The Blend SOP provides 3D metamorphosis between shapes with the same topology. It can blend between sixteen input SOPs using the average weight of each input's respective channel. It will also interpolate point colors and/or texture co-ordinates between shapes.
For best results, there should be the same number of points / CVs (and faces) in the geometry of each SOP. The points should have a similar order; if point 17 in one source is on the left side, and point 17 in another Source is on the right side, it will move across the object during the blend, which may create distorted and twisted shapes. To ensure that this doesn't happen, you can make all the pieces to be blended by editing point positions of one common base shape. Each of the pieces to be morphed must be in different SOPs.
For example, when editing the shapes in the Model Editor, each particular geometry can be saved and then read in, or input the Model SOP directly into the Blend SOP. You would then have the base model geometry entering the Blend SOP in the first Blend, then up to 15 Model SOPs feeding from the SOP containing the base model geometry, which in turn would feed into the Blend SOP. Remember that Model SOPs cannot be unlocked and, therefore, are a safe way to store data without using a File SOP.
The Blend SOP now only cooks non-zero-weighted inputs.
Group - Specifies a point or primitive group in the first input. If, for example, a group is specified containing the first and third points, then the first and third point of every input will be blended whereas the second, fourth, fifth, etc. points will be set to match the first input source. Accepts patterns, as described in: Scripting > Pattern Matching.
Blend 1-4 ... 13-16
/blend1 ... /blend16 - Channels controlling the contribution of the geometry inputs to the output geometry.
Differencing - Generates exaggerated blends between objects where values above 1 or less than 0 will result in over-scaled blends.
When this option is checked, the above channel values are not summed and scaled to 1. The first input is considered a reference; Blend computes the difference between the first input and the others for each point; values greater than 1 and less than 0 produce exaggerations of the shapes for inputs 2 and higher. The first input, however, cannot be exaggerated by its blend channel, /blend1, and it is considered to be the "base". When using Differencing, the /blend1 channel has no effect. If the geometry in the first input must be deformed, feed it into another input, where the blend channels have an effect.
Differencing - Enables difference blending.
Blend Position - When checked, the point positions of the inputs will be blended based on the weights of the blend channels. If not checked, the input geometry will not change, only allowing Blending of Colors, Normals and Textures if selected.
Blend Colors - When checked, the point colors of the geometry inputs will be blended based on the weights of the blend channels.
Blend Normals - When checked, the point normals of the geometry inputs will be blended based on the weights of the blend channels.
Blend Texture - When checked, the point texture co-ordinates of the geometry inputs will be blended based on the weights of the blend channels.
Inputs / Geometry Types
Accepts all geometry types for up to sixteen inputs.
Uses / Works in Relation With
This SOP can be used as a form of geometry deformation. Model your geometry, such as the mouth on the head of a character. This will become your first input.
Set the input to the first SOP to be blended. For example, if there are five shapes to be blended, the simplest set-up is the following:
Use four Model SOPs, and import the geometry from the base geometry to be deformed / morphed. In each of the Model SOPs, edit the geometry to suit.
Append a Blend SOP off of the SOP containing the base geometry and then proceed to connect up the four Model SOPs into the Blend SOP.
The amount that each input contributes to the resulting shape is controlled by each input's blend channel. Editing the values of the blend channels will result in each input to determine the final shape based on ratios.
For example, if the
/blend3 channel's value is 1, and the rest of the channels are set to 0, the resulting shape is the geometry in the third input SOP. Now change the value in the
/blend5 channel to 1, the resulting shape will be a 50/50 percent blend of the geometry contained in the input SOPs' three and five. You would get the exact same result by setting both channels to 0.3, or to any number if the numbers are the same in both channels, there will be equal contribution to the shape.