Spatial product production workshop / 6 MIN READ
Accept the exported material channels, not the authoring preview
The attractive shader in the design application is not necessarily the material that reaches the reader.
AI-assisted research and analysis for the SNS Creative Fieldbook
Sources checked 2026-11-07
Material acceptance starts with meaning: which finish is being represented, which exported inputs produce it, and whether the receiving renderer interprets those inputs as intended. A beautiful source render cannot substitute for inspecting the actual delivery material.
What you will leave with: A material-to-channel map and controlled comparison that separate export defects from presentation differences.
Use this when
- A procedural product material is being converted for real-time delivery.
- A finish looks unexpectedly shiny, flat or patterned after export.
This is a material handoff review, not physical material certification, photographic ICC correction or a promise of identical appearance on every screen. Transparency coverage and renderer-extension eligibility need their own checks. Sources checked on 7 November 2026.
Define the finish and its evidence
List the product regions that matter to recognition: coated body, exposed metal, fabric insert, printed mark or molded grip. Attach approved references and describe the important difference in ordinary language. A matte coating should not become a polished metal surface merely because the latter looks more premium. State whether a reference depicts the current product revision, a concept or an intentionally stylized illustration.
Ask the artist to identify each region’s source material and intended exported material. Consolidating materials can be reasonable, but record which distinctions must survive. Keep a short list of features that cannot be traded away during conversion, such as a contrasting insert or the direction of a recognizable grain. This becomes the comparison brief, not a request to reproduce a studio photograph pixel for pixel.
Map authored effects to exported inputs
Khronos’s asset guide describes baking layered authoring nodes into bitmap textures for deployment. The glTF specification assigns roughness to the green channel and metalness to the blue channel of its metallic-roughness texture, with linear encoding. These narrow conventions explain why a valid image file can still drive the wrong material property.
Have the technical owner produce a compact mapping: source effect, export texture, channel or factor, material assignment and intended receiver. Record unsupported effects explicitly instead of assuming every node transfers. Inspect the files that were actually packaged, not only their filenames. A label containing “roughness” does not establish which channel holds the data, and a correct bitmap assigned to the wrong product region is still a failed handoff.
Use comparisons that isolate the suspected change
Freeze a delivery candidate and review it under repeatable camera and lighting conditions. Capture the same product regions in the authoring reference and receiving renderer, while noting differences in the presentation setup. First ask whether the material assignment and feature locations are correct. Then investigate whether the finish response has changed. Changing lighting, texture and material factors simultaneously makes it difficult to learn which change helped.
For a suspected channel problem, ask the owner for a controlled diagnostic copy that isolates the input in question. Keep that copy out of delivery. Compare a patterned area and an unpatterned area, a broad surface and a small detail, rather than choosing only the most flattering view. Record observations such as “the coated panel responds like the exposed trim” instead of untestable directions such as “make it feel more real.”
Approve the interpretation and its limits
Classify findings by consequence. A missing identifying print or changed finish category deserves a different decision from a small highlight difference between rendering environments. When a source effect cannot transfer, choose an evidenced approximation, an alternative delivery method or a stated limitation with the responsible product owner. Do not invent an extra physical capability to explain an export artifact.
Retest the repaired material in the full product, since a diagnostic success may not reveal an incorrect assignment elsewhere. Store the accepted mapping, delivery revision, comparison views and unresolved receiver differences together. Later optimizations should return to these protected features. Keep material acceptance distinct from a blanket endorsement of every device: the record should say which renderer and configuration were actually reviewed.
Make the reasoning concrete.
A fictional travel-mug team exports a coated body and metal rim. The browser model makes both regions look mirror-like. The artist proposes darkening the body color, while the technical reviewer suspects the packed material inputs.
Compensate with a darker color
Opening: Change the body color until one screenshot resembles the reference.
Hypothesis: The selected hero angle could look less reflective.
Risk: The underlying material interpretation remains wrong and other views can still imply the wrong finish.
Trace the material mapping
Opening: Inspect the exported channels and region assignments before changing the approved finish.
Hypothesis: A controlled correction can restore the intended distinction between coating and rim.
Risk: Lighting differences still require a separate comparison after the mapping is repaired.
The team chooses the mapping investigation and reviews both regions after repair. Its acceptance record explains the tested representation without claiming that the screen measures the physical coating.
Take the questions into your next brief.
Use these prompts in your own notes. Nothing is submitted from this page.
Regions
Which product surfaces have meaningfully different finishes?
Mapping
What exported input and assignment represent each source effect?
Control
Which camera and lighting conditions make comparisons repeatable?
Diagnosis
What isolated test distinguishes an input error from presentation variation?
Acceptance
Which finish distinctions passed, in which receiver, with what limits?
Watch for the wrong conclusion.
- Approving only the authoring application’s render.
- Trusting texture names without checking assignments.
- Changing several inputs at once to chase a screenshot.
Your next action
Choose one important finish and trace its source effect through the packaged texture, material assignment and receiving render.
Sources & scope
Workflows, worksheets and fictional examples are original editorial analysis. Sources support only specifically attributed principles. Platform guidance can change; legal requirements depend on jurisdiction and context. No measured results or named human reviewer are claimed.
- Khronos: Real-time Asset Creation Guidelines, version 2.0.0 →Primary asset-production guidance · Supports: The Materials Summary describes baking layered authoring effects for deployment.
- Khronos: glTF 2.0 Specification →Primary format specification · Supports: Section 3.9.2 defines metallic-roughness texture channel assignments and encoding. The acceptance workflow is original analysis.