ACES 2 Rendering Transform¶
The core architectural change in ACES 2 is a complete overhaul of the rendering transform.
During ACES 1 development, features were introduced incrementally, thus leading to inconsistincies between transforms. However, ACES 2 was designed as a unified system, thus helping to ensure a more consistent appearance across different output targets while also providing a simple mechanism for users to create outputs to custom display setups. The out-of-the-box appearance through ACES 2 transforms is alo less prone to undesirable artifacts, reducing the need for corrective grading and allowing more focus on controlling creative intent.
Key design goals¶
The ACES 2 rendering transforms are designed to:
- Provide tone mapping with a slightly lower mid-slope contrast and gentler highlight rolloff than ACES v1 (See Figure 1)
- Minimize hue shifts across exposure levels within regions of uniform hue
- Include robust gamut mapping to eliminate harsh clipping and preserve perceptual continuity (See Figure 2)
- Fully utilize output encoding ranges, where appropriate (reach extents of display gamut volume)
- Maintain round-trip consistency (Output → Inverse ACES → Output)
- Use a unified algorithmic structure across all display-targeted transforms
- Offer simple parameterization for unsupported or custom display configurations
- Deliver a predictable, high-quality default rendering
Figure 1¶
ACES 2 (left) vs ACES 1 (right)
In ACES 2, the softer tone scale combined with gamut mapping tends to keep colorfulness in highlights longer.
Figure 2¶
ACES v2 (left) vs RGC + ACES 1 (right)
ACES 2 does not exhibit the same clipping artifacts that could often appear when using ACES 1. The comparison image even has the Reference Gamut Compression applied, which was developed specifically to minimize clipping artifacts with ACES 1.
ACES v2 preserves hue and maintains a more well-behaved desaturation as exposure increases.