In part one of this guide I tried to explain the uses of different coloured backgrounds to achieve good Photogrammetric models of small finds. This however is only one of the things to considered, and hence today I’d like to talk a bit about camera angles.
The way I see it there are two parts to the Photogrammetric process, and these are the photography and the modelling. In order to create good models what you need is not only a good program or experience with how the software works, but Photogrammetry also requires good photographs to process, and the only way you can get exactly what you need is to learn where you should take the photographs from.
I often say that the best position to take photographs is every 45 degrees in a circle around the object, then the same from a different height and a final shot of the top. Generally this works well, but there is a reason for these positions. It allows every point of the surface to be included in at least seven shots, four more than technically required. This way no matter what happens, you will always have enough coverage. In addition to this the photo from to top will nearly without doubt be placed, and it can be used to manually stitch the others together.
But let’s break it down even more. The ideal positions are described work perfectly if what you are photographing is a half-sphere, with the base placed on a surface, decorated with many different colours to allow the program to recognise points. The problem is that most archaeological objects are somewhat different from this description. A long object such as a sword reacts differently to Photogrammetry than a tall bulging object such as a pot. Because of this the methodology I described has to be adapted to fit these differences.
The sword will require some shots to be taken at a larger distance, or to have the long end photographed in two parts (not such a fan of this, but it seems to work). If it is engraved some close up shots of the decoration can help emphasise this, or if it is too shiny the camera may have to be moved to make the light more consistent. The pot instead can work easily with the consistent distance from the camera, but requires an additional circling around the top from a different height to process the inside. Handles may need extra photos to make sure all points are modelled, and a much lower circle may be necessary if the bulging is covering the bottom of it.
Similarly the lower spiral may need to be very low for a bowl, but in the case of the coin it can be as high as necessary, as the side is not as important.
Hence the main thing is to look at each object individually, realise the needs of the program and plan how to tackle any problems it may encounter. Realising that a certain area of a find could be a future issue before the photographs are taken means that the issue can be avoided before it even appears.