One of the advantages of being on an archaeological site during its excavation season is the ability to make 3D models of structures at different phases of excavation. During the course of excavation at the Ness of Brodgar, I will be recording the excavation of the curving, northern wall in structure 1, creating 3D models of the wall as it is removed. Sketch plan drawings are created during the course of the removal of a structure such as this, usually after each layer or phase of the structure has been removed, however creating a 3D model of the structure at these intervals can add an extra level of detail for the archaeological record. Indeed, one can view the wall as it was at a certain period of excavation and compare this with the plans which are normally sketched, to aid in interpretation.
“one could be under the impression that
creating the 3D model in sunshine would add
an extra, brighter level of detail,
however there are issues with this”
There are potential problems which need to be avoided when creating a model of a structure such as this. The first issue is the sun, one could be under the impression that creating the 3D model in sunshine would add an extra, brighter level of detail, however there are issues with this. Shadows can be problematic during any photographic work when recording archaeology, as they can hide details which would be visible in a more over-cast environment. Moreover, they can give an inaccurate perspective of a structure or trench when photographed at certain angles. This is often an unavoidable problem when recording on an excavation. However, this is only one issue caused by shadows when creating 3D models. Shadows can also potentially confuse the 3D modelling software when taking multiple images. If the sun is hiding behind clouds when I start the photogrammetry process, and then comes out halfway through, I have a model where half the model has shadows and the other half is overcast. The modelling software may then have trouble identifying the area as one structure and model.
Another issue can be the rain, if it had been raining and there are small puddles in the area to be recorded, the reflective surface of the water can also cause problems for the software. These issues can be overcome by either removing the puddles with sponges or removed from the model digitally, however this is far from ideal. Finally, if it is very windy and there are objects in the structure that are moving about, the software will again be confused by the movement and won’t be able to align the photographs in order to create a point cloud, or at the least will create a blurry, poor quality model. I’ll describe this process of point cloud creation in more detail in another blog, but as far as the process of taking all the photographs is concerned, all these issues are best avoided if possible. The perfect ‘Goldilocks’ conditions for a nice 3D model of a structure such as the wall, are overcast, when it hasn’t been recently raining and it is not too windy, conditions that are can be often hard to capture when working in Orkney, so there are always new skills to be learned in order to work around these issues.
For a view of the 3D model of the wall in structure 1, taken before layers have begun to be removed, see below.
You can find out more about The Ness of Brodgar, along with the work undertaken on the wall in structure 1, in the Ness of Brodgar People, Place and Perception videos, by Simon Gray here: http://www.nessofbrodgar.co.uk/dig-diary/video-diary/