03 August 2007
Strathmore drawing, 14" x 10". 20 minutes.
this is another life drawing group portrait. the model had strong features that i found more interesting than her sitting pose, and i had good vantage and lighting for the work.
in most learning and exploration, there are two kinds of questions. there are the questions that lead to specific answers (that can be answered), and there are questions that clarify assumptions (that lead to puzzles).
i have four puzzle questions about portraits:
1. how does a portrait look "wrong"? if you attend an art show of portraits, you will find yourself recognizing that some portraits are clumsy, or distorted, or mismanaged somehow -- as a representation of the person. how is that possible, if the person who is painted is unknown to you?
2. if a reasonably accurate, carefully drawn portrait can look "wrong", why does a caricature (a political cartoon, or a puppet) look hilariously "right"? if we see them under different expectations or frames of reference, how do we know which one to use? how many portrait frames of reference are there? can a painting create its own frame of reference, its own expectations?
3. why does a photo, which is always optically correct, look sometimes "wrong"? (we say, "oh, that's a bad picture of her" or "oh, her nose looks too big". the camera just recorded what came through the lens. so what went amiss?)
4. how does a portrait convey "inner state"? how do we infer the character or attitude or feelings of the model when the portrait was painted? what would a portrait that conveyed no emotion look like? like a life mask?
i looked into the literature on facial recognition software. some of that stuff is top secret now, but the stuff i found was not helpful. the face divides into characteristic distances: the distance between the two eyes; between the eyes and the mouth; between the nose and mouth; the width of the mouth; the height of the forehead, and so on.
Posted by Bruce at 8:59 PM