/**
* Multiplies a specific bone of a different WeightMap onto all boneWeights of this one.
* Use this for more exact fading
*/
public void Multiply(WeightMap map, string bone)
{
for (int i = 0; i < Width; i++)
{
for (int j = 0; j < Height; j++)
{
double fac = 0.0;
if (map.Weights[i, j].ContainsKey(bone))
{
fac = map.Weights[i, j][bone];
}
foreach (Bone b in Bones)
{
if (Weights[i, j].ContainsKey(b.Name))
{
Weights[i, j][b.Name] *= fac;
}
}
}
}
}
static void Main(string[] args)
{
/**
* General concept:
* Unity (well, the version used by VRChat) only supports up to 4 bones per vertex.
* Therefore I use this basic algorithm to generate the weights for bat-like wings
* by limiting the bones per vertex to 2 from the body side and 2 from the finger side.
* The hard edges in the weight map are then smoothed and all weights normalized.
* Then the weight maps are exported to PNG images that can be imported to Blender
* using some plugins.
*
* It works for me but I can't promise that it works well for anyone else. Use at your own risk!
*
* The code below generates the weights for the wing membrane between body and last finger.
* It shouldn't be too hard to adapt for the other wing membranes
*/
// By default the images will be saved in the bin directory
string exportPath = Directory.GetCurrentDirectory();
string exportBaseName = Path.Combine(exportPath + "bone");
double edgeSmoothingFac = 10;
int exportedImageRes = 100;
/*
* One bone for the entire body and one for the entire pinky.
* This will define the general influence from the body towards the finger and the other way
*/
HashSet <Bone> bonesBodyFinger = new HashSet <Bone>
{
/**
* Find the bone start and end positions (in pixel) on your UV map, enter them here.
* Change 1024 to the pixel size of your UV map.
*/
new Bone("body", new Vector(470, 1020) / 1024, new Vector(1024, 1020) / 1024),
new Bone("finger", new Vector(542, 585) / 1024, new Vector(995, 553) / 1024)
};
WeightMap weightMapBodyFinger = new WeightMap(exportedImageRes, exportedImageRes);
weightMapBodyFinger.CalculateDistances(bonesBodyFinger);
//weightMapBodyFinger.ExportToImage(exportBaseName);
// This defines the body bones
HashSet <Bone> bonesBodyDetail = new HashSet <Bone>
{
// I did the weights of the arm by hand (I don't know if this actually works well)
//new Bone("upperArm", new Vector(562, 968) / 1024, new Vector(549, 795) / 1024),
new Bone("upperleg", new Vector(1024, 1020) / 1024, new Vector(852, 1020) / 1024),
new Bone("hips", new Vector(825, 1020) / 1024, new Vector(733, 1020) / 1024),
// I got better results by combining spine and chest: 1 bone less
//new Bone("spine", new Vector(733, 1020) / 1024, new Vector(665, 1020) / 1024),
//new Bone("chest", new Vector(665, 1020) / 1024, new Vector(470, 1020) / 1024)
new Bone("spine", new Vector(733, 1020) / 1024, new Vector(470, 1020) / 1024)
};
WeightMap weightMapBodyDetail = new WeightMap(exportedImageRes, exportedImageRes);
weightMapBodyDetail.CalculateDistances(bonesBodyDetail);
weightMapBodyDetail.LimitWeightCount(2);
weightMapBodyDetail.SmoothEdge("upperleg", new Bone("", new Vector(.78, 0), new Vector(.78, 1)), edgeSmoothingFac);
weightMapBodyDetail.SmoothEdge("spine", new Bone("", new Vector(.78, 0), new Vector(.78, 1)), edgeSmoothingFac);
//weightMapBodyDetail.SmoothEdge("hips", new Bone("", new Vector(.69, 0), new Vector(.69, 1)), edgeSmoothingFac);
//weightMapBodyDetail.SmoothEdge("chest", new Bone("", new Vector(.69, 0), new Vector(.69, 1)), edgeSmoothingFac);
weightMapBodyDetail.Normalize();
weightMapBodyDetail.Multiply(weightMapBodyFinger, "body");
weightMapBodyDetail.ExportToImage(exportBaseName);
// This defines the pinky bones
HashSet <Bone> bonesBodyFingerDetail = new HashSet <Bone>
{
// I did the weights of the arm by hand (I don't know if this actually works well)
//new Bone("lowerArm", new Vector(549, 795) / 1024, new Vector(538, 610) / 1024),
new Bone("pinky1", new Vector(542, 585) / 1024, new Vector(680, 567) / 1024),
new Bone("pinky2", new Vector(680, 567) / 1024, new Vector(838, 556) / 1024),
new Bone("pinky3", new Vector(838, 556) / 1024, new Vector(995, 553) / 1024)
};
WeightMap weightMapFingerDetail = new WeightMap(exportedImageRes, exportedImageRes);
weightMapFingerDetail.CalculateDistances(bonesBodyFingerDetail);
weightMapFingerDetail.LimitWeightCount(2);
weightMapFingerDetail.SmoothEdge("pinky1", new Bone("", new Vector(.71, 0), new Vector(.77, 1)), edgeSmoothingFac);
weightMapFingerDetail.SmoothEdge("pinky3", new Bone("", new Vector(.71, 0), new Vector(.77, 1)), edgeSmoothingFac);
weightMapFingerDetail.Normalize();
weightMapFingerDetail.Multiply(weightMapBodyFinger, "finger");
weightMapFingerDetail.ExportToImage(exportBaseName);
}
