Were you initially, in movie terms, the "line producer?"

Yes, I was, but I had a very strong sense of that show's textural qualities. I became increasingly involved in all the creative aspects of it, including the scripts, and even wrote one or two. I was very involved in the rewrites -- in the sense that I would say to Steven, "If you want a scene to look this way on film, this is how to write it. The words are great on the page, but this is what you have to do to make them into great film."

At the time of their premieres, "Hill Street Blues" and "NYPD Blue" were considered iconoclastic in their treatment of police departments. But they really weren't debunking police -- they just made a viewer go through more to respect these cops, so you had an honest respect for them instead of just a reflex.

I think that's true. And of course the credit goes first to Steven and Michael and in "NYPD Blue" to David Milch and Steven. We all had great regard for those professions. But the shows weren't about "cops," they were about people who happened to be cops. Anymore than "L.A. Law" was about "lawyers," but was about people who happened to be lawyers. The idea was you couldn't help but be involved with people whose lives informed what they were doing -- people who informed their work with who they were. There was always a circular aspect to it that made them richer and more honest as characters.

"L.A. Law" was a more conservative show than either of the cop series, at least visually.

I think that content determines form. "L.A. Law" was much more conventional, not so much in the storytelling but in the way it felt on the page, in its words. I knew it was as much of an uptown show as "Hill Street" had been a downtown show. It was more classical -- elegant, brainy -- and the people it was about had more money. I knew I would shoot it in a much more normal way. When it came to "NYPD Blue," my first response was that I didn't want to direct it -- I didn't want to do "Son of 'Hill Street.'" But when Steven asked what I would do if I could bring myself to do it, I said this is the kind of visual style, and music and editorial style that I would give to it. And he said OK.

I knew it would rattle him! Steven is not as prepared to go off the deep end as I am. We banged heads a lot over the style, but he honored his agreement with me. He didn't like the camera zooming around but he agreed and he stayed by it. The technique had a lot to do with the pulse of the piece. If there was a relaxed scene, the camera might relax, but if there was a scene filled with anger and tension and fear the camera would reflect that.

Didn't "Homicide" do a lot of that?

That's always a point of a little bit of conflict between ["Homicide" creator] Tom Fontana and me. We had done it; "Homicide" got out on the air before us. They were really handheld. They did all kinds of ballsy, in-your-face things, which I applaud. I found it to be, however, their own worst enemy. In the beginning it was so jumpy, so self-conscious, so much so that you stopped listening to the story and the words because you were dazzled by the footwork.

I had lit on the style after seeing some commercials, particularly by a really famous and extremely well-regarded commercials director named Leslie Dektor. Dektor was doing the 501 commercials, the AT&T spots, and that camera was moving, and roving. And Dektor gave it a kind of elegance: The movements were so subtle, you didn't know what was happening with the camera unless you were really paying attention. I showed Steven some of Leslie's work. I said that there was a way to tell hour-long stories using this technique if it's not abused, if there's some sort of a philosophical and visual reason for it. If everything is motivated, and if it's guided by a principle of energy -- whether it's an eye flick or a hand movement or burst of physical or verbal energy, it all comes from somewhere. And Steven grudgingly allowed it and I, with unabashed enthusiasm, did it. It really did affect a lot of things.

In your features -- and "Frequency" is a good example of that -- you use a lot of your television techniques and show they can carry a two-hour movie. There's the way you layer in information, for example, so that you get an inkling of a theme or an incident from various sources before you get a big dramatic scene about it.

That goes back to the storytelling techniques and methods that Michael and Steven invented on "Hill Street Blues." Their notion was that "exposition" was the bad word; anything that smells of "show and tell" was immediately and viciously dispensed with. Instead we had bits of information, either visual or auditory, added together to create a whole that was greater than the sum of its parts. If you're really watching "Hill Street Blues" or "NYPD Blue" or "L.A. Law," or hopefully the movies that I've made, you're never getting it all in one piece -- you're getting hints, you're getting pieces of the puzzle as it comes along, and all off a sudden it becomes clear.

And in "Frequency" it's a mind-bender -- a very difficult journey if you're not paying attention. If you blink and get up and go to the bathroom, you can miss a piece of information that really ties it all together. You get a little bit of information off the AM radio at the beginning of the movie when the truck's going over the bridge -- from that, you heard about the nurses being murdered, New York Mets baseball, solar explosions, a number of things that all begin to add up as you move through the movie. But whether it's that, or what's on television or newspaper headlines or in dialogue -- it was all laid out rather carefully at different places to add up to a whole. And I find that fun to do, like a puzzle piece.

The filmmaking in "Frequency" is efficient; to my eye, you don't get as baroque as the fires in, say, "Backdraft." Did that efficiency come from TV, too?

Certainly in TV you are constricted severely by time and money; the upside is you learn to prepare, you have no choice, you get it done for a certain amount of time and money or you turn into a pumpkin. But the fact of the matter is, as I was describing the cops on those TV shows, this is not a story about a fireman, but about Frank Sullivan (Quaid), a man who happens to be a fireman. I had to rein myself in not to make the fires major events and just go bullshit with all the toys, all the bells and whistles, at my disposal. It was not about, say, Frank fighting a fire in a tunnel, but about him getting out; it was not about the fire in that big warehouse, but about him trying to get out with that girl. So -- less is more. There's kind of an adage in stunt work that would be well applied in any kind of film work, which is "get in late, get out early." That way you don't see the ragged beginning or the sloppy ending of it -- you only see the perfect stunt, and you shoot it with that in mind. And the same applies to a dialogue scene or a story sequence.

You first read the script to "Frequency" in November '97; what accounted for the strength of your response, and why did it take so long to launch?

My dad had died a year and a half before I read it and that was still pretty fresh in my mind. The notion of mixing wildly disparate genres and trying to make it work -- that was something I wanted to try again, because I had tried to do that in "Fallen" and it hadn't worked that well. I wanted to prove to myself that I could get that to happen. The bells and whistles were fun, the time travel was fun, but the heart of it was that relationship between father and son -- these two men who at a freakish moment in time are able to say something to each other that they would never say if they did not understand fully the fragility of the moment.

The movie was considered by the studio to be high-risk because it was such a combination of things -- they weren't sure how to sell it. It took a while to get a budget that was even remotely real. I knew going in, as did my producer, that this was an expensive proposition, and that we had no business making it for the money or for the budget that we had. But we were going to try.

The project had been adrift when I came aboard. It had a huge budget under other directors -- under one director of note, almost twice the budget I had. And they were not getting the cast to support that budget. I was interested in all kinds of principals for the leading roles, but I discovered that movie-star stars weren't interested in sharing the stage 50-50 -- and this movie is a two-hander. Frank's no bigger a role than John Sullivan (Caviezel). Some actors did not want to get close to the father-son stuff: It was too painful for them, or they were not ready to go there. At one point, I had a cast, but was not as happy with it as I had wanted to be, and one of them bolted.

In the script, the father was 30, the son 36. One of the problems I was having was finding 30-year-old male actors who had the kind of life experience or emotional weight that would make the scenes work. I realized I really needed a much more travel-weary, world-weary, experienced actor and human being to play the father.

So casting Quaid must have been key.

I had been very fond of Dennis Quaid's work for years and when I saw him in "The Parent Trap," I thought, this is a grown-up with a kid! You could see him dealing with fatherhood pretty honestly. I also knew he had a 7- or 8-year-old son, and that his marriage was really solid and terribly important to him. I said to myself, "There's the All-American guy." I also knew he would look good in a fireman's outfit -- Dennis has got size. Put that coat and that hat on Al Pacino and it would be a laugh. On Dennis it fits fine.

Now playing: Read all the recent movie reviews by Salon's critics