SPOM - Silhouette Parallax Occlusion Mapping

Silhouette Parallax Occlusion Mapping (SPOM) is an advanced rendering technique used to improve the edge appearance of standard Parallax Occlusion Mapping (POM), enabling surface details to have more realistic depth and accurate silhouette edges.

Normally, Parallax Occlusion Mapping creates a sense of depth on a flat surface, making it appear as if real geometry exists on the model. However, when viewed from an angle (side view), the edges of the surface still appear flat. The silhouette enhancement addresses these issues:

• Edge Depth: Ensures that edges appear to have real geometric depth even when viewed from the side.
• Enhanced Visual Quality: Accurately represents the object's outer contours (silhouette) using techniques such as Prism POM or custom ray-tracing approaches.
• Performance Efficiency: Delivers highly detailed surface appearance without relying on tessellation or truly high-polygon models, making more efficient use of GPU resources.

Features

• Ability to clip edges and form silhouettes
• Toggle between SPOM and POM
• Distance-based performance adjustment
• High accuracy with Depth Offset (HDRP only)
• Lighting support
• Tiling option
• Real-time surface displacement
• Compatible with URP & HDRP
• Shader Graph integration

---------------------------------------------------------------------------------

So, what is SPOM, and how does it differ from SSDM? You can access the package below and read the comparisons between the two systems.

SSDM - Screen Space Displacement Mapping

🔷 SPOM 🔶 SSDM

| PERFORMANCE

🔷 Performance cost depends on the number of steps. Fewer steps result in higher performance. In 1080p testing, the processing cost is up to ~0.5 ms.

🔶 Since the system is a screen-space technique, approximately 90% of the performance cost depends on screen resolution. In 1080p testing, the processing cost is up to ~1 ms.

| CLIPPING

🔷 Clipping is UV-based. Clipping occurs only at UV seams. The object can still be deformed using rigging or shape keys; however, clipping will still occur along the UV boundaries.

🔶 Clipping is screen-space based. Therefore, clipping behavior depends on the object's visibility within the screen rather than its UV boundaries.

| LOD

🔷 The number of steps can be reduced based on distance, allowing performance to be improved for distant objects.

🔶 LOD is used primarily for visual accuracy rather than performance optimization. Settings such as Pyramid Levels can be adjusted to control image quality.

| ACCURACY

🔷 Since the system operates inside the shader using UV-based mapping, it provides high / deterministic accuracy across the surface.

🔶 Since the system is screen-space based, certain artifacts may occur. Objects or geometry located in front of the SSDM surface can also affect the result.

| USE CASES

🔷 Due to its high accuracy, it is recommended primarily for interiors. It is well suited for close-range and controlled viewing conditions.

🔶 Due to its screen-space nature, it is recommended primarily for outdoor environments and areas where the SSDM surface can be clearly observed.

| DISPLACEMENT

🔷 Since the system is UV-based, it cannot move beyond the boundaries of the object's surface. Therefore, displacement always occurs inward, and the system cannot produce outward protrusions.

🔶 Since the system is screen-space based, pixels can be displaced outward as well. SSDM works by modifying screen-space displacement vectors, allowing displacement to affect silhouettes as well.

| PDO (Pixel Depth Offset)

🔷 Supports Pixel Depth Offset. However, PDO is only available in HDRP. Since URP does not support PDO, this feature is not available when using SPOM in URP.

🔶 Not currently available, since URP does not provide Pixel Depth Offset support.