Hi all,<div><br></div><div>I retrieved the normal vectors by </div><div><br></div><div><span class="Apple-tab-span" style="white-space:pre">        </span>vtkFloatArray* normals = static_cast<vtkFloatArray*>(streamer->GetOutput()->GetPointData()->GetArray("Normals"));</div>
<div> float normal[3];</div><div><span class="Apple-tab-span" style="white-space:pre">        </span>for (int i = 0; i < normals->GetNumberOfTuples(); i++)</div><div><span class="Apple-tab-span" style="white-space:pre">        </span> {</div>
<div><span class="Apple-tab-span" style="white-space:pre">        </span> normals->GetTupleValue(i, normal);</div><div><span class="Apple-tab-span" style="white-space:pre">        </span> vtkMath::Normalize(normal);</div><div>
<span class="Apple-tab-span" style="white-space:pre">        </span> cout << normal[0] << ", " << normal[1] << ", " << normal[2] << endl;</div><div><span class="Apple-tab-span" style="white-space:pre">        </span> }</div>
<div><br></div><div><br></div><div>However, I saw in the output from above many zero vectors (0, 0, 0), and also (nan, nan, nan).</div><div><br></div><div>So it seems there exists bugs in vtkStreamTracer's normal computation routine. Is there anyone who had this problem?</div>
<div><br></div><div>Yifei</div>