Our portrait form needs a basic consistency check before we discuss its face shape categories. The first question is whether changing the crop alters the label without changing the face itself. For the initial check, I would use one permitted portrait and keep its original dimensions documented. Each copy would retain the same facial pixels while adding or removing empty space around them. I am considering <a href=https://faceshapecalculator.top>face shape detector</a> as a reference while drafting the questions. The plan is still on paper, with no measured outcomes or confirmed behavior behind it. For each copy, I could note its border dimensions, category and any confidence information provided. What would you record to make a changed category understandable to someone reviewing the test? The acceptance rule should acknowledge ambiguity while still making unexplained changes easy to investigate.