish;
- a selective-varnish region with a hard boundary and, if supported by the validated queue, a declared grayscale ramp;
- an intentional transparent hole that must remain empty in W and V;
- a thin colour edge for measuring white choke;
- a varnish edge for measuring spread; and
- registration marks that let colour-to-white and colour-to-varnish displacement be measured in both axes.
Export the actual production row. Inspect colour, W, and V separately at 1:1; record output PPI, threshold, choke, amount, varnish mode, spread, plane names, polarity, dimensions, mirror state, and checksum. Then import that exact file into the exact RIP/version/queue and print the smallest physically useful sample.
Approve each interface separately: object-to-adhesive, adhesive-to-ink stack, white-to-colour registration, and colour-to-varnish registration. The controlled DTF test-file protocol supplies the broader go/no-go record; this UV-DTF proof adds a distinct V mask and transfer-stack inspection.
Sources
- NestSheet, UV-DTF workflow — current capability and roadmap boundary: https://nestsheet.com/uv-dtf
- NestSheet, Export matrix — supported colour, white, and varnish rows: https://nestsheet.com/export
- NestSheet, DTF spot channels explained — names, polarity, and RIP handoff: https://nestsheet.com/blog/dtf-spot-channels-explained
NestSheet defaults, channel counts, polarity, export-source selection, and mask operations were verified against the shipped repository on 24 July 2026. Physical stack order and process settings must be confirmed against the exact printer, ink set, transfer-film construction, and output queue used for the test.
