I'm doing some work for a sports network that requires the formatting of mass headshots into varying templates and sizes. I'm trying to build an action that would allow me to set up the placement of one headshot and the action would take care of the rest.
The problem is that the sources are of varying sizes and at different rotations, making a simple transformation unrepeatable in an action. My solution would involve a layer that's converted to a smart object, cropped, and then scaled up and down from there. However, you can't crop a smart object in Photoshop.
Does anyone know how to either: crop a smart object, or trim a smart object to a uniform size so that transformation functions will scale and translate the exact same way every time?