Every so often somebody asks why I do not just shoot it clean and add the color afterward. It is a fair question, and the answer is not nostalgia.

A gel changes what the light is before it arrives. Put a deep red on a source and the red does not land evenly: it picks out whatever in a face is already warm and lets the rest fall away. Skin holds it, hair does not, the whites of the eyes go somewhere strange and useful. You cannot get that by tinting a gray file later, because by then the information about what happened to each surface is gone.

The second reason is that it fixes the exposure decision in the room. A saturated gel eats a stop and a half. That forces me to open up or bring the light closer, which changes the fall-off, which changes the picture. The color is not applied to the photograph. It is one of the things that made it.

The third reason is the person standing in it. A room lit deep red is a strange place to be, and people respond to that. They stop performing the face they brought and start doing something else. Half the frames I keep come from the first two minutes after somebody notices the room has changed color.

When I do not gel: anything that has to sit beside other people's photographs on a page, and anything a client will end up printing on a shirt. Then it is one clean source and a lot of patience.

A singer came up from the city in June for a cover and two months of press, and for the first hour I could not make a single frame work.

She had brought a leather jacket and a folder of other people's photographs. I had built a teal wall and a warm edge to match it. The setup was fine. The pictures were dead. She was doing the face from the folder, I was lighting the face from the folder, and between us we were making a competent version of a photograph that already existed.

So I turned the color off. One source, no gels, gray wall, and I told her we were burning the next twenty minutes on nothing in particular. She stopped holding her chin at the angle she had practiced. She started looking at things in the room.

Then I put the teal back and we shot it again in nine minutes. Those are the frames she used. The gray ones from the middle, with only the warm edge left in color, are the ones I kept, and two of them ran alongside the interview.

What I took from it: the setup was never the problem. I had built the room around a photograph instead of around a person, and it took an hour to notice.

People assume a studio full of color is a studio full of equipment. It is usually one light and a lot of things standing in front of it.

Start with a single hard source and nothing else on. Everything a face does is already there: the shadow under the brow, the line down the side of the nose, the place where the jaw stops. Every light you add takes some of that away. A fill light does not add information, it subtracts contrast. Most portrait lighting is a slow negotiation about how much of a face you are willing to give up.

So I work the other way around. One source, and then I spend the time on what sits between it and the person: a flag to cut the spill off the background, a piece of foil punched with a window shape, a square of diffusion held by hand at whatever distance looks right. The frame on this page is a scrap of card with a rectangle cut in it, eighteen inches from the light.

The background is a separate problem and deserves its own source. A person lit red against a wall lit red is a red photograph. A person lit red against a wall lit teal is a photograph about a person.

What this costs is time. A setup like that takes twenty minutes to build and four to shoot. I would rather spend a session on three of them than an afternoon on nine that all look like a rental catalog.