In Sergei Eisenstein’s October (Ten Days That Shook the World), a film that made a life-changing impression on me, shots are used in a number of ways and with different purposes, and the effect at the time of the movie’s appearance was so dazzling that Eisenstein was catapulted to the forefront of cinema in the 1920s.
At the beginning of October, the statue of the Czar is shown in details; we see shots that break up the statue into discrete parts, almost in a cubist approach. Then the shots immediately begin a cause-and-effect mode, in which ropes are being pulled. Next, the shots are used to illustrate opposites in a sequence of comparisons: rifles compared to scythes. Almost illogical repeated action is then employed, the Czar’s chair falling over, in a sequence of shots, with time repeated. Then there are many close-ups of men singing, then shots of men intercut with shots of war explosions.
My point is merely that once you’ve reduced the possibilities to discrete shots, you are encouraged to then do something interesting with those shots rather than just let them be units of the progressing action unfolding. In October, montage takes the path of using the shot to make comparisons—wide shots cut against very close shots, rich people vs. poor people, and so on. Now that you’re working with unique shots rather than coverage, you can begin to use them in a variety of ways; repeated close-up (CU) shots of people waiting at the railroad station to build suspense, letting the audience think “Why are they waiting?” Then suddenly, medium shots of searchlights and then the title card “It’s him.” And this is paid off with a dynamic shot of Lenin and the red flag. Now just shots of flags and banners, carefully art-directed: “Down with the Proletariat” intercut with “Down with the Provisional Government.” The breakup of imagery into dynamic shots provides the ability to range through the multitude of possible ways to put them together. Very fast cutting, machine gun with rich lady; you almost think you hear the rattle of the gun purely through this illusion. The famous bridge scene of the film uses montage to expand time, opening in degrees. It’s not merely covering the opening of the bridge: the shots are controlling the opening of the bridge in degrees while you keep your eye on the dead horse. Your emotions are stimulated as you know eventually (in time) the horse will drop off the bridge, which it does in a beautiful moment.
The point is that once you break the action down into discrete and well-composed shots, you then have the ability to recombine them into a multitude of different patterns—enabling many different emotional reactions and responses far beyond what you could do if you merely broke a scene into its components and followed the obligatory rules of coverage. If you break water down to some atoms of oxygen and hydrogen, all you can make out of them is water and maybe hydrogen peroxide—but if you can break those down into protons, electrons, and neutrons, you can make anything!
Thinking in terms of shots and their participation in montage is liberating:
In the sequence where Kerensky (head of the provisional government) waits, we cut to a statue of Napoleon and crystal glasses and decanters. Then Kerensky crosses his arms. A crown cuts to a whistle (you can hear the screeches even though there is no sound). Then: Church, Icons, God, GOD. Laughing Buddha, ancient, bird of prey, monster, primitive, models, epaulets, return of the Czar statue—reconstructed. Eisenstein is using a cinematic language to give his opinion of Kerensky. He cuts to a crown, Napoleon, Kerensky, Napoleon looking at himself. Shots of tractors, gears of tanks, more tanks, whistle, smoke, locomotives, train, newfangled armaments. Comparison of Kerensky’s books against the counterforce of rifles and flags. Cinema is a new language in which shots are like words or sentences and can be organized, repeated, and compared at will.
10
THE STUMBLE-THROUGHS, TECHNICAL REHEARSALS, AND DRESS REHEARSALS
There comes a point when the well-rehearsed and energized cast of actors begins to stumble through entire acts, and soon through groups of acts, and, as soon as possible, the entire show. The process is very similar to what happens in theater, and the reason why I’ve chosen to have a theatrical stage manager calling the show. It is the stage manager’s voice that warns of minutes until going live, gives the countdowns and the lighting cues, uses signal lights to deal with actors’ entrances, and cues music and all the elements that will make up the show.
However, Live Cinema requires cameras, so it is the associate director and not the stage manager who is cueing them as to their positions, talking to them on the all-important intercom, and always one scene ahead, so that the cameras will be ready. There are any number of complexities when one is using a very large number of cameras (we had forty at UCLA). The associate director helps each camera operator by providing a card to mount on his or her camera with a list of that particular camera’s shots. That way operators immediately know what the next position is that they will go to after their camera is finished for a given scene.
Television cameras of the type used in live television shows, as in sports and award shows, each have a device known as a tally light. I’ve never used professional TV cameras for reasons of cost, size, dependence on zoom lenses, and the fact that they are 30p (the standard frame rate for U.S. television). However, a tally light would have been a great convenience at UCLA. The tally light merely indicates on the camera which camera is currently online or being used. If your tally light is on, you as well as the actors and crew know that this camera is online and shouldn’t be moved or readjusted, and the actors need to continue in character until the light goes off. I look forward to having the benefit of tally lights in the future.
In the first stumble-through and subsequent run-throughs, it is often unlikely, or even impossible, that the actors will be able to get from scene to scene in a timely manner. There are many tricks that have come to us from the Golden Age of Live Television, and they’re all fair game. For example, when, during a scene, the shot is cutting between the actors speaking to one another, often (on Playhouse 90, etc.) the shot will remain on the final actor talking to his counterpart; meanwhile, the counterpart has run away and is frantically having her wardrobe changed, a robe or jacket added or removed while the previous scene is still going on, and as she’s allegedly being talked to. As a result, she is ready to be there for the beginning of the next scene, which might even have further padding by beginning with a new character talking to himself essentially before she arrives.
There are other techniques for smooth transitions between scenes:
•One can use the EVS replay server to get the first few shots of the oncoming scene, and play them out, giving the transition more time to be accomplished.
•The entire oncoming scene can be pre-recorded on EVS servers so you are hopscotching from live to EVS to get through a difficult sequence due to extensive wardrobe changes or such.
•The new scene can begin on the back of an alternate cast member dressed to double the incoming cast member, enabling a transition in time.
It was my belief that sports’ extremely reliable and instantaneous EVS replay servers could be used to great advantage in Live Cinema, and so I tested some difficult situations in which they could be used. One was in the need for many extras, including babies, children, and animals, which can add a lot of expense and require a great deal of supervision. The many dressed extras needed in a period scene such as the one we explored at UCLA would require teams of wardrobe, hair, and makeup people; animals require handlers, and children are the domain of the state, whose representatives supervise their well-being, work hours, safety, and education. Babies may only work for 20 minutes a day (if you have twin babies, you can thus eke out 40 minutes), and only within specific times of the morning and afternoon. I wanted to see how I could manage period scenes that had many extras, including children, a baby, a dog, a cat, and a goat (in my story, the goat was a gift from an uncle in the Bronx, intended to be eaten at a baptism party, brought by two brothers on the subway). My question to myself was this: could I have one day with all those elements, and the accompanying hair, makeup, and war
drobe teams as well as the state supervisor of children and the lunches for large groups—only one day, so I wouldn’t be required to repeat all that for each day beginning with the stumble-throughs and technical and dress rehearsals? Could the football-type EVS machine make this possible?
What I learned was yes, one can intercut a live scene with the replay servers at will. At first it was difficult and tricky for my associate director and technical director to do this, but soon they were doing it fairly easily. On the multiview screen before me, I saw the many shots from the live cameras, and below those, the shots of the replay series. As I explained before, the live shots were numbered (CAM 1, CAM 2, CAM 3) and the EVS units displayed were (A)lpha, (B)aker, (C)harlie, etc. I found I could indeed cut whatever live scene was going on with the shots recorded on the EVS from that day when I had all those extras, animals, and kids. In fact, I did one scene with the big group, which was an improvisation of all those people wandering around the rooms, eating, playing with the kids and the goat—and even that was possible to use cuts from, giving the scene about to go on live the sense it was happening in the midst of all that crowd.
I am still evaluating the result of these effects, but I’d have to say that intercutting live cameras with playback servers appears very successful. This means that one wouldn’t have to have all that assembled expense on hand every day of the stumble-throughs, tech rehearsals, and dress rehearsals. One wouldn’t even need them for the final performance.
The one stunt we did at UCLA involved a boy being pulled by a falling radio antenna and falling from the roof. The story depicts two brothers who are early ham radio operators and have installed a kit that enables an early form of television. For a while they have it working, but there are loose wires to the antenna on the roof, and the older brother goes up to fix it. Ultimately this leads to a tragedy in which he falls through the hanging clotheslines strung between the tenements, and to his death. It could have been recorded on an EVS channel, but I thought that the fact that the cast and crew knew we were going to do the stunt live at the final performance would excite everyone, myself included, and so we did it that way. As with the baby and goat, having a live stunt like this was a way to ask this workshop to discover what it could and couldn’t do.
11
MARKS AND OTHER SMALLER UNRESOLVED PROBLEMS
The two experimental proof-of-concept workshops I conducted, at Oklahoma City Community College and at UCLA, were based on different samples of my script, and had different objectives in terms of exactly what I wanted to learn.
In Oklahoma I wanted to learn about cinema-style lighting: could it be placed entirely on the floor, and how could we use the new LED battery-powered lights in this scheme? I also wanted to learn if we could even do a complete performance of a roughly fifty-page script, and what was necessary to build shots as one does in film production. At UCLA I wanted to learn how modular settings (Gordon Craig panels) would function, and also what uses we could apply to the sports-style EVS replay servers. To what extent would a fully programmable mixing board like the DYVI switcher facilitate the difficult juggling of so many formats: live cameras, EVS servers, pre-cut sequences, green-screen composites? Could I shoot one day of many dressed extras, animals, and children and intercut live scenes with those elements? Could I have scenes in a foreign language (Neapolitan dialect) with ongoing live subtitles, and could I integrate a live stunt? As a result of these two workshops I came to the conclusions I’ve tried to discuss in this book.
The answer to most of my questions was “yes.”
The biggest problem I encountered in the OCCC workshop was cable management on the floor—sorting out the tangle of so many camera cables. The difficulty of camera placement was simplified by the lack of actual sets. The biggest problem I experienced in the UCLA workshop was camera placement. Frequently, optimal shots were not possible, and I had to settle for side angle and profile shots; true reverse shots were almost impossible. These issues were solved by pre-shooting reverse angles and using them in tandem with the live cameras. Importantly, I learned that it wasn’t daunting to intercut live cameras with EVS servers. And I learned that live stunts were no more difficult than they are in theater.
I discovered the need for an important new crew job: a cinema-style script supervisor who operates the EVS “IP Director” program to be able to quickly locate shots that were recorded earlier, or were from previous technical or dress rehearsals. In addition, I learned that the very important areas of sound recording, Foley work, sound effects, live music, and sound mixing essentially behave as in conventional cinema, with the obvious difference that sound must be synchronized or performed during the live performance.
With the choreography of many actors and camera positions constantly changing, marks—indications of where everything needed to stand at the various points in the story—resulted in a vast collection of multicolored points and arrows on the stage floor. These were obvious and apparent in all but the lowest camera angles. Since much of my story took place during a celebration, I chose to allow confetti to be sprinkled down from the grid, so that it would decorate the floor. While it didn’t hide all the many colored marks on the floor, the confetti served to disguise them and keep the viewer’s eye from singling them out.
The confetti solution may (or may not) have worked in the experimental workshop, but it certainly didn’t solve the question of marks in future Live Cinema productions. I considered a number of other solutions, for example, marks in invisible paint that would fluoresce when under black (ultraviolet) light. I had even found very small UV flashlights that I thought the camera crew might carry, although this would not help the actors, who also needed to see the marks. Finally, I just ignored the problem—and hoped the confetti on the floor would be sufficient to disguise the marks.
At this moment, I’m still not sure how to fix the problem of marks. Perhaps there’s another way of marking positions that would be invisible to the eye—magnetic, textural, or some other means. I feel confident that once I find myself attempting Live Cinema for real, I’ll come up with a solution to eliminate the thousands of colored angles and lines that decorate the floor.
I’m trying to think now what other unresolved problems I found along the way, but I come up with nothing—other than the sheer immensity of the number of pieces one must juggle to do such a production. At one point five days before the broadcast, my technical director (TD) Teri Rozic, who’s among the best there is, told me that in spite of the fact that our program was only about 30 minutes long, she thought I really should have had two TDs and two associate directors (ADs) to handle all the tasks. From Teri herself: “Here is how I can imagine two TDs working on a project like this. There would be a lead or main TD who would be the one to figure out all of the moving parts and how to put them together. They would be the primary operator of the switcher and therefore do the live switching. A second TD would help in the massive amount of programming, primarily the multiviewers.” It is a great feature that the EVS switcher can have multiviewers programmed per scene (stage), but it is a tedious process that requires a great deal of time, time that needs undivided and uninterrupted attention.
Then, during the show, the second TD could be responsible for following along with the show and routing the correct multiviewer layouts at the right times as well as helping to get the next scene (stage) ready. If I had had a second TD at UCLA I would also have had them cue my playbacks. As Teri told me, “It was too much on my plate to manage the multiviewers, cue the numerous EVS clips for roll-in, and keep my brain on the show to do what was meant and not what was said. If I were to do it again I would not have all of the playbacks that were rolled in under my complete control.”
I find myself still thinking about what that could mean—you might divide the performance into two leapfrogging video units, having each TD and AD handle 20 or so minutes, and then switch to another duplicate unit, enabling the first to reprogram and prepare for the next?
I
t reminds me of how old movies were shown, 20 minutes on one projector and then a changeover to a second projector with the next 20 minutes while the first projector was threaded up with the next reel. I doubt that alternating video units will ultimately be necessary, considering the nature of the digital controls and the potential to program the controls to handle vast collections of shots and sounds, but if it is necessary, how ironic that the most modern and sophisticated electronic cinema would need to copy the old changeovers of the classic motion picture projectors. Some things never change.
12
OBSTACLES AND OTHER THOUGHTS ON LIVE CINEMA NO MATTER WHAT THEY MAY BE
I have essentially discussed everything I have on my mind regarding Live Cinema, and everything I’ve learned during the two experimental workshops. Having thought a lot about all these things, I must ask myself, do I still want to do it? It’s a lot of work, a lot of worry, and the result may not have clear advantages over conventional cinema. With the control you have in conventional cinema, perfection is only a matter of your imagination and the financial and production resources you have at your disposal. Aside from saying that at least you have “true performance” at the heart of this new form, the response may be, “So what?”
Live Cinema and Its Techniques Page 8