camera-outputs-rt2

Identify and configure Real-Time Path-Tracing camera render outputs for Isaac Sim workflows.

Updated Apr 29, 2026
One-click install
npx skills add https://github.com/PubCyBerry/SO101-Sim2Real --skill camera-outputs-rt2-pubcyberry
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: camera-outputs-rt2
Source: https://github.com/PubCyBerry/SO101-Sim2Real/tree/main/.agents/skills/camera-outputs-rt2
Command: npx skills add https://github.com/PubCyBerry/SO101-Sim2Real --skill camera-outputs-rt2-pubcyberry

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps developers identify and use the correct Real-Time Path-Tracing camera render outputs without guessing formats, data types, or readback requirements.

Core Features & Use Cases

  • Camera Output Discovery: Provides available RT2 AOVs and RenderVar source names with their formats, dtypes, shapes, and intended uses.
  • Render Pipeline Guidance: Guides USD RenderProduct configuration and output selection while avoiding unnecessary memory overhead from extra AOVs.
  • Use Case: Use this Skill when building an Isaac Sim camera workflow that needs depth, normals, albedo, semantic IDs, or color buffers for robotics perception and rendering tasks.

Quick Start

Ask the skill to identify the correct RT2 camera output and explain how to configure and read it.

Frequently Asked Questions about camera-outputs-rt2

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
What are the available AOV camera outputs for Real-Time Path-Tracing in Isaac Sim?

Real-Time Path-Tracing camera outputs in Isaac Sim include color, depth, normals, albedo, and semantic ID AOVs. Each RenderVar source name provides specific formats, dtypes, and shapes for robotics perception and rendering tasks.

How do I configure a RenderProduct for RT2 camera output selection?

To configure a RenderProduct for RT2 camera output, select the required AOVs and RenderVar sources while avoiding unnecessary memory overhead. This ensures correct output format discovery and dtype validation for your simulation workflow.

Can I read depth and normals from an Isaac Sim camera without guessing data formats?

Yes, you can read depth and normals without guessing by identifying the correct RT2 camera outputs. This provides documented API usage patterns, dtype and shape validation, and reliable camera data readback for robotics applications.

What is the best way to avoid memory overhead when selecting AOVs for simulation rendering?

The best way to avoid memory overhead is to select only the necessary RT2 AOVs during RenderProduct configuration. This prevents loading extra AOVs while ensuring you capture the required color, depth, or semantic buffers for your workflow.

Why does my Isaac Sim camera readback fail due to incorrect dtype or shape validation?

Isaac Sim camera readback fails when using incorrect dtypes or shapes for RT2 outputs. Identifying the correct AOV formats and RenderVar source names ensures proper output format discovery and successful data readback.

Do I need to specify RenderVar source names for semantic IDs in my RT2 rendering pipeline?

Yes, specifying RenderVar source names for semantic IDs is required in RT2 rendering. This ensures the RenderProduct is correctly configured to capture semantic segmentation data with the right formats and dtypes for perception tasks.