I got started a bit late, but I thought I might share my attempts to create my own digital doppelganger. The reason I chose me as my subject is so I can assemble all the material whenever I discover I need it without scouring the web for resources. Also, once finished, I can provide a complete range of facial expressions for keyframing. Then I can tell everybody that I do all my own stunts.
I simply positioned my first three images around the model so that the mesh corresponds in orthographic to the images at Y+45, Y-0, Y-45, Y-90. From there, I just tried to capture the underlying topology as best I could. I should say “geometry” instead, right? I’m now starting to see the difference in the terms… “physical” topology as opposed to “mesh” topology.
Tell me what I’m doing wrong with this mesh! Please! I don’t want to mess up my own head!
There’s even a bigger advantage if you use your own head as a reference: you can draw little dots on your face, then take photos, so those dots will help you when modeling (although maybe you won’t need this as you also have a 45° portrait )
Your mesh seems to follow closely the topology of your head. Maybe another loop around the mouth could be a good idea - it should help to animate it/add shape keys.
At the third level of multires, I begin to take shape. I took the 45 degree shot and projected it onto the mesh, then did a bit of tinkering with the materials. It’s kind of creepy now, because it doesn’t quite look like me yet, so it’s like looking into a mirror and seeing someone else. eerie.