NVIDIA NCP-OUSD - OpenUSD Development

A reference composition arc brings scene description from another asset into the prim where the reference is authored, then combines that referenced data with local opinions on the destination prim. NVIDIA’s Learn OpenUSD references guide states that when a prim is composed through a reference arc, USD first composes the layer stack of the referenced prim, adds the resulting prim spec to the destination prim, and then applies overrides or additional composition arcs from the destination prim.

Option D is correct because /Root/Chair receives the composed contents of chair.usda, while the locally authored xformOp:scale = (1.5, 1.5, 1.5) remains part of the destination prim’s stronger local opinions. If the referenced chair asset also authored a corresponding scale opinion on the same property, the local opinion would win by standard USD strength ordering, where stronger opinions override weaker ones non-destructively.

Option A is incorrect because references are not bidirectional synchronization links; editing the referencing layer does not automatically modify chair.usda. Option B is too narrow because references compose all targeted scene description, not only variant sets. Option C is incorrect because a reference does not discard local opinions. This aligns with Composition → References, Local Opinions, Layer Strength, and Non-Destructive Overrides .

Namespace-prefixed attributes are the correct approach for representing grouped or compound source data in OpenUSD. NVIDIA’s Learn OpenUSD data extraction guidance states that when source formats contain compound or grouped data types, such as structs, records, or grouped fields, namespace-prefixed attributes provide the convention for grouping related properties together in USD. It also states that USD does not have a native struct type, so namespace-prefixed attributes serve as the standard approach for representing this kind of grouped data.

Option B is correct because attributes such as acme:sensor:temperature, acme:sensor:humidity, and acme:sensor:pressure can express ownership and logical grouping while remaining ordinary USD attributes with valid scalar or array value types. NVIDIA’s custom properties lesson reinforces this pattern, explaining that namespace prefixes logically group related properties on a single prim and are especially useful when mapping structured data types from other formats into USD.

Option A is incomplete because arrays represent repeated values, not named compound fields. Option C is incorrect because USD does not provide a native struct attribute type. This aligns with Data Exchange → Data Extraction, Conceptual Data Mapping, Custom Properties, Namespace-Prefixed Attributes .

Option A is the correct USDA pattern because a payload is authored as prim metadata using the singular payload keyword with a list-editing operation such as prepend. NVIDIA’s Learn OpenUSD payload exercise shows the canonical USDA form: prepend payload = @./red_cube.usd@, authored on the prim declaration. It further explains that payloads are similar to references in USDA, with the important distinction that the keyword is payload rather than references.

The key technical purpose of a payload is deferred loading. NVIDIA explains that payloads can compose scene description when loaded, or unload the targeted scene description beneath the payloaded prim. It also contrasts this with references, which are always composed and present on the stage, while payloads can be opened unloaded and selectively loaded later.

Option B is incorrect because payloads is not the USDA keyword and it is not authored as a property inside the prim body. Option C uses a reference, not a payload, so it does not provide payload load/unload behavior. Option D is not the canonical list-op form expected here. This aligns with Composition → References and Payloads → Working With Payloads .

A well-designed model kind hierarchy is important because it gives tools a reliable, high-level “table of contents” for a complex stage. NVIDIA’s Learn OpenUSD guide explains that model hierarchy is designed to prune traversal at a relatively shallow point in the scenegraph. Without model hierarchy, traversal may need to pass through all prims in a scene, which can be costly. The key pruning point is the component model boundary: ancestors of components participate in the model hierarchy, while descendants are outside the model hierarchy.

Option A is correct because pipeline tools can use kind metadata to find meaningful asset boundaries, traverse assemblies and groups efficiently, and avoid descending into every geometric or implementation-level prim. Option B is inaccurate because kind metadata is not primarily a composition-engine optimization mechanism; USD composition resolves scene description based on composition arcs and strength ordering. Option C is also not the primary purpose. Avoiding duplication is addressed more directly by references, payloads, instancing, inherits, specializes, and asset-structure reuse. This aligns with Content Aggregation → Asset Structure → Model Hierarchy → Model Kinds, Components, Assemblies, Groups, and Efficient Traversal .

An inherit arc lets prims receive scene description from a source prim, commonly a class prim, and enables modifications to that inherited source to broadcast to all inheriting prims in the relevant composition context. NVIDIA’s Learn OpenUSD inherits lesson states that after an inherit arc is established, the source prim and its descendants can be modified in a stronger layer, including across references, and those modifications are broadcast and applied to all prims that inherit it. ( docs.nvidia.com )

Option B is correct because it captures the defining production use: central refinement of many related prims without editing every instance individually and without modifying the original referenced asset globally. NVIDIA’s strength-ordering lesson also describes inherits as the arc that allows opinions on one source prim to affect all prims that author an inherit arc to that source. ( docs.nvidia.com )

Option A is incorrect because inherits are not simply a replacement for references; they solve broadcast refinement and reusable opinion-sharing problems. Option C is conceptually adjacent but imprecise: inherits may resemble inheritance patterns, but USD inherit arcs are composition mechanisms, not an object-oriented programming system. This aligns with Composition → Inherits, Class Prims, Broadcast Refinement, Encapsulation, and LIVERPS Strength Ordering .

The most suitable USD data type for representing texture files is an asset path . NVIDIA’s Learn OpenUSD glossary defines an asset as a named reusable resource and explicitly includes textures among asset examples. It also explains that USD provides a specialized string type called asset so that attributes and metadata referring to external resources can be identified robustly. In USDA syntax, asset-valued strings are delimited with @, such as @textures/albedo.png@.

Option C is correct because texture files are external resources that should participate in USD’s asset-resolution system. Using an asset path allows the resolver to map the authored identifier to an actual file location, versioned asset, package member, or site-specific storage location. Option A is incorrect because tokens are compact identifiers best suited to enumerated values such as purpose, interpolation, or role-like names. Option B is less appropriate because ordinary strings do not communicate that the value is an external asset dependency requiring resolution. This distinction is essential for interchange, packaging, relocation, and pipeline portability. This aligns with Data Exchange → Asset Paths, Asset Resolution, Texture Dependencies, Sdf Value Types, and External Resource References .

The valid model kind hierarchies are A, B, and E . NVIDIA’s Learn OpenUSD material defines group, assembly, and component as model kinds, while subcomponent is explicitly outside the model hierarchy. The hierarchy rule is that all ancestors of a component must be group or a subkind of group, such as assembly, and a component acts as a leaf or pruning point in the model hierarchy; therefore, it cannot contain descendant model kinds such as another component, group, or assembly. ( NVIDIA Docs )

A is valid because CarA is a component and every descendant has unset kind, so no nested model violates the component boundary. B is valid because CarA is an assembly, organizational containers use group, and the leaf model assets are component prims. E is valid because CarA is a component, and internal important parts are marked as subcomponent, which is permitted inside a component and does not count as a model kind. C is invalid because a component contains descendant component prims. D is invalid because Wheels (group) appears under a component, placing a model kind below a component boundary. This aligns with Content Aggregation → Asset Structure → Model Hierarchy → Model Kinds .

A practical mixed-format strategy is to use human-readable USDA for lighter, structural, or non-geometric layers, while using binary USDC for heavier geometry and animation data. NVIDIA’s Learn OpenUSD layer guidance identifies .usd, .usda, and .usdc as USD layer formats, where each layer is a modular source of scene description that can participate in composition. ( docs.nvidia.com ) This means a scene can compose different layers with different encodings through sublayers, references, payloads, and other arcs.

Option B is correct because it reflects the common pipeline division: USDA is convenient for inspection, debugging, review, and hand-authored structural opinions, while USDC is better suited for dense numeric data such as meshes, point caches, and time-sampled animation because it is compact and efficient for machine consumption. The official OpenUSD toolset also supports conversion between formats using tools such as usdcat, allowing pipelines to choose the representation that best fits each layer’s purpose. ( openusd.org )

Option A is too rigid; shading data does not require USDC. Option C reverses the usual debugging distinction because USDA is the readable format commonly preferred when inspecting scene text. This aligns with Data Exchange → USD File Formats, Layer Composition, USDA, USDC, and Pipeline File Strategy .

Both cones should point along the Z-axis . The decisive authored property on each Cone prim is uniform token axis = "Z", which defines the cone’s local geometric axis. The differing upAxis metadata in cone1.usda and cone2.usda does not automatically rotate referenced geometry when those assets are composed into coneScene.usda. NVIDIA’s Learn OpenUSD units guidance states that upAxis, metersPerUnit, and kilogramsPerUnit are manually handled metadata during composition, while only timeCodesPerSecond is automatically reconciled. It specifically explains that USD does not automatically reconcile geometric unit metadata across composed layer stacks.

Option A is correct because the root stage is Z-up, the DCC applies no corrective transform, and both referenced cone prims explicitly author their primitive axis as Z. Option B incorrectly assumes that the source layer’s upAxis = "Y" causes automatic conversion into the root stage’s coordinate system. Option C and D likewise invent orientation changes that are not authored. In production, pipelines must validate or explicitly correct up-axis mismatches through transforms or import/export policy. This aligns with Pipeline Development → Stage Metadata, upAxis, Asset Assembly, References, and Manual Unit Reconciliation .

The valid asset-structure principles are Legibility , Navigability , and Modularity . NVIDIA’s Learn OpenUSD asset-structure guide identifies four principles of scalable asset structure: Legibility , Modularity , Performance , and Navigability . It defines legibility as making an asset structure easy to understand and interpret, modularity as enabling flexibility and reuse, and navigability as making it easy for users to find and access the features and properties they need. ( docs.nvidia.com )

Option A is correct because clear names, organization, and intent make assets easier to troubleshoot, exchange, and maintain. Option C is correct because downstream users and tools must be able to locate meaningful prims, properties, variants, payloads, and interfaces efficiently. Option D is correct because reusable modular components support parallel workstreams and scalable aggregation. Option B is incorrect because redundancy is generally discouraged; NVIDIA’s modularity guidance emphasizes reuse and avoiding duplicated data. Option E is incorrect because “compressability” is not one of the stated principles. This aligns with Content Aggregation → Asset Structure Principles → Legibility, Modularity, Performance, Navigability .