Texture extraction overview

Why extract a texture map?

Sculpting and subdividing models in Mudbox can produce models with very high polygon counts resulting in large file sizes. While you can easily export your models as either .fbx or .obj files for use in other 3D software and games applications, these software applications may not efficiently handle models with polygon counts in the range of tens of millions.

The recommended workflow is to extract a normal or displacement map (or both) that represents the high resolution surface(s) from Mudbox and export the texture maps so they can be applied to a simplified (low resolution) version of the model in another 3D application. As a result, the fine high resolution detail is maintained on a simpler version of the model.

The type of map you extract depends on your particular needs and production workflow. The Maps > Extract Texture Maps menu item lets you extract high resolution detail from a sculpted mesh in many image formats or in Ptex format. You can extract the following:

The Transfer Paint Layers option lets you transfer multiple paint layers from one model to another, regardless of whether the models have different topology or UVs. For more information, see Extract Texture Maps properties.
An ambient occlusion map is a shading effect that calculates the attenuation and occlusion of light between surface details and other objects in the scene. Mudbox can calculate this information as a high resolution image map and assign it as a paint layer to the selected target model to simulate a realistic shading effect. For more information, see Ambient occlusion maps overview.
A vector displacement map provides both height and directional information for each pixel which provides the ability to displace vertices off-normal allowing users to faithfully recreate complex overlapping shapes. For more information, see Vector displacement maps overview.
A displacement map represents the height difference between the two surfaces as a gray scale image. For more information, see Displacement maps overview.
A normal map extracted in Mudbox represents the difference of the surface normals between the sculpted source model and the target model and represents these differences as a color map. For more information, see Normal maps overview.

For most texture extractions, the source and target meshes can be low and high resolution versions of the same mesh or even arbitrary meshes that have differing topology but suit the production pipeline requirements.

When extracting vector displacement maps, the meshes must either have the same base topology, or one mesh must be identical to a subdivided level of the other.

It is also possible to extract detail from 3D laser scans of a real model and produce a map that can be applied to a digital mesh that has UV texture coordinates at a lower resolution.

How texture extraction works in Mudbox

Extracting texture maps requires two polygonal meshes - a source and a target.

The source model contains the high resolution sculpted detail. Mudbox samples the source mesh to produce a texture map that can be applied to the target model.

The target model receives the output texture map. The target model is normally much lower in resolution than the source mesh. In most situations, and unless you are saving the map in Ptex format, the target model needs to have UV texture coordinates that are non-overlapping, unless the UVs are arranged on the model to produce a symmetrical color map.

Most texture map(s) are extracted in Mudbox and applied to the target model using another 3D application (for example Maya, 3ds Max, or a games console) but can be previewed on the model within Mudbox. Vector displacement maps are an exception. Since they are 32 bit floating point, they can be previewed using the Image Browser or another application capable of viewing 32 bit floating point images.

An extracted displacement or vector displacement map can also be used in Mudbox to displace the vertices on another model (see Sculpt using a displacement map). Vector displacement maps that are exported in tangent-space can be used as stamps or stencils, as a way of easily copying sculpted detail.

Mudbox extracts normal and displacement maps from a polygon model by sampling the detail and offset between the source and target versions of the user-defined models. The difference between the source and target model is written in the UV texture space of the low resolution (target) model. For this reason, the target model must have UV texture coordinates that are suitable for the texture extraction. For more information, see Prepare a model for sculpting.

In Mudbox, you control the behavior of the source and target meshes during extraction, which is crucial for professional results and for obtaining good maps under varying circumstances.

For normal and displacement maps, two extraction methods are possible:

The Ray Casting method extracts a texture map by recording the distance (as a pixel value) between the source and target models based on sample rays that get cast along the surface normals of the target surface. The number of sample rays that occur is dependent on the texture map resolution. This method should be used when extracting texture maps between arbitrary meshes (meshes with differing topologies) and requires that the models be manually aligned as accurately as possible prior to extraction to produce accurate results. Artifacts in the resulting texture map can sometimes occur (particularly with overhanging features) that will require editing afterwards.
The Subdivision method extracts a texture map by recording the difference in position of a sampled point on a model between one subdivision level and the corresponding point at a higher subdivision level. This method should only be used when extracting texture maps between different subdivision levels on the same model. The resulting texture map will contain fewer artifacts compared to the Ray casting method, requires no setup or alignment between separate models, and performs the extraction quicker. The Subdivision method does not capture the height differences as accurately as the Ray Casting method particularly when the vertices on the higher subdivision level have been translated off normal compared to the lower subdivision level such as might occur when using the Grab or Pinch tools.

NoteFor vector displacement maps, the Subdivision method is always used. For this reason, vector displacement maps can only be extracted between different levels of the same mesh (or topologically equivalent meshes).

Mudbox lets you extract maps between polygonal and subdivision surfaces in any combination depending on the method selected. For best results, the low resolution and high resolution surfaces are treated as subdivision surfaces at extraction time. The resulting maps align when your 3D rendering application applies a smoothing operation (Catmull-Clark subdivision) to your low resolution polygon model at render time.

Important

If you are exporting a mesh that has been used as the source in a texture map extraction (for example, if you used this model as the source to extract a displacement map), be aware that any extracted texture maps are automatically exported with your model.

This can result in unexpected deformation when you import the model into another application, as the texture maps are automatically assigned to the model on import. If you experience this issue, you can remove any unwanted texture map from the associated channel (such as the displacement or bump channel) after importing the model.

Supported extraction file formats and bit depths

Mudbox can extract maps in a variety of bit depths and file formats for compatibility with your software rendering application and production pipeline, whether you render using Maya, mentalray, RenderMan, and so on.

Note

Vector Displacement maps can only be extracted as a TIFF, OpenEXR or Ptex file.

File format	Bit depth	Number of channels
BMP	8	3
JPG	8	3
PNG	8	4
TGA	8	1, 3, 4
TIFF	8, 16 integer, 32 floating point	1, 3, 4
OpenEXR	8, 16 float, 32 floating point	1, 3, 4
Ptex	8, 16 integer, 32 floating point	1, 3, 4

Note

If you specify either 8 or 16-bit depth for an extracted texture, the extraction feature samples the surface in 32-bit depth and then converts it to either an 8 or 16-bit normalized set of values that range between zero and one. Vector displacement maps are automatically extracted as 32 bits-per-channel floating point files and can be either in TIFF or OpenEXR format.