We can use the same camera nodal point, roll, pitch, and yaw that we used on the T.I.E. and X-wing to provide that viewer frame of reference motion, but because this miniature is in a different scale than that of the X-wing and T.I.E. ships, our linear track, crab, and boom moves will have to be reduced, not only in length, but in speed, as well. The linear moves will have to be programmed from scratch, so in order to make this program, we refer to the "bi-pack" black and white negatives of the X-wing, and T.I.E. ship, that we used earlier to test the composition and speed of the two ships. By studying this presentation of the shot, we can then get an idea of what the background should look like. We program this move into the control system and shoot another black and white negative and process it. We can now tri-pack (triple thickness) the X-wing black and white negative, the T.I.E. ship black and white negative, and the newly generated Death Star surface negative. Looking at this combination in the viewer tells us if the independent moves recorded on each of these pieces of film will sync together in a realistic-looking way. When we have a good-looking move, we then commit it to color negative. The Death Star surface we are using for this shot is too large to be placed in front of the blue screen. In order to make a matte to be used in optical to hold out the stars, we now make yet another digitally controlled move on the Death Star surface. But this time we front light the Death Star and overexpose the color negative a minimum of three stops. This provides us with a matte of the Death Star surface as a white, or clear cell, image against black. After the proper optical steps, the negative of that print will provide the optical element used to hold out the stars in our final composite. We now set aside the normally exposed image of the Death Star surface, and the overexposed image of the Death Star surface for delivery to the laboratory.
Finally, we prepare to photograph the stars for this composite shot on yet another piece of film. The Dykstraflex camera could be used to photograph the stars, but the star photography does not require the complexity of the ship and surface moves. Instead we will take the tape record of the camera nodel point roll, pitch and yaw to the Dykstraflex's sister camera and control system. This second system does not have the versatility of the Dykstraflex but exactly matches the motor speed per degree of angular motion of the Dykstraflex so that, by loading the digital information from the tape of the nodal point angular moves used on the surface, we can photograph the stars with an exactly matching set of angular moves. No track moves are necessary on the stars, as they are at an infinite distance and traveling toward them or away from them will not noticeably change their position in frame. Because no program changes were necessary in any of the axes, the star—move is photographed directly onto color negative and set aside for laboratory delivery.
The processed black and white negative of the X-wing, T.I.E. ship and the Death Star surface, along with the tape of the digital information comprising those moves, are appropriately catalogued for future reference. The black and white negative will be used later for synchronization of all these elements, and the tape is held so that any element can be duplicated again should any harm befall the negative in its future travels, or should another shot require a similar move on a different miniature.
The negative with its latent images now becomes the responsibility of "Control". Control shepherds the film from that stage through all the steps that it must take from this point on to become a final composite. Control now sends the negative to the laboratory, ordering the appropriate prints for later use. In this case, each of the elements will be printed twice, with the exception of the X-wing engine effect, which will have a third, low contrast print made. This print will be used later in optical composite.
All of our prints are made at our specified printer light, and the laboratory graciously consented to run sensitometric color strips generated by us in order to allow control of the blue screen, and subject color. These strips were all shot at the same time and stored under ideal conditions until sent with our exposed negative.
This method allowed us to determine whether or not subtle color variations seen in daily projections were due to slight laboratory fluctuations or to some change in our system, such as aging lights or aberrations in the camera mechanics or electronics. We also ordered a quantity of print stock with a camera 1866 pitch and shape perforation. When contact-printed on a machine used for making internegatives, the print we received could be used for making rotoscope garbage mattes, and in some cases this print would actually be used in the optical printer as a printing element.
The morning following the photography on stage, we select takes in dailies. The individual takes may be different by design. One longer than another or perhaps a filtration variation. Shooting this variety helps expedite the completion of the shot should the pacing or visual nature of the shot require modification later.
The selected takes than appear via "control" at the viewer for synchronization. The viewer which we designed and built for our next step incorporates a pin-registered movement capable of projecting three thicknesses of our special 1866 printer stock, or six thicknesses of the mylar-base black and white negative stock. Because of the negative pitch stock and the pin-registration, this device can give a very accurate indication of matte fit and steadiness, even though the print we are viewing was continuos printed. The only variation that we generally find is a very slight bit of image weave. Using this machine, we now bi-pack the color print of the Death Star surface and its second-pass matte to make sure it fits. Now the color print of the stars is run against the matte of the Death Star to verify that the motions of the stars and surface match appropriately. We now check the X-wing color print against its second-pass engine effect to be certain of its matching up.
The black and white negatives of the X-wing, the T.I.E. ship, and the surface that we generated during programming are now bi-packed with each of their color print counterparts to make sure that they are exactly the same. The reason for this last step is that the synchronization of each of these elements can be varied slightly (two or three frames) from what the original match up was at the time of photography. This slipping of synchronization can improve the feeling of the shot. The color print is a white or clear image of a ship against a black or dark blue background. When bi-packed, it is very difficult to see the ship, which falls behind or in front of the heavy density of the blue or black areas of the print. The black and white negative, with its black ship image on clear cell background, allows you to see all the ships, or surface relative to each other easily.
Once it has been determined that the black and white negative and the color print duplicate each other exactly, we set aside the color print, and run black and white elements in Bi-, Tri-, or Quadra-pack. When the composition, position, and speed of each element relative to each other element seems right, the elements are punched with a holepunch on a common frame. This sync punch is then transferred to the matching six color elements and sent back to control with specific instructions for "Optical", or "Rotoscope" regarding its optical combination.
The next step involves returning the color print elements to control with their sync marks and the instruction card. Control then records the key numbers from each piece of print and sends the prints with their instruction card to the optical department. The optical department attaches their own optical sync mark to color print elements and determines what pieces of film will have to be provided by rotoscope. In this shot, lasers firing from the T.I.E. ship at the X-wing will have to be generated. The lasers will cause reflections on both the T.I.E. ship that fires them and the X-wing which they will pass by. Because of the precision artwork this requires, optical may choose to make a registration print of these two elements. This print will more exactly match the position of the X-wing or the T.I.E. ship on the original negative, than the continuos print. The blue screen elements of this shot will also need garbage mattes. With these—requirements determined, the appropriate elements will be sent to "rotoscope."
First, rotoscope will generate the garbage mattes. The T.I.E. ship and the X-wing are the two elements that require garbage mattes for this particular shot. As you may remember, these two ships were shot against blue backing. To help reduce the blue-light reflections on the miniatures, we covered the majority of the blue screen. This left a small patch of blue around the miniature and a large black area that may have lamp stands or other equipment included in it. This black area with its attendant "garbage" is the reason for these garbage mattes. The rotoscope department will now put the color print of the X-wing in a device called a rotoscope camera which is capable of projection and photography through the same lens with the precision of a camera movement. The print will be projected frame-by-frame onto a surface with animation registration pegs. A cell is placed on this surface and a drawing is made which outlines all of the areas that we will not want seen in the final composite image. These drawings are blacked out in the areas to be eliminated. The drawings are then returned to the rotoscope camera where they are re-photographed off the same surface and animation pegs they were originally drawn on. This gives us a piece of film which will be used in conjunction with the X-wing element in the optical composite step to keep from printing the unwanted portion of the frame. A similar garbage matte is then generated for the T.I.E. ship element.
With the garbage mattes complete, we will now move on to the laser beams and reflections. With the frame-by-frame drawing of each of the two ships, we can now establish their position relative to each other in any frame of the shot.
We discuss the choreography of the lasers and establish their speed and position. We also determine their number.
In this shot the T.I.E. ship pursuing the X-wing fires a volley of lasers, starting on the 12th frame from the sync frame. These lasers miss, but some pass behind the X-wing, and some pass in front of the X-wing. To achieve this effect, two sets of lasers will have to be created—one set that passes over the X-wing and one set that passes behind the X-wing. The two sets are necessary so that when the optical department composites this shot, the set of lasers behind the X-wing can be covered up by the X-wing hold-out matte. Lasers typically last between three, and six frames. In this shot the volley which we mentioned will include seven lasers. These lasers will come from the T.I.E. ship in a rapid fire fashion. The first laser fired will appear on frame 12, and the second on frame 14. The remaining five lasers will appear in two-to-four-frame intervals.
The interval and duration of the lasers provide us with the frames that we will have to be concerned with. In the same fashion that we generated the cells for the garbage mattes, we now generate a new set of frame-by-frame drawings. This set of drawings will include the positions of both the X-wing and the T.I.E. ship on one cell. There is a cell for each of the frames which will include lasers. The first cell contains accurate drawings of the X-wing, and the T.I.E. ship, as they will appear in frame 12 of the final composite.
The artwork for the laser is now created. This cell will have only the beginning of the first laser on it. The second cell contains the ships as they appear in the following frame (frame 13). The move that the ships make between this frame and the last is inspected and the artist creating the second frame of laser must compensate the position of this second frame of the first laser to allow for our apparent camera move.
[ continued on page 4 ]