I am learning RC software couple of days and I guess that I am making something wrong (btw. I am using the demo version).
The thing is that I want to have a full mesh without holes. The quality dont have to be super precize, I can later tweak it to “render” it for hours. So I manage to make a good looking sparse cloud point but after Normal Detail reconstruction my model is lack of some places. I tried to increase the triangles ammount but it didnt helpt but I did a trick that was a solution to my problem but dont know if this is right or wrong.
So on the first image (bird01) you can see my mesh with Tie Points visible. You can clearly see that on the tail there are some point but on on the mesh. The settings here were on every camera I had “Downscale for depth-map” = 1
On second image you can see full tail with my settings. Here I set all cameras to “Downscale for dept-map” to 2 and then I just selected couple of them and made changes the Downscale to 1. You can see the selected cameras on last image.
So this is strange for me because I thought that having all cameras with Downscale to 1 I should get best results but I found that it is better to set evey single camera to 2 and then in problematic plasce to 1. Why is that ?
I thought that maybe I was limited with triangles but the first image has 10.1 M tris and the last 3.8 M tris.
Any ideas why is that?