Realistic PBR Surface for UE5

Visual Target

RealisticPBRBaseline

What we are building: a clean realistic surface setup that respects PBR ranges, uses packed masks, exposes useful instance controls, and avoids accidental stylization unless the art direction asks for it.

Image placeholder: realistic PBR material boardGenerate later: material balls or surface swatches showing stone, metal, painted plastic, and rough wood, neutral lighting, 16:9 landscape, no text.

Realistic material work starts with boring discipline: correct values, correct scale, and predictable instances.

Use Cases

RealisticProps and environment assetsUse when surfaces need to behave predictably under production lighting.

PipelineMaterial master baselineGood as a starting point before specialized wood, metal, fabric, or stone materials.

ReviewPBR sanity checkHelps catch albedo, roughness, and metalness values that drift outside believable ranges.

Mental Model

PBR is a contract with lighting

The material should respond predictably when the lighting changes. Baked shadows, fake highlights, or over-painted contrast inside albedo break that contract.

Source textures do the heavy lifting

A realistic master should not rescue broken scans or hand-painted maps. It should expose small, safe corrections that help assets fit the level.

Realistic still needs art direction

Realism is not raw data dumped into a shader. Scale, value range, roughness response, and detail intensity still need taste and review.

Graph Steps

Scale checkAlbedo rangeORM masksNormal detailInstance biasLighting QA

1. Lock texture scale before beauty tuning

Check texel density, UV scale, and real-world material size. A good scan at the wrong scale will look fake no matter how clean the shader is.

Beginner check: compare the material next to a known prop size, door, wall tile, or character height.

2. Validate albedo range

Keep base color free of baked shadows, direct highlights, and extreme values unless the material is intentionally stylized. The lighting should come from the scene.

Senior note: albedo mistakes often hide until the asset moves from neutral preview lighting into a real level.

3. Use packed masks with clear channel meaning

Follow project convention for ORM or packed utility masks. Document whether red is AO, green is roughness, blue is metalness, or if the project uses a different order.

4. Treat roughness as a physical response

Roughness is not just a contrast map. It controls reflection spread and material identity. Review roughness under moving camera and different light sizes.

5. Normalize normal and detail strength

Expose normal intensity, detail normal amount, and macro variation separately. Do not let detail normals overpower the sculpt or create noisy silhouettes.

6. Review under multiple lighting cases

Check indoor, outdoor, warm, cool, bright, and low-light setups. A robust PBR material should not collapse when the lighting changes.

Artist Controls

ColorTint StrengthSmall correction, not a replacement for correct albedo.

ValueAlbedo Brightness BiasUseful for level matching; keep range limited.

SurfaceRoughness BiasBalances response per lighting setup without repainting maps.

NormalNormal IntensityKeeps scan or sculpt detail from overpowering the asset.

DetailMacro VariationBreaks repetition on large surfaces while preserving material identity.

PipelineUse Detail Normal / Use MacroStatic switches or feature toggles for cheaper background instances.

HLSL Sketch

float3 albedo = Texture2DSample(albedoTex, albedoSampler, uv).rgb;
float3 orm = Texture2DSample(ormTex, ormSampler, uv).rgb;
float3 normalTS = UnpackNormal(Texture2DSample(normalTex, normalSampler, uv));

float occlusion = orm.r;
float roughness = saturate(orm.g + roughnessBias);
float metalness = saturate(orm.b);

albedo = lerp(albedo, albedo * tintColor, tintStrength);
normalTS = normalize(lerp(float3(0, 0, 1), normalTS, normalIntensity));
return float4(albedo * occlusion, 1.0);

Senior Notes

Realistic does not mean uncontrolled

A realistic master still needs art direction controls. The difference is that controls should preserve plausible ranges instead of hiding broken source maps.

Review the asset family, not one hero sample

A good PBR master needs to handle rough wood, painted metal, stone, plastic, and clean/dirty variants without becoming a parameter graveyard.

Memory can matter more than math

Texture resolution, packed masks, streaming behavior, and duplicate material instances can cost more than a few scalar operations. Audit texture use as part of material review.

Common Mistakes

Using albedo with baked shadows or highlights.
Treating roughness as a beauty contrast map instead of reflection behavior.
Letting metalness sit in gray values for ordinary non-layered materials.
Adding too many tweak controls that hide bad source textures.
Approving the material under one lighting setup and never checking it in the target level.