The Line Is Always the Line

In 1878, a man named Eadweard Muybridge lined up a row of cameras along a track in Palo Alto and rigged the shutters to trip wires. A horse named Sallie Gardner ran past at about forty miles an hour, snapping the threads as she went, each shutter open for roughly a thousandth of a second.

The pictures settled an argument nobody could settle with their eyes. All four hooves leave the ground. Not stretched out flying, the way every painter had drawn it for centuries. Gathered underneath. (Smarthistory)

Every painting of a running horse before that moment was wrong.

Teresa put that on the screen in an Artist Salon and asked the room what the camera setting had to be. High shutter. One one-thousandth. Stop the motion or show the motion, pick one, and know which one you're picking.

That's the whole reason she teaches this to people making AI images.

Her Artist Salon is not a prompting class. It's like a photography school, run by somebody who went to art school in San Francisco and shot film in Amsterdam and danced on stage with the bands she was photographing. She walks through the exposure triangle. Aperture, shutter, ISO. She explains what film grain actually was and why you might still want it. She explains that the standard lens sits around fifty millimeters because that's roughly how a human sees, and that's more or less where AI will park itself unless you tell it otherwise.

Then she says the useful part: you can just tell it otherwise. Ask for twenty-four millimeters and get more world in the frame. Ask for a shallow depth of field and let the background fall away. Ask for the golden hour and know that the morning one runs bluer and sharper than the evening one, which runs gold. These aren't magic words. They're camera settings that have meant something specific for a hundred and fifty years.

She put up a portrait of Sarah Bernhardt made in 1864 and asked where your eye enters the frame. Somebody said the forehead, because it's bright. She agreed and then showed her own path through it, down the nose, into the contrast, across the fabric, back up. She pointed at the blacks and said there's information in every one of them, even the hair, you can still read curls. She pointed at the whites and said nothing is hot. Hot means so much light hit the negative that nothing got recorded. There's nothing there to recover.

Composition, she says, is arrangement. Rule of thirds. Golden ratio. Use your edges, don't be afraid of an arm running off the frame, it makes a better picture. AI wants to plant your subject dead center every single time and you do not have to accept that.

Then she hit a wall, live, in front of everyone. She'd been building an art history resource and she stopped at 1936, because copyright runs about ninety-five years and she'd walked right up to the edge of the public domain. She wasn't going to pretend otherwise or quietly grab the images anyway. She just told us where the wall was.

Her homework was the best assignment we've handed out. Take your phone. Decide in advance how many frames, twelve or twenty-four or thirty-six, like a real roll. Then shoot the whole set without looking at the screen. You know where the lens is. Trust it.

For bonus points, hold the phone upside down so the lens sits at the bottom and you get more sky.

The point isn't the photos. The point is that you'll stop putting everything in the middle. And once you've felt that in your hands, you'll start writing it into your prompts without thinking about it.

Teresa apologizes for repeating herself about the line, session after session. She shouldn't. She's been looking at pictures for a lifetime and she's still finding it. That's not repetition.

That's the practice.

‍