Hello,
Just got my Inspire and threw myself straight into a new project.
I’v managed to start getting some good/acceptable scans after trial and error.
I wanted to scan the inside and outside of a cover.
Tried to make one big scan using a rotation table for a whole day until I gave up and checked YT.
Most videos seem to be scanning and merging two scans from different sides.
I thought I’d give that a go.
Made two scans and tried merging both with markers and with feature.
I’v tried merging both in point cloud and meshed with the same result.
I’v also tried putting the marker on the “right side” of the object with the same result.
The idea is that the purple object should be inside the yellow object
Yellow is top/outside. Purple is the inside of the cover.
marker’s mode works with 2 corresponding points on different clouds. If you don’t have enough overlapping ( more than 10% ) and you need them, you have to tell the app what point is the same point on different clouds/meshes. But here you are telling the app that your “inside” points are the same as points that are in another cloud… the “outside” cloud.
Then the app will try to match them, but clearly that can’t happen… because they are from different clouds. It doesn’t matter how many you add… you need some overlapping to tell the app “this point here (yellow mesh)” is the same point as "this point there “purple mesh”.
I hope that makes sense. You are not going to be able to merge these inside RS5, even you make a new scan with overlapping… or you try these 2 in some other software like zbrush.
How accurate has the marker alignement to be? Only 2 ore more roughly corresponding points or exactly the same points of the surface? The latter is often difficult…
I’ve scanned it on either side so that the top and the ends are a common feature, however, feature alignment tries match the top and the opposite sides:
Marker mode is no use since all points either have to be spot on or it just appears to average the error - I’m pretty certain that marker mode used to have a level of intelligence in that it would use the markers to get rough orientation and then analyse the point cloud. As it stands, I would much prefer total manual control to manipulate the scans for merging.
