Continuity

I remember when I was an extra on that Gene Hackman/Barbra Streisand film, Gene Hackman was on set. "Were my hands in my pockets or out?" he asked the script supervisor, wanting to maintain continuity. We've been shooting (and mis-shooting) these certain scenes so long now that the continuity is becoming ingrained in my head. In the first scene, Ben is wearing the maroon shirt. His jacket is open. The scene with him looking and smoking is where he wears the orange shirt. The continuity comes from having shot it, captured it, edited it, like I remember the moment, a different kind of continuity.

Great expectations

I think the previs was extremely useful for developing the narrative. All sorts of narrative problems and gaps came to light. So the pre-vis supplies an overall plan. It supplements the written treatment. it's our script. It sets the stage for performance and listening.

We start shooting on set. But a lot of the first encounters with the set is to find out what wants to be shot because the pre-vis shots don't work. The Ben standing and looking at the spinner just doesn't work because of framing. Others come to mind. Then each scene begins to form and tell us what it wants to be. We then start to name each scene. In a sense, typical previsualization is like naming the scene first and dictating how it has to function. Within our structure, the pre-vis dictates only generally. We listen and then name the scene--the ritualistic scene, the watching the spinner like a TV scene, etc. This is totally different from the shot list which was broken up into days.

Maybe this is what an experienced crew does only much much faster. It takes minutes, whereas for us, it takes weeks.

So setting expectations is important. Rather than assuming we can just go out and shoot something and make it work, you set the expectation that for a few days or weeks you'll be searching. Then the shot will disclose itself. Then you move on.

Running like a well-oiled machine

Saturday night's shoot went well. The spinner works great now. You just turn it on, let it "equalize' (spin smoothly) and it's good. I'm getting a better handle on the lighting. This is where I thought we'd be a few weeks ago--actually shooting and getting some footage we can use. Here's some footage I edited to see how everything cuts together [VIEW]. (Levels in AE, deinterlaced, 80% speed).

Ben sent a note wondering if the set should change to show a progression:

I thought while looking at the images that every film that has a sequence similar to this one has even the "work station", develop in a significant way. For instance when the guy first discovers the formula or the magic or what ever, his work station looks one way, and then as his under standing develops it changes. Elements are replace or removed as his understanding of the phenomenon take shape

Usually this sort of progression is good. But in this case, it may not be needed. This perpetual motion thing is like building a circuit-bent instrument. If it works, don't touch it! You don't know what you might be screwing with. Of course, there's that other question--how do you know if you actually invented a perpetual motion? Scientifically speaking, the only way to tell is if you can verify that the output energy is greater than the energy consumed. But in a home inventor scenario, all you may have is a 'virtual' perpetual motion machine... a machine that runs seemingly forever, but who knows if it does and why? So I like the idea of little changes we had talked about--things on the table move slightly and change--as if there's activity going on, Ben's facial hair changing. That may be all that's necessary.

One thing I'm always interested in is the psychological aspects of working. Shooting at 10 pm as we've done a few times makes a difference. We went from 10 to 12:30-- 2.5 hours--almost nothing. And yet it seems like we were working a long time. Seriously if you look at all the time we've spent fixing the spinner, shooting, etc., it probably doesn't even add up to three or four 8-hour days. So part of the project is playing with the psychology--dealing with the feeling that we're far behind, dealing with the feeling that we've been shooting on the same set for SOOO long. It really hasn't been that long.

Another thing I'm warding off is the feeling of perfectionism. There's a part of me that is concerned we're being overly perfectionistic in trying to get "the perfect shot." But in reality, I think, we're still in the "listening" stage. That is, we're shooting not to get it perfect. We're shooting to figure out what we want to shoot. There's still a searching going on. That's productive I think. It seems like we're settling on getting four or so modes of activity from this set--

1. The iconic, dark, ritualistic, Ben as priest shot.
2. The smoking, it's been doing this forever shot.
3. The top lit more-staring at the spinner shot.
4. The finding the stars shot.

I'm liking the improvisations that are going on. I'm also liking the "edges" of the footage. That has always been the favorite kind of footage for me--the places where both the camera and the performer are improvising, searching and then come together for a brief moment. You'll find an example of that in the sample edit--the close up in which Ben is smoking. BTW, I was reading a book that said the only way to see smoke is to backlight it which is what we did. Somewhere there is a class that teaches these things and apparently I missed it.

Another note: the framing on some of these shots is odd, almost like I'm composing for 4:3 in a 16:9 frame. My original idea was to shoot everything either really wide or really close. Medium shots were only for normal, real-time. Still thinking through that one.

What the fog?

Spent part of the day today trouble-shooting. Focus problem: it's a little embarassing mentioning this but I'm using the Portrait mode setting on the Sony. This setting promises to create a soft, out-of-focus background which it does nicely. But the depth of field is extremely narrow so it's easy for Ben to go out of focus--and, of course, you can't check focus on the mini viewfinder. Geez, what next? Do I need to get Erik to be a focus puller using the remote? The only shots in which Ben is in focus is when he's bending over with his face near the spinner (shown above). I admit I focused on the spinner instead of his face thinking it was close enough. Note that Ben is conveniently obscuring the unwanted wood hammered into the wall.

Also, in the above image (color-corrected in AE/Levels) notice that part of the spinner rig is visible. It is not visible in the viewfinder so the captured image is slightly larger than you can see on the Sony. No problem. Just scale it up a bit, but it does add to rendering time.

Checked back on the original model image shot a few weeks ago, the one shown in miniature at the left. The background in that one is quite dark. There is quite a bit of contrast in that one. I have to not be too afraid of using contrast.

Also reviewed the shots done with the fog machine. Tip for would-be filmmakers: A $30 DJ smoke machine from Guitar Center is probably NOT going to give you the effect you want. The smoke looks more like a puff of steam from a tea kettle and then quickly dissipates upwards.

Lights, camera, "Open file...!"

Last night we shot masters of the spinner scene. The spinner is operating fairly reliably now so we were able to get everything done. Still some problems--Ben's looking slightly out of focus, and a piece of wood that he used to create the spinner rig is visible, nailed into the wall. However, I think there's some usable stuff in there.

I'm a fan of moving quickly and shooting fast. It prevents you from thinking too much. And indeed, that's the way a lot of low/no budget films are shot--you know Brian Foy's adage "you can't see the teeth on a buzzsaw." But in this case, when the shot and lighting are so much of the performance, we can't do that. With our slow-moving, still-like shots, there is no strong narrative or action to pull a viewer into the story. So I think a lot of my production strategies have come from an intuitive recognition that we're doing a certain kind of visual arts-influenced cinema within significant budgetary restraints. Two of these intuitive strategies:

>Buy not rent.
Typically filmmakers rent their equipment. But in general, my tendency has been to buy something less expensive, learn it inside/out, and have it available for shoots and reshoots. The hypothesis is that when the technology is decent, familiarity and availability trump superior technology. There's no way we could have run this many tests using a rental unit (remember the VX100 problems?) That's also why I bought the Lowel lighting kit.

>Shoot on a long-term set
Our major sets are all designed to be available for extended periods of time (the garage has been up for over a year). This enables us to fine-tune the imagery. That wouldn't have been possible with a standard production schedule and locations.

So this is the tension. On the one hand, as I've written earlier, I don't want to create an overly-controlled production design that suffocates any sense of performance. And yet, I seem to be treating the sets as "Photoshop documents" that I can go back to and tweak over long periods of time.

The shot is the performance

I keep going back to this test footage that we shot the other week. I really like it despite its problems. Yes the spinner is a little fast and wobbly. The left side light isn't flagged off. Erik is holding an ordinary non-temperature corrected garden flood light (with a couple of layers of diffusion) to get the shadow on Ben's face. The lighting on Ben's face is flat. (Siwaraya also thinks his face is too red). But as an image it comes together for me.


The final lighting test using Erik as a stand-in just didn't work as well. That got me thinking about my experience doing music. One problem you always hear about pop music is that the demo sounds better than the polished final version. That's because the demo often has an energy that is difficult to recapture. Yes, the final product has fewer mistakes and is seamless, but there is often something lost in the translation.

So that got me thinking that the performance for this video doesn't start and end with Ben; the shot is the performance. And like a performance, you can't dial in settings and expect everything to work. You can't recapture performance; performance has to happen. As Steve Buscemi playing the director in Living in Oblivion says, "we can't think about that. That moment's gone. We'll get another." You have to look and listen and see what you have in front of you. So I wonder if it's less about getting everything set up perfectly and writing down all the camera settings and expecting Ben to walk in and "act." Rather, it's knowing the craft--lighting, the camera, the acting--fluently enough to be able to look and listen while on set and make the proper adjustments such that everything comes together. The set--and its attendant technologies--is the instrument.

Should we have completed the entire scene on that day we shot test footage? Probably not. Testing out some of our other ideas has been fruitful and provides a good reason for doing tests. We were talking about putting a light in the spinner's box and I just assumed that it wouldn't look like much. Sure there would be a little ellipse of pure blown-out white coming out the top of the box but what was the big deal about that? I didn't even try it for the final (Erik) test. But after putting the light in the box later I discovered that light leaks out the sides of the box where the cover meets the bottom. It looks really great and adds to the "magic" that we were talking about.

If the shot is the performance, then we have to think of the tests not as tests so much, but as rehearsals. If you overrehearse, you lose performance value. If you underrehearse, the product looks too raw and unfinished. And directing is finding the optimum point between the two.

Seeing is believing

(If you want this to make sense, be sure to read the previous post first.)
I just watched X: The man with the X-ray eyes. It screens very differently from what I was expecting. It reminds me a lot of the original version of The Fly. A scientist goes too far in his quest for knowledge--with tragic results! Very modernist. The second half of the film with Don Rickles looks and feels like a Twilight Zone episode with its sinister carnies and doctor-on-the-lam-turned-faith healer. Also, I guess I was expecting something that looked like Wasp Woman. Black and White. Stark. But this was a color widescreen picture with pretty good production values. Some stock footage exteriors. Location interiors and a few appropriately motivated flimsy sets. Just some clumsy direction here and there and a clunky, abrupt ending. So all the elements that I was responding to in the film synopsis were left unexplored, buried beneath the surface. These unexplored ideas might make a good idea for a project: someone who could see things from all angles simultaneously in true, objective, Google-like, God-like vision. Like a Hockney photomontage in time. Where X ends, this would begin. Perhaps beneath the heart of every B-movie lies the soul of an experimental film.

What if you found out that your so-called art films were really just B-movies?

I've been reading cult/indy/schlock director Roger Corman's book "How I made a hundred movies in Hollywood and never lost a dime." In it, I found an outline for a script that sounds just like something I would write. I've never felt that way about a screenplay before. Here's the synposis of The Man with the X-Ray Eyes:

Dr. James Xavier injects himself with an experimental serum he has developed to expand human eyesight and he goes insane because he soon sees everything. Obsessed with the God-like possibilities, he pushes a colleague out of a window when he tries to inject X with a tranquilizer. He is now a mad genius on the run from the law.

Lead-reinforced goggles shield him from overpowering light but his life becomes a pathetic, tormented odyssey. Aided by a con artist played by Don Rickles, he works as a sideshow mind reader and healer before winning $20,000 in Vegas so he can develop an antidote. In a tussle with security guards, his goggles fall off and he panics, tossing the money into the air. When patrons riot, he flees to the desert.

Driving toward the desert, he sees through buildings and describes "a city unborn, flesh dissolved in an acid of light, a city of the dead." The car crashes and X staggers, blind into a desert tend revival. The preacher, offended by X's claim to see at the center of the universe the light of God, quotes Matthew: "If thy right eye offend thee, pluck it out." Tortured by voices in his head, he bends down and gouges out his own eyes.

Hmmm. A few months ago, I was reading about how cult/indy/schlock director Lloyd Kaufman was strongly influenced by film artist Stan Brakhage (see troma.com). It might be obvious to some people but I guess I never made the art film/b movie connection before. Think about our bird hitting the windows scene--it's very similar to a scene out of the upcoming remake of The Omen. And the fire at the end... that's straight out of Corman's The Fall of the House of Usher. Now where did I put that bucket of fake blood?

Lighting with a histogram

Yeesh, more lighting tests. I've been developing a technique for lighting using the Sony's on-screen viewerfinder histogram. It immediately triggers my experience in Photoshop since I know how an attenuated histogram decreases the amount of data used in an image. I love its familiarity.

The idea is to move the key closer or farther in order to get a reasonable histogram reading keeping in mind that lighting rule about lighting dropping off (or increasing) expotentially and setting the corresponding exposure. Then you set the other lights. It's really just treating the camera like a light meter.

The image above doesn't have enough light. The bump in the histogram goes too far off the scale on the left. I moved the lights closer to generate a more reasonable histogram after this test. As always, it's the relationship of the lights to one another--the contrast--that counts.

Mission: control

One of the things you always hear about filmmaking is how you can't completely control the process. Making a film is less about imposing a vision and more about setting the stage for things to happen.

I believe this is true.

At the same time, filmmakers from visual arts backgrounds tend to think differently. Artists are used to having control over every aspect of their work. Think about those all-blue screen films like Sky Captain and the World of Tomorrow, or Mirror Mask (both made by people with art backgrounds). Or think about Star Wars I-III made by George Lucas who thinks like an artist. All of these films are heavily composited, and feature sock-puppet performances in which the actors are essentially objectified. All of these films have a kind of deadness that comes from the director's total control over the image, tone and performance. We are not watching performances so much as watching sculpture-in-motion.

My tendency is similar. Consider the fact that I just have one character. This character is placed into rooms/sets that I control to the extent that one set has been standing for over a year (the garage) just so I can continually run tests and fiddle with the lighting/production design/set dressing. At the same time, I wonder if this is why Ben and Erik and I are all feeling so strongly about doing practical (vs. digital effects). I also like very much the fact that while the sets are all transformed, they still start out with their own character. They are dressed but not totally created.

So it will be interesting to see how this all turns out. Will we end up with another one of those artist-made, visually suffocating, performance-challenged films? Or will it like somewhere between a virtual creation and a typical film? The overriding question from the very beginning of this project was "where does the performance exist?" At one time I had the idea that the performance existed in the fact that the film would have no cuts. I still think there's a certain a truth to that. But at the same time, Hitchcock's Rope doesn't seem to have more or less performance value than his other films. The editing, in a sense, was done in rehearsal, and in the actors' training. So performance value seems to be less a matter of method or approach, but of participating in the process of life come into existence. This isn't so much about cuts or blue screens, but knowing when, where and how this sense of liveliness occurs.

Added 3/10
I was reading a book about production design today. A designer said you have to be cautious of making a set too perfect or it can detract from a sense of reality. I also remember reading something like that in Cinefex in which some vis effects supervisors complained that visual effects were often too perfect, taking away from the sense of reality. The example was the elephant squashing the car in Jumanji--the elephant steps too squarely in the middle of the car roof and that never happens in reality.

It would have been easier to invent a REAL perpetual motion machine

On the third iteration of the spinning practical effect, it finally seems to work. Version 1: use of the chopsticks and paper cone is too inaccurate to allow for smooth motion. Version 2: addition of wood rig creates stability but the small size of the circle attached to the drill motor causes too much side to side motion. Plus, the spinning is uneven... spin, spin, spin...slow down. spin, spin, spin... slow down. Where's Dan (physicist) when you need him? Version 3: a charm. Use of a larger circle (a 12" diameter piece of wood) attached to the drill seems to work pretty well. It allows for slow motion and doesn't causes too much sway or uneven spinning. It took us about 16 hours to get the thing working.

This time spent caused us to question how important the whole visual effect was. Our conclusion: it was important. More thought on the spectacle thing. Last week I analyzed the student projects in terms of four factors: impressiveness, spectacle, poetry and authority. Students tend to like videos that are high on impressiveness and spectacle, lower on poetry and authority. Documentaries tend to be low on impressiveness and spectacle, high on authority. For our film, we're going for low on impressivness but high on everything else. That's why it's important to get the spinner working. Film needs spectacle. But spectacle without the impressiveness, and without being MTV Jackass.

Open wide...

This 16 x 9 takes some getting used to. First there was the framing problem solved by going back to the HDV sony. Then there was the monitoring problem. Well, there is no monitor. I don't even know how to hook one up with this setup. So the computer becomes the monitor and after capturing, I realized the spinner was a little wobbly and going too fast. So after more discussion with Ben and Erik, more changes are in order. We're going to try to increase the size of the disk in the spinner rig to see if that gives us more control. And after looking at the test footage, I've been realizing how I am not used to filling up the 16 x 9 screen. My framing, as with the previs is all for 4 X 3. So I'm wondering if I can take advantage of it. Maybe it would be nice to project a spinner "shadow" onto the back wall like a spatial montage? You can see the wobby spinner test here.