The Jailhouse/Factory interior set.
I knew this one would be a killer. In the storyboard, I've put a lots of hanging lights in the big, wide space and I've copied the concept into the 3D environment. I've added a fenced gallery, but there were no walls in the animatic set, so I knew I'll have to improvise. The idea was to use the Courtroom set and overpaint it in greenish-grey, to turn the pillars upside-down (I've been prepared for that), or turn everything upside-down and get away with it. It's quite possible it'd work.
But, something happened in the meantime. I've reshaped the exterior,
and I had pretty precise drawing of the Jail's/Factory's exterior. To the point that I even made the model in the pre-production. So, let's take a look at the shot preceeding the shot Nr.101 (what an ominous number, it's like it was doomed from the start): there's the jump-cut at the end of the shot Nr.100, to the barred window (so we know it's a jail). Now, that exterior doesn't cut to the interior wide-shot directly, but I wanted to match both views at least to a point of some resemblance. And that's where the troubles have begun. Moreover, I wanted the interior to match some reference photos I collected, and that was guite an ambitious task. However, the chain of events that made everything so time-consuming, started with the fact that Kristijan did something I specifically told him not to - he made the camera moving, pulling back through the rows of lamps, over the rows with elfs. Oh, that was spectacular! I knew moving the camera in such a composite shot (and this one is a textbook definition of a composite shot, one ot he few really unavoidable ones in this picture), that's why I've tried to avoid it, but when I saw what Kristijan made, I couldn't let it go, no matter what!
And that's how it all began.
I started to rise the set pieces, but there was hardly anything to work with. I've abandoned the topsy-turvy grey Courtroom concept as soon as I became aware of the extent of changes needed to match the exterior in any degree. The Courtroom pillars were the only remaining detail from that concept, and they looked rediculously small for the shot. I made some conceptual sketches, and nevertheless incorporated the said pillars into the equation. I did the background unbelievably cheap and simple. I couldn't avoid making the windows, though, but also in the cheapest and fastest way imaginable. That whole backwall perhaps weighted no more than four or five pounds. But, that's only the far wall. So, what to put around it? It's not like I can fill the space with so much smoke that I'd need nothing on the far side. Even with the smoke added in post, I needed something to fill that wast space. Also, there should be that gallery for guards, all around...
O.K. Let's put the gallery on... For that matter, I had two unused perforated wooden beams from earlier commission, in metal-like finish, with rivets. Those seemed ideal to use for the sides. And it was so, because beams were rigid, sturdy enough to support all kinds of additions. But at the same time, the beams were so heavy I could't attach or lean neither of them to the background piece. That thing hardly supported itself, was so soft and fragile that the beam would go right through, if attached.
Therefore, I've placed the beams on their own supports. And all of the sudden, the studio was occupied to the fullest. That left almost no space for the lighting, or for me to squezze through "backstage", for changing bulbs, etc.
Time to shoot. I wasn't sure upon lighting, not at first, so I had to try different things. At the same time I've been testing my improvised "ride" for the camera. I had to come up with something, 'cause my usual stop-mo machine has been about seven feet too short! I had to construct an improvisation, so the new slider was a hand-operated, mini bench with stop-mo machine sitting on it, both sliding on plastic rails. With hand-drawn measure stops. It was rather primitive, but functional enough. The downside was it filled the studio to the extent that when having the camera on it's end position, I could hardly open the studio door.
And I started to shoot one camera pass after another. But, I've been the one operating the stop-mo machine and the slider manually, so I had to walk my way through the cramped space for every frame, every 240 of small camera movements. In theory, I could be able to make as much as two or three camera passes a day. That was a happy thought.
But...
There were some little adjustments to the set, or to the lighting, or I had to repeat the pass, or whatever, every time. After just one pass, I was a wreck, incapable to endure another, not without serious danger of hitting, or bumping to something in the process, which would instantly put all the efforts, all the passes shot to štarting point. So, with every pass more that was shot, the stakes get higher. It was nerve-racking! All in all, I did 75 passes. Simple mathematics suggests it can be done in no less than 75 days, maybe a few more. Right?
Well, mathematics is good, but...
As much as 75 days can sound for just one ten-second shot, the truth is, nevertheless, grimmer. Much grimmer.
How so?
First couple of passes were a bit short. I've found a way to make them a bit longer, to achieve maximal close-up at the beggining, to the widest frame composition possible at the end. That was of the utmost importance, so I had to repeat the passes with the new "settings". Ok, that's reasonable. Then, I still wasn't sure about the lighting, so I made a couple of passes for correction. I started to add some elements to the set. New passes. Every group of elements needs another pass. It's of no use to shoot them in single pass, because I needed them in layers, anyway.
Even if it wasn't like that, it wouldn't be possible, because I made limited number of elements (time and money factor), enough for just one layer at the time. I built them so I can use them multiple times. Ok, that's reasonable. But, to work as predicted, each group of elements needed chroma key additional pass. And that was a bitch to prepare, putting any kind of sheet behind any part of the set, without destroying everything. If I had merely slightly hit some part of the set, or the camera bench/slider, it would reset to start. Nerve-racking!
So, I've shot the multiple passes for number of set elements, the ones that needed to be shot regularly for colours and with chroma-key screen behind them so they can be transformed in layers at a later time.
The gallery, the roof trusses, the bridge crane, the supporting pillars, parts of the floor, and finally, the hanging lamps, in three layers. And that was only the set, a mere background. Static objects. By then, 75 days became a long lost dream. And I didn't even started the real thing: animating the puppets.
The only good news was I could have finally removed all the additions to the back wall, to gain some space, when I was done with the set.
So, that'll be several more days for the puppets, right?
Right...
First, there was the matter of preparing the elfs. I've learned by then: latex wouldn't last long on prepared puppet, clay will harden, too. The best course of action is - to prepare the puppets fresh before shooting.
Ok, that's reasonable.
So I prepared three puppets and six work positions for the first row.
It would be too cramped to have all six of them together, and to animate them, with and without greenscreen, and all in 25 fps*, right? So I did the pass. In 25 fps. 240 frames. That took days.
Then I had to reset the camera, place another three elfs on unused work positions (every other), and do it all over again. Blast! More days have gone. Maybe it'd be better to do them (the elfs) in one pass? Second row had eight posts. Not enough bodies, not enough heads. Recycle? Produce?
News was, I needed another doll, and the package for it. Designs.
Did it in one pass. Eight puppets with machines, one doll, 25 fps. Nerve-racking.
Finally, the third and the last row of elfs - according to the animatic.
Ok. But I've already used all the heads, some more than once. They need some changes. And the machines will look the same. They too need some changes. New doll needed, new box. Done. Animated. Another eight puppets with machines, third doll, 25 fps. Nerve-racking. Stakes were dramatically high by then. Everything could still reset to start. Wasn't over, there was still Santa.
But...
The three rows didn't seem enough... Oh, come on, yet another row?
Another eight elfs, fourth doll, plus box... I must have been jinxed!?
So, I did it. All. Was it enough?
Almost...
I still had to mask the bottom part of the elf's machines. Improvisations, improvisations.
Partial pass. With simple animation, with chroma key.
And there it was, I was done with elfs, finally.
There was only one thing left to be done, and that one was crucal: it was Santa, and he alone was more important than all the elfs and lamps and backgrounds together! And if I mess this one up, it can still send me back to start. Nerve-racking!
Wouldn't it be better if I shot Santa first, and add things around later, you may ask? Affraid not. I've been aware of his position all the time, but even though Santa had to be in the central position here, I had to deal with building the set first, and then stripping it after. That was the way to go. Ok, and how about the elfs? Couldn't
I, at least, shoot them at the end? Well, not really. I had to do lots of carpentry to place the elfs in place. And that was risky. Animating Santa shouldn't be that risky, but should've been the most demanding thing in this shot, instead. But to do that, I had to train all the other things perfectly, like moving the camera, or moving through the studio. So I had the "learning curve": animating the camera only in first passes, then animating less important elfs together with the camera. That way, I was rather trained to animate that demanding puppet and to even make him walk, along with moving the camera. Fine.
So, that must be over, then? Hell, no.
I still had passes for the lamp on Santa's desk, and one pass for backgrounds that should've been digitally placed to the sides, behind supporting pillars, and hardly visible behind layers of digitally added smoke effects.
But, wait! What about the guards? There was still the gallery pass I'd have to do. That's only for the last 50 frames, and in front of a green screen, but I had to set it first. It was only on the third try, that it looked good enough.
Total score: approximately a year. Imagine if I had to reset everything, somewhere near the end? Sounds impossible? Not really. I've actually managed to bump into the camera's workbench just after I took the last frame. How about that?
That was the ordeal with the shot Nr.101.
Is it over?
Mmmm,... no.
That was only the second part of the shot. There's still a prequell of a sort - similar view, similar composition, starting position of the camera movement, if you like. But, the risky part is over. I could reposition the camera. There was that screen I had to come up with, a screen with the red suit painting around Santa's face. I had to be able to animate it's exit from the shot, as well as one small Asian in the black tux. I've tried to be as faithfull as possible to the template Kristijan prepared.
And, finally, was it over?
No?
No. There was still the shot Nr.102: the elfs in the sideways camera move, in a way a repeated shot from the North pole Workshop.
Oh, my... that one was like nine shots put together. But it was a school. There was no need for repeating backgrounds, this time. This one background would do. And I've been trained animating elfs four times prior to this one. I couldn't miss. So, that worked fine.
I was done with the Jailhouse-slash-factory. It took over the year!
Sevety-five days sounds like a dream, compared to that nightmare.
When I reflect on that now, I wonder how it was even possible. It was really like a bad dream...
But, was it over with the factory set alltogether? Not exactly.
I still had to make automated production line for dolls.
That could've been shot on any background, but this one was ideal and already there, so I used it. I had most of the "mechanical" parts ready, but I still needed to improvise some elements. More about machines in the "models" section:

Uz štafelaje: The important part of the shot were designs on these drafting tables, arranged here in array of display stands. I've planned at first to use simple display stands, but that suddenly started to seem somewhat inappropriate for, otherwise very mechanical appearance of this set. More about drafting tables here:

hit counter