The specular weight. Influences the brightness of the specular highlight.
The color the specular reflection will be modulated with. Use this color to 'tint' the specular highlight. You should only use colored specular for certain metals, whereas non-metallic surfaces usually have a monochromatic specular color. Non-metallic surfaces normally do not have a colored specular.
Controls the glossiness of the specular reflections. The lower the value, the sharper the reflection. In the limit, a value of 0 will give you a perfectly sharp mirror reflection, while 1.0 will create reflections that are close to a diffuse reflection. You should connect a map here to get variation in the specular highlight.
The 'microscopic' features of a surface affect the diffusion and reflection of light. This 'micro surface' detail has the most noticeable effect on specular reflections. In the diagram below, you can view parallel lines of incoming light commence to diverge when reflected from rougher surfaces when each ray hits a part of the surface with a different orientation. In summary, the rougher the surface becomes, the more the reflected light will diverge or appear 'blurred.'
'Microsurface' detail represented as a general measure of roughness (this surface would have a high Specular Roughness value).
The brightness of the Specular highlight is automatically linked to its size due to the Standard Surface Surface shader's energy conserving nature. In the example below, all of the materials are reflecting the same amount of light, but the rougher surface is spreading it out in multiple directions. However, with low amounts of roughness, the surface is reflecting a more concentrated amount of light.
To get variation in the highlights of the surface, a map should be connected to the Specular Roughness. This will influence not only the brightness of the highlight but also its size and the sharpness of the environmental reflection.
You may need to connect the file texture to a Range shader to see any results in the Specular Roughness.
'Fingerprint' texture connected to Specular Roughness
The specular roughness affects both specular reflection and refraction. There is also a Transmission Extra Roughness parameter to add some additional roughness for refraction if required. You can, however, use Coat to create a rough reflection layer over a sharp refraction.
'Scratches' texture connected to Specular Roughness via a Range shader
The IOR parameter (Index of Refraction) defines the material's Fresnel reflectivity and is by default the angular function used. Effectively the IOR will define the balance between reflections on surfaces facing the viewer and on surface edges. You can see the reflection intensity remains unchanged, but the reflection intensity on the front side changes a lot.
Specular IOR with Transmission
The default value of 1.0 is the refractive index of a vacuum, i.e., an object with IOR of 1.0 in empty space will not refract any rays. In simple terms, 1.0 means 'no refraction'. The Standard Surface shader assumes that any geometry has outward facing normals, that objects are embedded in the air (IOR 1.0) and that there are no overlapping surfaces.
When rendering refractive surfaces, it is very important that the normals of the geometry face in the right direction. In the example below (left), you can see the difference between normals that are facing in the right direction (outward), versus those that are facing inwards (incorrect). This is especially important when rendering surfaces with double-sided thickness, such as glass. However, with air bubbles in glass (below center), the reverse is true. The bubble geometry normals should be reversed and the bubbles should be combined with the glass geometry. Normal direction is equally important when rendering single sided surfaces such as a car windscreen (right).
Normals pointing outward (correct). Rollover image to view normals pointing inward (incorrect).
Air bubbles in glass: Normals should point inward. Rollover image to view normals pointing outward (incorrect).
Windscreen model (single sided). Rollover image to view normals pointing inward (incorrect).
If you see any black where there should be refraction, you may not have a high enough Transmission Ray Depth value (found in the Ray Depth section in the render settings). The default value is eight, which is sufficient for most cases.
Anisotropy reflects and transmits light with a directional bias and causes materials to appear rougher or glossier in certain directions. The default value for Anisotropy is 0, which means 'isotropic.' As you move the control towards 1.0, the surface is made more anisotropic in the U axis.
Anisotropy is suitable for materials that have a clear brush direction such as brushed metal which has tiny grooves in which form a 'stretched' anisotropic reflection.
Many small discs form together to create an effect which is the anisotropic highlight
Anisotropic reflections are suitable for brushed metal effects such as in the example below:
Texture assigned to Specular Anisotropic Rotation
You may notice faceting appear in specular highlights when using anisotropy. It is possible to remove the faceted appearance by enabling smooth subdivision tangents (via Arnold subdiv_smooth_derivs parameter). Take into account this requires a subdivision iteration of at least one in the polymesh to work.
Enable Subdivision and increase Subdivision Iterations to remove anisotropic specular faceting
More information about Specular 'Anisotropy' can be found here.
The rotation value changes the orientation of the anisotropic reflectance in UV space. At 0.0, there is no rotation, while at 1.0 the effect is rotated by 180 degrees. For a surface with brushed metal, this controls the angle at which the material was brushed. For metallic surfaces, the anisotropic highlight should stretch out in a direction perpendicular to the brushing direction.
It is possible to assign textures to the specular rotation. When doing so, it is advisable to avoid texture filtering. This means disabling MIP-mapping and disabling the magnification filter, which by default is set to "smart bicubic." One way is to set the mipmap_bias of the image node to a strong negative value, like -8, which means "use 8 MIP levels higher resolution than usual".