AVOID OVERLAYS WHEN COMPARING FACES

by Joelle Steele

A client asked me to compare two photos of women taken in the 1860s. The photos were of two great-great-grandmothers, and were labeled with different names, but the resemblance between the two women made the client question if they were instead photos of the same woman.

At first glance, it was easy to see how much they looked alike. Many amateurs just size the faces to make them the same and then overlay one photo over the other and claim they are the same woman. But after comparing faces for more than 40 years, I know better than to make a snap decision like that, and I know not to depend on something as unreliable as an overlay. There are so many people who are unrelated and look very much alike in photographs! The only way to know for sure whether or not they are the same person is to measure their faces and compare the measurements.

But, before you can accurately measure any face, you have to first prepare the photos by doing two things: 1) align the pupils horizontally, and 2) size the irises on both photos so that they are the same size. This preparation is absolutely critical to accurate comparisons. It is also the most immediate way to rule out a match before you waste your time measuring.

In this case, preparing the photos ruled out a match. Here’s why: When the pupils were aligned horizontally and the irises were made the same size – in this case half-irises due to the heavy upper lids – it was extremely easy to see that this was not the same woman. Mary, on the left, has eyes that are smaller in proportion to her face. Susannah, on the right, has eyes that are larger in proportion to her face. So, in the end, this makes Mary’s head noticeably larger than Susannah’s.

To measure these faces is unnecessary, but I did anyway for the purpose of this article. The following are not a match: chins, noses, foreheads, hairlines, cheeks, pupillary distances, and jaws.