In the second stage of creating this edition we used the programme Agisoft Metashape Pro. Agisoft Metashape is a “software product that performs photogrammetric processing of digital images and generates 3D spatial data” (Agisoft, n.d.). As part of the course Creating Digital Collections we got a number of resources and classes to learn how to use the programme. At first, we put the images we took in two chunks based on whether they were depicting Ken upright or upside down. After that we aligned the pictures of both chunks and created a point cloud for each, after which we created a preliminary model of low quality to have first glance. However, from this model Ken’s legs appeared distorted and riddled with holes and even though we tried used the “filling holes” and “refining model” features, Ken’s legs did not turn out better. In order to find the issue, we examined the masks generated by the Metashape. Here we found that the software misinterpreted Ken's leg dimensions. Therefore, we manually edited the masks of all the pictures with the aim of remaking the model from the new masks. Unfortunately, this did not work, and we realised that we had to take all the pictures again to create a new model.
After the setback of the previous step, we deliberated on enhancing our model's quality by retaking the pictures. We determined that clothing was necessary for Ken to conceal shiny uniform skin colour areas, like his legs and torso, as Metashape struggled to recognise them despite the matting spray we had used. After some looking we eventually found some clothes for him (although he had to borrow some pants from Barbie) and covered his arms in baby powder for our second attempt at the 3D model. We also decided to take picture from four different heights to be sure to cover all parts of his body with our pictures. With this new understanding of how we should approach the pictures we reproduced the steps of the first picture taking process and were able to create a 3D model. This time the model looked much better, but areas such as his arms remained difficult for Metashape to process in chunk 1. However, we were able to take the head of chunk 1, which turned out really well with detailed features such as the hair and face, and the body of chunk 2, which had better arms and the feet, and combined them into our final model. This appeared to be our best option as his legs had slightly moved when we took him off the lazy susan and turned him around. As such, we are happy with how our 3D model turned out, acknowledging the difficulties encountered along the way despite its remaining flaws.