COMPUTER ANIMATION AT LAWERENCE LIVERMORE LABORATORY

S.R. Levine Lawrence Livermore Laboratory

In any discipline where the desired results of print can expended effort can be used instead of high be pictorial in nature, we speed Ektachrome used in most direct find . In those color film particular recorders (described below). There are color cases where the results can be displayed as a print sequence that have a resolution in excess of pictures so as to give the illusion of 200 line pairs/mm of motion, we find computer in any color. That contrasts animation. The sharply with the 20-50 line pairs/mm obtainable uses of animation made by any discipline are with high as varied speed Ektachrome. This becomes important as the disciplines themselves. when we recognize the At the Lawrence fact that a high quality Livermore Laboratory, we film recorder is capable of imaging have been producing computer animated over 150 films for line pairs/mm on 16mm film. Thus, it is clear well over 10 years. These films have been used both that the indirect approach can potentially as a means to aid researchers with their produce a far work and as a means superior product. to communicate to others A second point is the results of their calculations. that during the optical reduction, it is possible to use the Until a few years ago, the only means of for special direct computer effects that are not easily programmed. film output was 35mm black and For example, fades and dissolves white microfilm. This film was can be added either viewed for essentially no additional effort. A far directly on a 35mm projector or else reduced better to control over color can be expected. If 16mm on an in-house optical printer. The the user is unhappy use of 35mm proved unsatisfactory with the choice of , for the obvious the film can be reprinted with different colors. reason that 35mm projectors are not commonplace This like 16mm involves no additional computer runs. projectors. Whereas we actually have There is one special a 35mm projector at LLL, effect we have used it was recognized that at Livermore on a number of occasions that most other laboratories that we exchange film provides a result with will that is not easily accomplished not have a 35mm projector. by programming. As a simple The 16mm motion picture example, consider film is the most two distorted grids that overlay one another. reasonable medium in which to exchange film Let in the one to be in front colored red and the the scientific community. The main reason one to be in back is the abundance of colored green. Suppose further 16mm motion picture film that these grids move with respect projectors. Also recognize the to one another. fact that there At all those points that the grids cross the is a single standard for 16mm film. It is this film will standard show yellow (red + green = yellow). that allows computer animated films This array of yellow to enjoy the popularity dots gives an undesirable they do. effect that can distract from the information The production of color computer generated the programmer was film is a simple trying to convey. It is, extension to the optical re- in effect, a type of optical noise. duction process. The user produces a number To eliminate of 35mm black this effect by programming and white films each correspond- means computing all of the intersections ing to the particular color desired. of These the two grids, breaking the green lines into films are then superimposed one upon another onto segments with the proper spaces left for the 16mm color film with the appropriate red lines. This is color filter inserted a very time-consuming process into the light path. and not a trivial one to be done This technique is currently the efficiently. most popular A far simpler solution is to first make method for the production of scientific computer a negative animation. copy of the black and white film The major drawback is the time that is to be colored red. required to make the film. We call this a negative The users generally mask. This film is clear except for the black expect to see their results soon after running lines. Now their program. the film to be colored red is printed It is for this reason that onto color film with the red computer films are more often filter. After used to show the color film is backed up in the printer, final results rather than as an alternative the film to be to other forms colored green is printed through of computer output. the negative red mask. This These are however, a few is accomplished advantages to by loading both films into the optical printer production of film in this manner. The quality with the mask of the finished closest to the color print film. product is better than by any The black lines from the negative other technique. This is red mask because high resolution prevent light from passing through the "green"

81 film at exactly those places where the two cross. If we choose to avoid the drift by changing Thus, the red lines will be unbroken and appear the filter for each element as required, then in front of the green lines. Since the negative the time to generate a frame becomes excessive. red mask was made from the original film to be The second system used by film recorders colored red, there is no registration problem. consists of movable arms attached to rotary These techniques are used by Hollywood to insert solenoids or stepping motors. These arms are a scene into a different background. arranged so that more than one filter at a time We also use our optical printer for the pro- can be inserted into the optical path. This duction of color stereo movies. The printer is allows the use of subtractive color. The advantage used to place the images side by side as required here is that seven colors (white, red, green, by the Bolex stereo system we use at LLL. It blue, yellow, cyan, magenta) can be produced is clear that the calculation of the stereo without the need for multiple exposures. For pair is only a time-consuming extension to the pictures that contain a collection of colored calculation of a single perspective view. The lines and characters, this can be a timesaving optical reduction is no more difficult then device. We have found that five or six colors the 35mm to 16mm reduction. The lack of popular- is generally sufficient for most of our work. ity of color stereo movies in the scientific The loss in resolution due to more than one community is not because of the lack of interest filter in the optical path is more than made in such film nor the ability to produce them, up for by the poor resolution inherent in high but rather because of the great difficulty of speed Ektachrome. showing them. Given that we can produce the film, The major advantage of the direct color there is the problem of a special lens for recording systems is one of fast turn around projection, poloroid glasses for viewing on a time. After the film is recorded and processed, special type screen that is probably not available. the user has a color film. Even with our in-house In addition there is not a standard for a 16mm optical printer, a color film is at least 24 stereo format. hours away (more typical is one week). The quick The advantages are not all on the side of turn around time means color film can be used the optical printing method. A number of instal- as another form of output and for many users, lations employ a film recorder with direct color this far outweighs the advantage of added quality recording capability. The essential differences obtainable by the indirect procedure. are in the CRT phospor, lenses, and use of a There is, on the horizon, a new technique filter changer. that promises us the best of both worlds. The The white phospor CRT is usually a P4, P24, technique is direct color recording using lasers. P48 or other braod band phosphor. What is impor- The laser has enough light output to directly tant is that the CRT be capable of emitting expose slow speed high resolution color print relatively equal amounts of red, green and blue film. Use of HeNe, Argon, HeCd give red, green light. The lens must be corrected for this and blue light sources eliminating the need for light. Most high quality black and white film filters. Lasers can be focused down to spot recorders employ a P11 phospor and use a lens sizes less than 1 micron. This compares to a corrected for the blue light this phosphor minimum of 20 microns on a white phospor CRT emits. and can provide all the resolution the color There are two systems for generation of print film can handle. color film. The first is called additive color The color laser film recorder is in some and is generally implemented with a filter changer sense, a far simpler device than the color that consists of a single disk containing at CRT film recorder. The three laser beams are least a red, green and blue filter. passed through intensity modulators and folded It is clear that any color can be generated into a single optical path. Using the modulators, by the proper combination of red, green and blue the three lasers can produce a light beam of light. However only three colors; namely, any color. There is no need for multiple and blue can be generated by a single exposures. red, green by exposure. Generation of the secondary colors A color laser recorder has been built require two exposures for each element to be CBS and is operating in a production environment other colors requiring up to at their studio in New York City. It was colored, with and is three exposures. For the reproduction of designed to transfer videotape to film tone images, this triple exposure therefore adjusted to record approximately 500 continuous (30 frames per element is required in any case and a color lines on 16 or 35mm film in real time disk is generally satisfactory. Here each per second). sequentially and only three This particular recorder is capable of over image is recorded time filter changes are required per frame. However 1000 lines resolution. Because of the real where the computer film consists of colored recording requirement, the device becomes lines and points, the color disk has some serious bandwidth limited before it is resolution disadvantages. limited. If one were to slow the frame rate The major drawback is the multiple exposures down, CBS assures us resolution of 4000 pixels required for all but the primary colors. If we presents no major problem. Our only problem choose to minimize the filter change time (between would be in computing all those points. 100-250 ms for most disks), this requires process- The laser recorder has the fundamental times to compute the red, advantage in that it is much faster than a ing the data three no green and blue component. The drift in the CRT comparable CRT system. There is apparently system will cause some mis-registration problem in building a machine that can record deflection a rate resulting in a loss of resolution which is very colored spots with varying intensity at in pictures consisting of many fine exceeding 7.5 million points per second. This noticeable conventional lines. is two orders of magnitude faster than

82 high resolution color film recorders. record a 200 x 200 grid composed of 40,000 short The major drawback to Laser recording today vectors in less then 10 seconds. It will also is deflection of the Laser beam. The CBS system be capable of recording continuous tone type solves the problem by deflecting the beam images with up to 4096 x 4096 points. We have horizontally with a rotating minor and vertically found however, that 1024 x 1024 is more then with a galvanometer. This works well for videotape adequate for any computer movies we have made. but leaves much to be desired for most computer In addition, since the device will not record generated data. Forcing the data to be organized on print film, we end up resolution limited in a raster means that vector and character data by the Ektachrome film. The resolution on the must be scan converted to a raster format. This CRT will be at least 1500 line pairs across the means either special purpose hardware or time- image area. consuming software. In addition, there is the The major limitation to quality is the problem of "jaggies." film itself. We expect to use the high resolution When a vector is broken into a string of capability of the CRT with film points on a finite raster, there will be dis- such as 70mm or 4" x 5". We have found in continuities. For a still picture, this may or talking to users and vendors, a lack of under- may not be objectionable. For movies, the standing of the interplay of CRT phospher, discontinuities along the line move and appear lens design and film response. We were advised as stair steps moving along the lines and are by a vendor that if we purchased their color extremely distracting. This is another form of recorder we would be undertaking a large in-house optical noise. R and D project just to learn to use it. These effects can be suppressed in several A major concern was the elapsed time from ways. The most straightforward approach is to running the program until the film was processed. use a raster finer than the resolution of the For movies, the major portion of that time is film. The only problem is increased computation the recording itself. For continuous tone, and storage costs which rise as the square of three color 1024 x 1024 pictures, at least 30 the raster. We have found that for black and seconds recording time will be required for white 16mm microfilm, a raster of 4096 x 4096 each frame. That comes out to 12 hours per is required to remove the "jaggies". minute of film. We felt that a good preview A second technique that has been used requires capability was essential to prevent recording a recorder with gray scale capability. The screen film generated by a program with a bug. Note is considered to be a grid that is intersected that a laser recorder could record the same by a straight line. Each cell that the line film in real time. passes through is assigned an intensity propor- The video subsystem is designed to preview tional to the area subtended by the line. More movies before committing them to film. It is sophisticated approaches take into account attached to the same minicomputer as the film slope and adjacent cells. This produces a recorder. The primary element is a video disk smooth appearing line. However, the line such as those used by the television networks appears thicker and hence, resolution is reduced. for the instant replay. It will have at least RCA, among others, is developing a solid three independent channels each capable of state deflector for laser beams based upon a recording 300 frames. The three channels can TeO2 crystal. The amount of deflection is be used to record an RGB picture to be viewed small but can be opticly amplified. Current on an RGB monitor for the highest quality crystals have a resolution of 512 points which pictures. This allows us to look at a movie 10 is not enough for high quality film. Deflectors seconds at a time. We can select frames at with 2048 point resolution are operating in a random for checking or certain whole sequences laboratory environment and will be available to examine timing. soon. If it is required to look at a longer The solid state deflector has the obvious sequence, we can encode the RGB picture into advantages over a rotating mirror deflector. NTSC format and record it on videotape. In order A recorder built with these crystals will neces- to do this with standard television equipment sarily be more a complex device. This is due means the picture can contain only 480 scan lines in the main to the fact that the deflection instead of the expected 512. The commercially is a function of the wavelength of the incident available digital color television systems use light. We believe that these problems can be a 512 x 512 picture which is not compatible solved and a high quality color laser recorder with standard television. It cannot be encoded will be built soon: Its only a matter of money. and recorded with standard television equipment. We believe that the color laser film recorder We have given up those mystical 32 lines for the will provide a quantum jump in the production sake of compatibility. There are many people and quality of color computer generated film. here who would willingly trade the ability to LLL is now in the process of building a record their material in color for those 32 lines. complete system for the generation of color Some habits are just hard to break. computer generated film. The objectives were It is important to note that the NTSC to provide the user both fast turn around time recording does not have the resolution of the and high quality color film output. Since a RGB image. This is due to the bandwidth limitation color laser recorder was not within our budget, of the 1" video tape and the NTSC signal. We do a dual system has been designed. It will consist not envision this to be amajor problem since of a high quality direct color film recorder the video is for a quick check only with the and a video subsystem. The film recorder will quality image to be recorded on film. We do, use the subtractive color technique to provide however, expect to'find a class of user for whom for seven separate colors without the need color video is adequate for their needs. for . It will be fast enough to The recent advent of the digital time base

83 corrector gives us one new dimension not possible a few years ago with a 1" video recorder. We can record our RGB picture as three separate images and using the time base corrector, dump the three images onto the video disk in perfect time registration for high quality viewing. Of course only ten seconds at a time can be viewed. As far as we have determined, this is the first time this will have been tried. This entire system (video disk, video tape, video switches, etc.) is under computer control. We feel that we will have a most advanced system for the production and editing of computer generated

84