Q: Why do you paint from the live model, rather than from photos?

A: For me, painting from life is a dynamic conversation; a 4-way conversation between me, the picture, the model, and the environment.

The picture I'm painting is continually changing in response to the model (who is always changing, even if imperceptibly) and the light (non-studio natural light, which is always changing).

That keeps it interesting.

Also, I can choose the composition, I can choose the framing, I can choose the coloring, the style, the level of detail, I can move my head and body to get a different perspective, I can ask the model to move (if it doesn't disturb him or her, naturally), I can change the lighting; in other words, it is entirely malleable and new, every time. That makes it dynamic and exciting, like traveling down a real ski slope rather than playing a videogame of a simulated ski slope.

Painting from a photo is similar to painting by numbers; you can get good results, and it can be interesting and informative and a good exercise to learn the tools, but it can look like a copy, and I'm not interested in copying unless it's re-expressing a picture that is on the wall of a room in which a subject of mine is located.

Like this:

   http://www.flickr.com/photos/dnsf/4828194855/

If I were a teacher giving a lesson, I would provide an exercise in which a student paints from an upside-down photo.  The results will be better than painting from a right-side-up painting.

Why? Because our minds tend to tokenize things we see. We distill the rich visual detail into a kind of cartoon that is memorable; that's why they have police lineups to identify criminals.  It's not just "a man in blue wearing a red hat" (that's memorable), but the actual shape of the man's face; the type of hat, the actual material of the blue coat, his skin tone, etc. These can be forgotten and replaced in the mind by the icon/token/cartoon that is easier to remember.

Scott McCloud's Understanding Comics book explains this:

   http://blog.visualmotive.com/wp-content/uploads/2009/12/mccloud-uc-triangle.jpg

To paint a nose from a photo normally results in a person painting the idea of  a nose.

To paint a nose from an upside-down photo results in a person painting the colors and shapes and forms and textures that are seen in the photo; then when you turn the painted picture back around (right-side-up); you will find that you have painted something that looks like the nose in the photo; not your idea of a nose.

Painting from photos is not interesting to me as art because it is like watching a play; the words have already been written, and we simply experience a re-interpretation of the play.

As an exercise, it's good, and even essential -- as an exercise.

On the other hand, painting from life is a dynamic conversation, and the outcome cannot be predicted with certainty.

And it's great fun.