Verify the model’s physical size after export

Define exactly what each dimension measures

Obtain the approved product revision and choose meaningful dimensions with unambiguous endpoints. Overall height, body width and the distance between two distinct features provide stronger evidence than a single number. Record whether handles, feet, removable accessories or an open lid belong in each measurement. A packaging dimension is not an acceptable substitute for the product’s geometry, even if both appear in the same catalog record.

Give the asset owner a small dimension table with units, reference state and uncertainty noted. If two supplied drawings disagree, resolve that conflict before changing geometry. Do not let the export technician select whichever dimension makes the model easiest to approve. The product owner should decide the authoritative reference; the production team should show how that reference travels through the pipeline.

Follow the conversion through the delivered scene

Khronos specifies meters for glTF linear distances. Its transform model combines a node’s local transform with its parent’s transform. These are format facts, not proof that an individual exporter has converted an authoring scene correctly.

Ask the technical owner to report the authoring units, export conversion and effective dimensions of the delivered object. Include inherited scaling and the chosen scene or product state. Keep the untouched source alongside the candidate export. A component may have sensible local dimensions while a parent changes its final size. Conversely, a displayed scale value alone may be harmless if conversion was already baked correctly. Diagnose the effective result rather than imposing a cosmetic rule that every number must equal one.

Compare numerical evidence before judging presentation

Open the exact delivery file in an inspection workflow that can report its effective dimensions. Compare the agreed endpoints, not merely a bounding box that includes lights, staging geometry or an extended animation pose. If the inspection tool cannot distinguish these, have the asset owner produce a clearly scoped measurement report and retain the file revision it describes. Note which dimensions are directly checked and which remain unresolved.

Then review the receiving viewer in its documented scale states. Record the initial presentation and any explicit reset or size control relevant to the intended experience. Do not resize the asset by eye until it resembles a nearby chair or person. That creates a second undocumented adjustment and can conceal the original defect. Keep presentation observations separate from numerical acceptance, particularly when automatic framing makes differently sized files look alike.

Repair the first incorrect stage and reissue evidence

When all dimensions share one discrepancy, investigate unit conversion or inherited scale before remodeling. When only one dimension differs, inspect the reference state and geometry rather than applying a uniform correction. Route the defect to the stage that introduced it and describe the expected values. A downstream viewer adjustment should not silently compensate for a defective master that other channels will reuse.

After correction, repeat all selected measurements on the new delivered file and inspect the receiving experience again. Record the accepted asset revision, reference revision, measurement method and remaining limits. Keep written dimensions available to readers independently of the visualization. Acceptance means this asset passed the stated checks; it does not turn a consumer spatial preview into a surveying or installation tool.

Make the reasoning concrete.

A fictional desktop-organizer brand supplies a model for a unit that is 30 centimeters wide. The export’s effective width is three meters, but automatic camera framing makes its preview look normal. Height and depth show the same discrepancy.

The brand selects the conversion repair and retains the before-and-after measurements. It does not describe the preview as proof that the organizer will fit a customer’s particular shelf.

Take the questions into your next brief.

Use these prompts in your own notes. Nothing is submitted from this page.

1. Reference

   Which product revision and endpoints define each approved dimension?

2. Conversion

   Where do authoring units become delivery units, including inherited transforms?

3. Measurement

   Which exact delivery file and effective dimensions were inspected?

4. Presentation

   What scale state did the receiving viewer actually show?

5. Disposition

   Where did the mismatch begin, and what evidence closes it?

Watch for the wrong conclusion.

• Approving size because the model fills the viewport.
• Using carton dimensions for product geometry.
• Correcting one viewer while leaving the shared asset wrong.

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: glTF 2.0 Specification →Primary format specification · Supports: Section 3.4 defines linear units; section 3.5.3 defines parent/local transform composition. The reconciliation workflow and fictional example are original analysis.