### Camera rig and file formats Images are collected using a DI3D system (www.di4d.com). This is a passive stereo photogrammetry based solution for the creation of accurate, ultra-high resolution, full colour 3D surface images using six standard digital stills cameras (Canon EOS100D; lenses: Canon EF 50 mm f/1.8 STM). The software DI3Dcapture (version 6.8.4) is used to capture participants' faces from six different angles (see figure 1). The frontal image captured by the middle top camera will be used for 2D analyses and stimuli creation. The 3D images are generated using DI3Dview (version 6.8.9), which creates both a texture map in the BMP file format (at a resolution of 1MP minimum) as well as a three-dimensional mesh from the raw data that will be exported in the Wavefront OBJ file format (see figure 2 for import and export settings). ### Capturing setup Participants will be seated at a distance of 90 cm from the rig. The height of the camera rig is adjustable via a remote-controlled lifting column. Two remote-controlled flash units (Elinchrom D-Lite RX 2) are used for lighting. See figure 3 for a schematic drawing of the setup. ### Capturing procedure #### Preparing participants On entering the lab, participants will be asked to wipe their face using a hypoallergenic face wipe to remove any makeup. A minimum of 10 mins will be left after this before pictures are taken. Participants will then be asked to wear a white lab coat that will cover their clothing. They will be instructed to remove any glasses, earrings or other body jewellery (if possible). Participants will be further asked to wear a headband to ensure their face and ears are not obstructed by hair. The headband will be pulled back just far enough for the hairline to be visible. #### Taking the images Participants will be asked to sit upright and look into the top middle camera. For neutral images, participants will be asked to maintain a neutral facial expression, with their back teeth lightly touching, and their lips lightly touching (neither should be clenched). For smiling images, participants will be instructed to smile as they would do naturally. Correct positioning is checked via the capturing software’s live view. Participants will provide separate informed written consent regarding the use of their images (see pdf 1). ### Pre-Analysis Plan Extraneous parts of each scan will be removed using MeshLab (Visual Computing Lab ISTI-CNR, http://meshlab.sourceforge.net) and Blender (Blender Foundation, http://www.blender.org). Faces will be delineated in MorphAnalyser 2.4 (Tiddeman, Duffy, & Rabey, 2000). In a first step, a template with 54 landmarks will be used (see pdf 2). An inter- and intra-rater reliability analysis will determine whether any of the points should be removed or if their definition needs to be clarified. MorphAnalyser will then be used to align the landmark templates for all digitized faces in orientation, rotation and scale using Procrustes superimposition. MorphAnalyser resamples each faces model in accordance with a standard face delineated with the same set of landmarks (Holzleitner et al., 2014, Holzleitner & Perrett, 2015). This process establishes homology of each face scan's tesselations across the entire sample and allows to carry out procedures such as averaging on the whole surface of each face model (instead of being confined to landmark templates). --- ### References Holzleitner, I.J., Hunter, D.W., Tiddeman, B.P., Seck, A., Re, D.E., & Perrett, D.I. (2014). Men's facial masculinity: when body size matters. *Perception, 43*,1191-1202. doi:10.1068/p7673 Holzleitner, I.J. & Perrett, D.I. (2015). Perception of strength from 3D faces is linked to facial cues of physique. *Evolution and Human Behavior, 37*(3), 217-229. doi:10.1016/j.evolhumbehav.2015.11.004 Tiddeman, B.P., Duffy, N., & Rabey, G. (2000). Construction and visualisation of three-dimensional facial statistics. *Computer Methods and Programs in Biomedicine, 63*, 9–20.