Hello 
We are just trying to get some good results with a face (180°).
We made 84 pictures in four angels.
The pictures are sharp and there is enough overlapping.
But still we get extremly bad results.
The lighing is not realy perfect, but I already got waaaay better results, even when the light was not perfect.
I attached an example picture and the result.
I just dont get it - never had so strange results - but also never made a human face…
Resolution is 4k
5D Mark III
EF40mm /2.8 STM
Well, the person is moving slightly while u made 80+ photos. the software has problems with that. the object has to be perfectly static and thats only possible by shooting with many cameras simultaniously
Hello Markus,
I’m guessing you’re taking the images sequentially, not all at once, right?
Because there’s always some subtle body movements you’re obviously going to end up with shifted geometry.
Notice the good results for the carpet in the background and the bad results for the moving model in the front.
The only way around that is keeping the model as still as possible so there’s no breathing related movement and be fast while taking the pictures.
If you’re only after the geometry you could also try painting the models face, which would probably lead to better results.
Or have multiple cameras and take them all at once.
Also in my opinion there’s a lot of noise in your first picture, which surely doesn’t help with a bland surface like skin.
Good morining
Thx for your answers, Julian and Lukas!
Makes sense.
We tried to avoid movements by putting the backhead agains the wall.
I try to find a solution to kinda fix the head even more and do quicker photos.
Yes Lukas. the model is more important since we gonna change the look completly, but of course based on the original texture.
Even the model I can edit easily in Mudbox - but as better the result, as less scuplting and time 
Thx again, I gonna try again with quicker photos and less movement of the model.
Markus
Hey,
just wanna give a little update.
I resized the pictures, cropped out only the face, made some Light adjustments and saved everything again (Photoshop).
The result already looks way better. Gonna check it in High Quality now, but with that I guess I could work so far…
Okay, so RC relied more on the facial features which moved, than on the fixed background.
Actually good to know that quick masking helps in such a case.
You could also try using Photoshop’s automatic subject detection masking to batch process the pictures
and get rid of the entire background.
