How do you get the vertex data from the spatial meshes?

rwblodgett

I'm trying to be able to do a couple of things.

1. know where a point intersects the spatial meshes.
2. place a virtual object on the floor.

I'm not using Unity.

I found the example in the Universal Windows Samples that shows the mesh.

How do I get any of the vertex data out of it, in order to, I guess, do triangle/ray intersection on it? (if this is the best way to find a point on the floor).

This is what I'm trying, but my array is just filled with mostly 0s and extremely large exponent numbers (mostly zeros though).

m_PositionFloats = ToArray(m_surfaceMesh->VertexPositions->Data, &m_PositionFloatsLength);

My ToArray function is defined by:
`

float* ToArray(Windows::Storage::Streams::IBuffer^ buffer, int* len)
{
auto reader = ::Windows::Storage::Streams::DataReader::FromBuffer(buffer);

int length = reader->UnconsumedBufferLength;
unsigned char* data = new unsigned char[length];
memset(data, 0, length);

reader->ReadBytes(Platform::ArrayReference<unsigned char>(&data[0], length));

float* floats = ToFloats(data, length);
*len = length / 2;

delete[]data;

return floats;

}
`

My ToFloats function is defined by:
`

float* ToFloats(unsigned char* data, int length)
{
int floatLength = length / 2;

float* floats = new float[floatLength];

uint16_t * p = reinterpret_cast<uint16_t *>(data);

for (int i = 0; i < floatLength /2; i++)
{
    uint16_t temp = p[i];
    uint16_t in = temp;

    uint32_t t1;
    uint32_t t2;
    uint32_t t3;

    t1 = in & 0x7fff;                       // Non-sign bits 
    t2 = in & 0x8000;                       // Sign bit 
    t3 = in & 0x7c00;                       // Exponent 

    t1 <<= 13;                              // Align mantissa on MSB 
    t2 <<= 16;                              // Shift sign bit into position 
    t3 >>= 8;


    t1 += 0x38000000;                       // Adjust bias 

    t1 = (t3 == 0 ? 0 : t1);                // Denormals-as-zero 

    if (t3 > 0)
    {
        t3 -= 15; // remove exponent bias for 16bit
        t3 += 127; // add exponent bias for 32bit
        t3 <<= 23;
        t1 |= t3;
    }


    t1 |= t2;                               // Re-insert sign bit 


    float f = *reinterpret_cast<float*>(&t1);

    floats[i] = f;
}

return floats;

}

`

Patrick

I've done this recently. I tried to do the SHORTN4 conversion myself, but wasn't successful. The XMxxx apis know how to handle this though. I'm sure there's a better way than this, but this is what I've got working:

XMSHORTN4* rawVertexData = (XMSHORTN4*)GetDataFromIBuffer(m_surfaceMesh->VertexPositions->Data);
float3 vertexScale = m_surfaceMesh->VertexPositionScale;
int InputVertexCount = m_surfaceMesh->VertexPositions->ElementCount;
for (int index = 0;index < InputVertexCount;index++)
    {
        // read the currentPos as an XMSHORTN4. 
        XMSHORTN4 currentPos = XMSHORTN4(rawVertexData[index]);
        XMFLOAT4 xmfloat;

        // XMVECTOR knows how to convert the XMSHORTN4 to actual floating point coordinates. 
        XMVECTOR xmvec = XMLoadShortN4(&currentPos);

        // STore that into an XMFLOAT4 so we can read the values.
        XMStoreFloat4(&xmfloat, xmvec);

        // Which need to be scaled by the vertex scale.
        XMFLOAT4 scaledVector = XMFLOAT4(xmfloat.x*vertexScale.x, xmfloat.y*vertexScale.y, xmfloat.z*vertexScale.z, xmfloat.w);

float4 nextFloat = float4(scaledVector .x, scaledVector .y, scaledVector .z, scaledVector .w);
}

Patrick

I've done this recently. I tried to do the SHORTN4 conversion myself, but wasn't successful. The XMxxx apis know how to handle this though. I'm sure there's a better way than this, but this is what I've got working:

XMSHORTN4* rawVertexData = (XMSHORTN4*)GetDataFromIBuffer(m_surfaceMesh->VertexPositions->Data);
float3 vertexScale = m_surfaceMesh->VertexPositionScale;
int InputVertexCount = m_surfaceMesh->VertexPositions->ElementCount;
for (int index = 0;index < InputVertexCount;index++)
    {
        // read the currentPos as an XMSHORTN4. 
        XMSHORTN4 currentPos = XMSHORTN4(rawVertexData[index]);
        XMFLOAT4 xmfloat;

        // XMVECTOR knows how to convert the XMSHORTN4 to actual floating point coordinates. 
        XMVECTOR xmvec = XMLoadShortN4(&currentPos);

        // STore that into an XMFLOAT4 so we can read the values.
        XMStoreFloat4(&xmfloat, xmvec);

        // Which need to be scaled by the vertex scale.
        XMFLOAT4 scaledVector = XMFLOAT4(xmfloat.x*vertexScale.x, xmfloat.y*vertexScale.y, xmfloat.z*vertexScale.z, xmfloat.w);

float4 nextFloat = float4(scaledVector .x, scaledVector .y, scaledVector .z, scaledVector .w);
}

Patrick

BTW, writing your own physics can be a pretty daunting task. The UWP Marble Maze sample might help. I haven't tried, but you might be able to get the Bullet physics engine to work.

rwblodgett

Thank you, Patrick. Huge help.

rwblodgett

Any idea how to do this in c#? I"m trying to port the spatial meshes example to c# ....

druidmechanics

I'm looking for FBX model import into c++ so that I'm not relying on Unity, and I'm interested to hear from anyone doing the same.