<<

SMPTE TUTORIAL Searching for the Perfect Aspect

By Mark Schubin

A debate is currently taking place over the appropriate for some degradation of the imagery advanced displays. Any selected aspect ratio is inherently involved.' There are only three basic incompatible with any other and will require the use of some form of methods of accommodating existing accommodation technique. The derivation ofthe 16:9 (1.78:1) aspect ratio material in a fixed aspect ratio on a display of a different fixed aspect ratio, from accommodation techniques and display modes is explained, as is the though the techniques may be com­ relationship between aspect ratio and display memory. Research into the bined. These three basic techniques are history ofaspect indicates that the 1.78:1 aspect ratio was adopted shown in Fig. 1. by the Standards Committee of the Society of Motion Picture Engineers Figure I a shows the truncation (SMPE) in 1930. It also indicates that the factors that may initially have method, a variant of which is some­ led to motion picture systems may no longer be applicable. times referred to as "." The research for this paper found no clear indication ofa preference for When going from a wider aspect ratio any particular aspect ratio for moving images nor any physiological rea­ to a narrower one in this method, the son to favor one over another. The research did show that cinematogra­ heights of the two images are matched, phers have not always favored the same aspect ratio. and any excess width in the wider image is removed from the display. The A 1988 paper entitled "Another That incompatibility became most position of the displayed rectangle in ./'"'\..Method of Aspect-Ratio Conver­ noticeable in 1961, when the 1953 the pan-and-scan mode may vary either sion For Use In Receiver-Compatible movie, How to Marry a by gradual panning (or tilting, in the EDTV Systems" begins: ''Two systems Millionaire, was broadcast on the NBC case of accommodation of a narrower with different aspect ratios are inherently television network.' Intended to be seen aspect ratio) or by rapid repositioning incompatible."' (EDTV is extended-def­ at an aspect ratio of 2.55: 1 (and with (cutting) between frames. inition television.) The statement bears image composition intentionally filling Figure 1b shows the shrinking looking into. the wide frame), the movie was truncat­ method, referred to as "letterbox," due For the purposes of this paper, aspect ed to television's 4:3 (1.33:1). Almost to the shape of the shrunken image ratio will be defined as the ratio of an immediately, technical publications window when a wider aspect ratio is image's width to its height. Ever since began to carry information about how being accommodated on a narrower there have been rectangular images, best to deal with the "conversion" of display. The black bands need not be there have been aspect ratios (and it one aspect ratio to another.' evenly spaced. It is often the case that may be argued that even elliptical the lower band (when a wider aspect images have aspect ratios). Aspect Ratio Accommodation ratio is being accommodated) is made We are surrounded daily by multiple In fact, imagery is not "converted" larger for the purpose of carrying subti­ aspect ratios not seeming to cause any from one aspect ratio to another; one tles, and, as will be discussed later in incompatibility problems. Images in aspect ratio is merely accommodated this paper, when a narrower ratio is newspapers and magazines have a vari­ by another, almost invariably with being accommodated, the elimination ety of aspect ratios both greater and less than one; the same is true of paintings and photographs. Even some computer display screens may be rotated from a horizontal aspect ratio (landscape) to a vertical one (portrait). When theatrical motion picture and television screens are !IJJJ OJ considered together for the purpose of displaying the same imagery, however, the inherent incompatibility becomes ..r---···------, more clear.

Presented at the 137th SMPTE Technical Conference in New Orleans (paper no. 137·61) on September 8. , J0 1995. Mark Schubin is a technological consultant in New York. NY 10036. An unedited version of this paper appeared in Moving Images: Meeting the B c Challenges, SMPTE. 1995. Copyright © 1996 by the A Society of Motion Picture and Television Engineers. Inc. Figure 1. Aspect ratio accommodation: (a) truncation; (b) shrinking; and (c) .

460 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL of one of the side bands offers the pos­ preserves the original image composi­ oped for excitation of the unused areas sibility of stacking additional images in tion but reduces the visual angle avail­ with signals that vary to match average the other side band, a technique that has able to the viewer and, often, resolution picture level, however, and those tech­ been referred to as multiple picture-in­ as well. Detail that is just perceptible in niques appear to eliminate image strip­ picture (MPIP). an image when it is viewed at a particu­ ing." No investigation of viewer accep­ Figure lc shows the distortion lar will be lost if the tance of display stripes with varying method, whereby the linearity of the same image is shrunk on the same dis­ brightness was found in the research for geometry of the image is changed to play. this paper. squeeze it into a different display In most cases when the viewing The differential phosphor luminance shape. In a recent variation on this tech­ screen is -based, this shrinking decay issue is related only to displays nique, a nonlinear distortion is used, results in noticeably empty portions of using phosphors, such as those based affecting the edges of the image more the display device, a condition that has on typical direct-view or projection than the center (e.g., in two of JYC's been considered objectionable to audi­ cathode-ray tubes. Some video projec­ consumer widescreen projection ences by some television program­ tors, such as the Schlieren-optics-based receivers). mers." One television manufacturer Eidophor,!" have never used phosphors, As can be seen from the two rows of (lVC) has introduced a widescreen and advanced television displays may Fig. 1, the same basic methods apply model with a mechanical masking sys­ be able to take advantage of other phos­ whether the original image is wider tem that covers the unused portions of phor-free technologies. 11.12 than the display or narrower. In fact, the display much as drapes mask There are two other techniques asso­ the same techniques apply whether the unused portions of some motion picture ciated with aspect ratio accommoda­ two aspect ratios are both , both theater screens, possibly resulting in the tion, but they require that either the video, or one of each. From 1961 to reduction or elimination of such objec­ image or the display be effectively non­ date, however, generally only the tech­ tions.' fixed in shape. One of these techniques niques of the upper row have been A potentially more serious problem is sometimes used in video walls. As seen, as widescreen movies have been related to the shrinking technique is shown in Fig. 2, a video walI com­ shown on narrower video screens in differential phosphor luminance prised of 4:3 image modules can create homes, aircraft, or other venues. Unless decay, a reduction in the light output a 4:3 image when stacked in a 3 x 3 or otherwise specified, the word wide­ of the cathode ray tube phosphors in 4 x 4 module configuration, but the screen, for the purposes of this paper, the active picture section relative to same modules can create a 16:9 will be used as defined by the British that in the blank section, often affect­ (1.78:1) image when stacked in a 4 x 3 Kinematograph Sound and Television ing blue phosphors more than red or configuration. Society (BKSTS): "in general, pictures green." As a result, when the full dis­ When the goal has been not aspect presented with an aspect ratio greater play area is viewed, the shrunken ratio accommodation but the creation than 1.4:1."5 image area can become visible as a of a different aspect ratio than is com­ AlI of the accommodation tech­ stripe somewhat yellower than the rest monly used in a particular medium, niques of Fig. 1 are problematic. of the display. The effect is greater in similar modular-screen techniques have Sometimes aspect ratio accommodation projection displays than in direct-view been used in many film and video pro­ is demonstrated with so-called "neu­ displays due to the higher beam cur­ jection systems. These range from the tral" imagery: pictures that appear no rents of the former. 19th-century Cineorama system (using less desirable when cropped. Motion It has been suggested that the differ­ ten interlocked motion picture film pro­ pictures and television shows are not ential phosphor luminance decay prob­ jectors)" to the current Geographica shot to be neutral, however. The trunca­ lem may be eliminated by making the theater (using three synchronized video tion technique clearly causes portions inactive sections of the display gray sources) at the National Geographic of the image to be lost, and the variants instead of black, but in one experiment, Society's Explorers Hall in Washing­ associated with pan and scan introduce the outline of an inactive section of a ton, D.C. The original motion or cutting never intended in the direct-view picture tube was visible widescreen movie process, using three original. after 5,000 hours, even though that sec­ synchronized film projectors, is proba­ The distortion technique clearly tion had been excited with a 50% gray bly the most famous of these systems.' changes the shape of not only the signal. Techniques have been devel- It has been suggested that, at some image but also people and objects con­ tained within it. An informal survey conducted in association with the research for this paper found that dis­ tortion in the range of 2 to 6% may be considered acceptable, but that is much less than the amount needed to accom­ modate a typical widescreen movie on a conventional television display or vice versa. The shrinking technique (letterbox) Figure 2. Modular display configurations.

SMPTE Journal, August 1996 461

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL future date, consumers will be able to This restriction can affect not merely avail themselves of low-cost, nonfixed­ aesthetic shooting preferences but also aspect-ratio displays, but at the moment, plot. In the play, Largely New York this technique does not appear to be (1989), for example, a character applicable to most homes. trapped in a television signal tries to get Action out by pushing on the edges of the frame; if that material was captured in a The other technique of aspect ratio shoot-and-protect system, in at least 1------Fluff accommodation may, in fact, be the one of the aspect ratios those frame most common used today, but it cannot edges would not be properly located, so be used for existing material shot with the plot device could not be used. just one aspect ratio intended. The tech­ Dramatic and comedic timing is some­ nique is sometimes called "shoot and times affected by the moment when a protect." It has been used to accommo­ particular character enters a frame; in a date different aspect ratios both for the­ shoot-and-protect system, a character F atrical film projection and for video may appear at different times on differ­ I display. ent displays. u Action In such a system, during production There is an additional problem asso­ f f the captured image is framed so as to ciated with shoot and protect of the make images appear in a desired fash­ form where the action area is the ion in one aspect ratio while additional wider aspect ratio (upper portion of area is suitably protected in the overall Fig. 3). This problem relates to theatri­ Figure 3. Shoot and protect. frame to allow the images to be seen in cal projection framing. With no visual a different aspect ratio without lighting indication of the upper and lower lim­ had been created in a shoot-and-protect instruments, masking, microphones, its of the wider frame, a projectionist mode. puppeteers, or the edges of set pieces must guess at the correct framing, and Unfortunately, such presentation becoming visible. Such framing is that framing is not necessarily intend­ often shows viewers image areas facilitated by the presence of reference ed to be centered in the protected aper­ intended by the director and cinematog­ I lines (reticles) on the viewfind­ ture. 7.1R.19 Despite all of these prob­ rapher not to be seen. It is possible to er for the desired (action) aspect ratio. IS lems, because home video, alone, has see a microphone intruding into the top Thus, the inner action area is some­ resulted, since 1986, in greater domes­ of the image in Hatari! (1962), for times referred to as the reticle region, tic wholesale gross revenues for movie example, when that movie, intended to and the outer frame is sometimes distributors than has theatrical be seen on a wider aspect ratio theatri­ referred to as the aperture." The area release," there is a strong financial cal screen, has its full exhib­ between the reticle and the aperture, incentive for this form of aspect ratio ited on a 4:3 display. Theatrical projec­ where significant action is to be avoid­ accommodation, whatever its prob­ tion masking would have kept the ed, has been referred to as "fluff." lems. Aside from its other negative microphone out of the shot. Nudity Reconsidering Fig. 1b, in a shoot­ aspects, in 1987, pan and scan (and intended not to be seen in Bonnie and and-protect system, the black bands associated) costs ran as high as $8,000 Clyde (1967) is similarly a result of would not be black but would contain, per feature at one cable television net­ full-frame exhibition of material shot to instead, a continuation of the back­ work.' There is also a need to consider be shown with much less of the film ground of the image, the continuation accommodation of shot in very frame visible. In Psycho (1960), set area avoiding anything critical to the wide aspect ratios on much narrower masking is visible when the full 4:3 action. This is shown in Fig. 3. theatrical screens." frame is presented.' Such visible micro­ A shoot-and-protect system allows It should be pointed out that the phone booms, masking, set edges, and aspect ratio accommodation without viewfinder markings of a shoot-and­ even lighting instruments have been image truncation (and its associated protect system may be used even when attributed, in some cases, to sloppy additional pans or cuts), image shrink­ there is only one intended aspect ratio, filmmaking; the real cause is exhibition ing (and its associated reduced viewing simply to allow the use of imaging in an aspect ratio never intended by the angle, reduced resolution, objectionable equipment designed for a different director or cinematographer. blank screen bars, and potential differ­ aspect ratio. The Sony Jumbotron In to the shoot-and-protect ential phosphor decay), and/or image screen at the Skydome in Toronto has systems of Fig. 3, it is also possible to distortion. On the other hand, it creates an aspect ratio of 10:3 (3.33: 1), but its create a shoot-and-protect system major restrictions in the way sets can be images come from conventional 4:3 matching the shape of neither of the dressed and lit, the way sound can be television with appropriate aspect ratios needing accommodation picked up, and the way action can be viewfinder reticles. In fact, many but providing both with "equal pain." framed. A character cannot be posi­ widescreen films shot this way (with a Such a system would have an outer tioned at the edge of a frame, for exam­ reticle framing a widescreen image in a protection frame (aperture) as high as ple, if that edge will not appear in one conventional 4:3 frame) have been the narrowest desired aspect ratio and of the aspect ratios. shown on 4:3 screens as though they as wide as the widest (when both have

462 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

Other beneficial properties have been claimed for the 16:9 aspect ratio. ,....-r---- -_ .. ";, . , The two most common widescreen . nonanamorphic theatrical projection aspect ratios worldwide are 1.85: I and II 1.66: 1.'9 A linear average of the two is ~ ., : La __ [I , . OJ 1.755: I, again close to 1.77: I (but clos­ ... --- - _ -~ : .. .. cr still to the theatrical projection ratio of 1.75: I adopted by some major film distributors" and standardized by A B C 1956).'R There is also an interesting mathematical progression from 4:3 to Figure 4. Equal-area shoot and protect. 16:9 (4/3 * 4/3) to approximately 2.35: I (4/3 * 4/3 * 4/3). equal area), and an inner action frame 16:9 It has been reported that the 16:9 (reticle) as high as the widest aspect Figure 4c shows the actual equal­ aspect ratio was unanimously approved ratio and as wide as the narrowest, The area-based shoot-and-protect shape by the Working and that a num­ outer and inner frames would have the proposed by Kerns Powers, then with ber of cinematographers were same aspect ratio. This new shooting RCA Laboratories, at the meeting of involved.":" Such reports have also aspect ratio is derived from the follow­ the SMPTE Working Group on High been disputed." Some of the conflict ing formula: Definition Electronic Production on may simply be semantic (e.g., what makes a person a cinematographer"). D=W*(IIW)"'/(I/N)'" (I) May 4, 1984. It is clearly closer to that of Fig. 4a than to that of Fig. 4b. The For the purposes of this paper, the argu­ where D is a derived compromise aspect ratio was derived from the above ments over who knew or approved aspect ratio, W is the widest of a range formula through the substitution of 4:3 what when are not significant. of desired aspect ratios, and N is the for the narrowest aspect ratio and Interested readers are referred to narrowest of the range. 2.35: I (then the projected shape of sources listed here and elsewhere in The benefit of such a system would anamorphically expanded 35mm the­ this paper.":" be to reduce the amount of nonaction atrical movies, commonly referred to as Even before the Working Group linear dimension in all aspect ratios "scope," from the CinemaScope sys­ meeting, in 1983 Joseph Nadan, then between the narrowest and widest tem) for the widest. The resulting ratio with Philips Laboratories in the U.S., selected. Since the shoot-and-protect is just over 1.77:I, which was rounded illustrated possible advantages for a systems of Fig. 3 have no nonaction up to 16:9 (somewhat less than 1.78:I) consumer high-definition television area for one aspect ratio (the reticle), for convenient circuit design. (HDTV) set that had a display shape of however, that aspect ratio would actual­ At the same time, the Motion Picture 16:9. In a "polyscrcen" mode, the 16:9 ly suffer an increase in nonaction area Association of America was evaluating shape could be divided into twelve 4:3 through this plan. Of course, if a dis­ a new anamorphic projection format images (an inverse of Fig. 2's video play screen happened to have the same using an anamorphic expansion ratio of wall); in MPIP mode, it could provide aspect ratio as the derived shooting sys­ 1.5: I instead of the scope 2: I to three stacked 4:3 images adjacent to tem, it could be perfectly framed, with improve screen illumination efficiency one larger one." (Note that some in the no nonaction area. and reduce projected jitter. That format consumer electronics industry refer to Figure 4 shows three possibilities for would have had an aspect ratio of l.5/2 this mode as "picture-outside-picture," such equal-area-based shoot-and-pro­ times the 2.35: I aspect ratio, or or POP.) These display modes arc shown tect systems. Figure 4a shows what I. 7625: I, very close to the derived in Fig. 5. might be considered a sublime example shoot and protect 1.77:I. 16 It was also pointed out that 16:9 was of such a system. The two extreme As a proposed electronic production "friendly" to two international recom­ aspect ratios being accommodated are format, 16:9 need not ever have been mendations: that HDTV have twice the so close that the inner action frame is used in a shoot-and-protect mode. resolution of non-HD TV; and that dig­ proportionally even larger than that of Given sufficient resolution, a 16:9 elec­ ital component non-HD TV have 720 the SMPTE Safe Action area accom­ tronic imager could be used to capture active picture elements (pixels) per modating overscanning in TV sets. single-aspect-ratio moving pictures in scanning line. Twice 720 is 1440 for a Figure 4b shows what might be con­ any desired aspect ratio. For the 4:3 aspect ratio. For a 16:9 aspect ratio, sidered a ridiculous application of such extremes of 4:3 and 2.35: I, the 1.77: I it would be 1920, and equivalent verti­ a system. The inner action frame is so shape would allow minimum waste of cal picture clement resolution (square small relative to the outer protection photosensitive area (for a multiaspect­ pixels) would dictate 1080 active scan­ frame that it is difficult to see how any ratio imager) and simplify lens design. ning lines, for a total of 2,073,600 pix­ 21 director or cinematographer could In single-aspect-ratio use, the action els. A 2-Mpixel memory has 2 pixels: make meaningful use of such restricted area could fill the intended display 2,097,152, a near-perfect match." Even framing. aspect ratio. a slight aspect ratio increase to the

SMPTE Journal, August 1996 463

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL common widescreen 1.85: I would lar display shape modification shown ratios, and Table 1 shows that there are require 2,157,840 pixels, a poor fit to in Fig. 2. many other possibilities. common random access memory One more rationale for the selection Still more combinations are possible (RAM) sizes. of 16:9 was probably unknown to pro­ if there are multiple columns of MPIP The 16:9 aspect ratio also allowed ponents of the ratio in the 1980s. In or if the images are not contiguous on for a form of simplified dual aspect 1930, the standards committee of the the screen. Though it has a 16:9 screen, ratio transmission. If a 16:9 image is Society of Motion Picture Engineers for example, a recent RCA television transmitted at the common composite (SMPE), recommended a new method receiver allows up to five MPIP video digital sampling rate of four of projecting large-screen movies from images, not merely three." times the color subcarrier frequency wider, 50mm film. The screen shape The display-memory-size benefits of (4fsc)' a receiver can recover the full used for the Society's viewing purpos­ 16:9 HDTV are also predicated on the 16:9 image by reading its memory at es was 41 ft by 23 ft, 1.78: I (the very specific requirement of doubling 4fsc or can get a 4:3 truncated version Society rounded off its widescreen film the 720 active horizontal pixels of ITU­ (potentially in a pan-and-scan mode) by aspect ratio recommendation to 1.8:I). R Rec. 601 for a widescreen display. If reading the memory at 3fsc. 27 Only The aspect ratio was said to be in line the common practice of using 704 pix­ aspect ratios that have a 4/3 relation­ with the desires of the Academy of els to represent the picture width is con­ ship (as do 16:9 to 4:3 and 2.37: I to Motion Picture Arts and Sciences sidered instead, even a 1.85: I display 16:9) can make use of this technique (AMPAS).l' In 1953, an aspect ratio of can use common memory devices with common sampling rates. approximately 16:9 was again consid­ (twice 704 divided by 4/3 is 1056; Since "1.85 is far and away the most ered as a standard ratio for theatrical 1056 2 times 1.85 is 2,063,002, well common aspect ratio for motion pic­ projection." within the 2-Mpixel limit of tures filmed in the .''" the This plethora of beneficial aspects of 2,097,152). Furthermore, while the proximity of 16:9 to 1.85:1 (less than 16:9 has sometimes been carried too 1920 active pixels/line of some 16:9 4% difference) could also be consid­ far. It has been claimed, for example, HDTV systems have a very simple ered beneficial for the display of that 16:9 is the only aspect ratio that relationship to the 720 active pixels of movies. Circuit design generally causes the inner reticles and outer aper­ Rec. 60 I, there is no such simple rela­ requires values for multipliers tures of Fig. 4 to have the same shape; tionship between 1080 active scanning and dividers. The 1.85:I ratio can be a mere glance at Fig. 4 indicates that all lines and the active scanning lines of expressed as the complex 37:20. The equal-area shoot-and-protect aspect either 525/59.94 or 625/50 television simpler 9:5 ratio is a very close 1.8:I, ratios have the same property. systems. Even if only 480 active lines but it could not make use of the simpli­ The linear position of 16:9 between are considered for 525/59.94 instead of fied dual-aspect-ratio transmission sys­ 1.66:I and 1.85: I is also of question­ the more traditional 483 or 484, the tem described in the last paragraph, nor able benefit. A linear average of the resulting simple relationship, 9:4, is dif­ would it be able to double digital com­ extreme aspect ratios of 4:3 and 2.35: I ferent from the horizontal relationship. ponent resolution and still fit in a 2­ is just over 1.84:1, a near-perfect match If a relationship with Rec. 60 I is Mpixel memory. The 16:9 shape is the for the existing theatrical widescreen ignored, a 2:1 display offers a perfect closest aspect ratio to 1.85: I offering aspect ratio of 1.85:1 (though it may be match to 2-Mpixel memories (2048 x those other electronic system design argued that, to obtain the benefits of a 1024), albeit with somewhat less verti­ benefits. 4/3 relationship with 4:30, 16:9, less cal resolution than 1080 active lines. The 16:9 aspect ratio is also well than 4% smaller, might have been cho­ Benefits derived from 3-perf produc­ matched to the technique of economiz­ sen even if the desired aspect ratio were tion have also been questioned. The 3­ ing by shooting film frames three per­ 1.85:I). perf format has been said to be poten­ forations high instead of the usual The polyscreen and MPIP advan­ tially more unsteady than 4-perf, to 28 four. ,29" o Again, it is the 4/3 relation­ tages of 16:9 may also have been offer poorer audio frequency response, ship between the 4:3 aspect ratio and overemphasized. While it is true that and to require difficult projector con­ the 16:9 aspect ratio that makes 16:9 only 16:9 yields a polyscreen of twelve version." For the moment, it also has an appropriate three-perforation (3­ 4:3 images and an MPIP of 3, Fig. 6 additional costs associated with its perf) aspect ratio. The same 4/3 rela­ indicates some poly screen and MPIP being a nonstandard format." tionship also makes possible the modu- possibilities of the 2: I and 5:3 aspect A rarely discussed issue is associated with the concept of using identical scanning characteristics in both an equal-area shoot-and-protect produc­ tion format and a display. In the case of 16:9, for example, the shoot-and-pro­ tect system would allow the extraction of a 2.35: I image with 25% nonaction area at the sides or a 1.33:I image with 25% nonaction area at the top and bot­ tom. A 16:9 display occupying only the Figure 5. 16:9 alternative display modes. action area would be perfectly framed,

464 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

economic issues that may be associated with it (from the mass manufacture of - single-aspect-ratio electronic imagers or from potentially increased bit rates - for constant-vertical-resolution trans­ - mission of wider aspect ratios, for example), is one that should clearly be favored by both filmmakers and view­ ers. Filmmakers have long complained of the need to compromise their theatri­ cal framing to accommodate tele­ vision.v? Consumers are already being offered choices for aspect ratio accom­ modation in some displays and videodiscs. The cable television chan­ nel American Movie Classics (AMC) currently offers its widescreen movies in repeated showings on the same day, Figure 6. Alternative MPIP andpolyscreen. once in a pan-and-scan format and the following time in letterbox; if ATV with zero "fluff." If, however, the cam­ (ATV) distribution and displays at the receivers offered the capability of local­ era and display have identical scanning Artists Rights Symposium, portions of ly accommodating different aspect characteristics, the display cannot be which are excerpted here: ratios, consumers could opt for their perfectly framed in the action reticule; "While the ASC would prefer an choice at any time. it must occupy the entire aperture and aspect that matches our current widest A digital transmission system, as is will, therefore, suffer 43% nonaction screen production standard of 2.40: I, supposed to be used for HDTV, lends area - considerably more than either a we realize that practical engineering itself to the encoding of movies in any 4:3 or a 2.35:1 display. and manufacturing requirements must aspect ratio. A few bits can indicate to This problem is not specific to the also be considered. Thus, the ASC receivers what the picture aspect ratio 16:9 aspect ratio, only to the concept of advocates 2: I as an adequate, if not is as well as carrying any pan-and-scan an equal-area shoot-and-protect pro­ ideal, standard ratio." [Author's note: or other accommodation information duction system sharing scanning char­ The 2: I expansion of the current 35mm that the filmmaker, distributor, and/or acteristics with a display. Even the 4:3 anamorphic projection aperture yields a programmer choose to offer. The dis­ aspect ratio suffers similarly in some 2.4: I image rather than 2.35: I, and the play memory (necessary in the receiver film shoot-and-protect systems." proposed anamorphic theatrical projec­ for inverse bit-rate reduction) can be That shoot-and-protect display scan­ tion system with a 1.5:1 squeeze/expan­ read in whatever manner the viewer ning issue notwithstanding, for any or sion, therefore, now yields a 1.8: I desires, subject to the capabilities of the all of its benefits, real or imagined, and image rather than 1.76:1.]38 TV set, of course. perhaps for other reasons (such as polit­ "Every original film work would be ical or economic considerations), with­ mastered and distributed over U.S. 2:1 in a few years after its 1984 introduc­ ATV (Advanced Television) in its The second ASC position, adopting a tion as a production technique, the 16:9 native aspect ratio. This might be 2.4: I, 2: I , is not as obvi­ aspect ratio had been adopted for 2.2:1, 1.85:1, 1.66:1, or 1.33:1 (or ously beneficial to either filmmakers or ATV/HDTV display not only in the 1.78:1 ifin the current HDTV format)." viewers. A 2: I display aspect ratio U.S. (except for some HDTV transmis­ "The ATV system should be solves none of the aspect ratio accom­ sion system proponents) but also in deployed so that all ATV receivers modation problems, except that it Europe and Japan, where different have a 2.0:1 aspect ratio, at any and all favors the widest aspect ratios to the aspect ratios had originally been pro­ standards at which they might detriment of narrower ones. posed.":" By 1994, 16:9 appeared to be operate.'?" As was noted previously, a 2: I universally accepted, not only for It is worth separating this position aspect ratio either uses picture memo­ HDTV but also for widescreen video of into the two parts that affect the choice ry circuitry poorly or must use hori­ ordinary resolution in such systems as of a 16:9 aspect ratio. First, the ASC zontal resolution lower than twice that the European PALplus and Japanese would like all "films" (the term is clear­ of digital component video. For any Clearvision. ly used loosely, since it is meant to given screen width or diagonal mea­ include HDTV productions) mastered surement, a 2: 1 aspect ratio will pro­ Questioning 16:9 and distributed in the aspect ratio for vide a smaller image than will 16:9. In April 1994, however, the which they were shot. Second, they Even a screen of equal area will American Society of Cinematogra­ would like ATV receivers to have a 2:1 appear shorter. Only a screen of equal phers (ASC) presented positions on aspect ratio display. height will clearly be bigger, but, if it aspect ratio of advanced television The first position, excluding any is direct-view picture-tube-based, it

SMPTE Journal, August 1996 465

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

Table 4 shows display resolution Table 1 - Polyscreen and MPIP Possibilities reduction caused by letterbox and fixed display memory size. As in Table 3, Aspect Ratio :9 :1 Polyscreen MPIP screens may be of any size. For a fixed 8:3 24 2.67 2 1 memory size, a narrower aspect ratio 2:1 18 2 6 2 offers more vertical resolution and a wider aspect ratio offers more horizon­ 16:9 16 1.78 12 3 tal resolution. The data have been nor­ 5:3 15 1.67 20 4 malized so that the narrowest image 8:5 14.4 1.6 30 5 (1.33:1) on the narrowest display (4:3) 14:9 14 1.56 42 6 shows zero vertical resolution reduction and the widest image (2.4: 1) on the 32.21 13.7 1.52 56 7 widest display (2: I) shows zero hori­ 3:2 13.5 1.5 72 8 zontal resolution reduction. Again, the 2: I display offers the best results for 2.2: 1 and 2.4: 1 images and will also be deeper, heavier, and more letterboxed 1.85:1 image than would a the worst for 1.33:I images. Again, the expensive. 2: I display. As it was intended to do, 16:9 display offers the best results for The cost of direct-view picture tube the 16:9 ratio provides approximately 1.85:I images. displays is related to a of fac­ equal-sized images for both extremes Tables 2,3, and 4 all show, as proba­ tors. The costs of phosphor screens and of image shape. bly seems obvious, a perfect fit for shadow masks are related to their area. A nonpicture-tube-based display 1.33:1 images on 4:3 display screens. The cost of electron beam deflection technology might not have a diagonal There is a very large installed base of circuitry is related to display width. The size-related cost basis, however. Tables 4:3 displays in TV sets and computer cost of a glass bulb is related to its vol­ 3 and 4 compare the same image and monitors worldwide, as well as very ume, a specification derived from display shapes for screens ofany size. large libraries of roughly 1.33: I mov­ screen area times depth, with depth Table 3 shows the amount of screen ing image programming - virtually all based on deflection, which is width­ area left blank to fit images into dis­ pre-1953 movies, virtually all television related. The overall display cost, there­ plays of a different aspect ratio. The programming to date, virtually all fore, is based somewhere between area same figures indicate the amount of nontheatrical films, and many post­ and width, or, roughly (for aspect ratios resolution reduction from the maxi­ 1953 movies. In 1958, 750 pre-1948 greater than 1:1), on diagonal measure­ mum that the display can deliver. The movies were sold by Paramount to ment. boxed area indicates reduction of verti­ MeA for $50 million; in 1994, another For any given diagonal measure­ cal resolution (traditionalletterbox); the Paramount transfer, this time of 4,000 ment, the largest possible rectangular unboxed area indicates reduction of 4:3 TV episodes (as well as feature display will be square. Narrower aspect horizontal resolution (blank screen films and other desirable properties) ratio displays (of 1:1 or greater) will be areas at the sides of the image). was valued at $9.6 billion." larger in area than wider ones. Table 2 Again, the 2:1 display offers the best If ATV brings a wider aspect ratio shows the combined effects of a fixed results for the widest aspect ratio into the home, it may be expected that diagonal-based screen size and "letter­ images and the worst results for the sports coverage, shopping programs, box" image display on overall image narrowest. Again, for 1.85: I images, game and talk shows, other entertain­ size for five common motion picture the 16:9 display is superior to all oth­ ment programming, and even news image shapes (expressed as ratios to ers. Again, the 16:9 display provides coverage will eventually migrate to the one, as is common currently in cine­ roughly equivalent results for the wider aspect ratio. The older movies matography) and five common existing extremes of image shape. and television programs that form the or proposed display aspect ratios (expressed as integer ratios, as is com­ mon in display engineering). The data Table 2 - Image Sizes for Letterboxed Equal-Diagonal Displays have been normalized so that a 1.33:1 Display image on a 4:3 display is considered 100%. 4:3 100% 80% 72% 61% 56% While the 2: 1 display provides the largest 2.4: 1 and 2.2: 1 images, it pro­ 3:2 85% 87% 78% 66% 60% vides the smallest 1.33:1 images. The 5:3 73% 92% 83% 70% 64% center column, representing the most 86% 72% 66% common U.S. theatrical film aspect 16:9 67% 83% ratio of 1.85:1, indicates that not only 2:1 55% 69% 77% 76% 69% does a 16:9 display provide a much 1:33:1 1.66:1 1.85:1 2.2:1 2.4:1 larger image than would a 2:1 display but even a 3:2 display provides a larger Image

466 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

Table 3 - Blank Screen Area and Screen-Based Resolution Reduction for Letterboxed Displays Display 4:3 0% 20% 28% 39% 44% 3:2 11% 10% 19% 32% 38% 5:3 20% 0% 10% 24% 31% 16:9 25% 7% 4% 19% 26% 2:1 34% 17% 8% 9% 17%

1.33:1 1.66:1 1.85:1 2.2:1 2.4:1 Image

Table 4 - Fixed Memory SIze Resolution ReductIon for Letterboxed DIsplays VertIcal Resolution ReductIon HorIzontal ResolutIon ReductIon Display 0% 20% 28% 39% 44% 4:3 18% 18% 18% 18% 18% 6% 15% 24% 36% 41% 3:2 23% 13% 13% 13% 13% 11% 11% 19% 32% 38% 5:3 27% 9% 9% 9% 9% 13% 13% 17% 30% 36% 16:9 29% 12% 6% 6% 6% 18% 18% 18% 26% 32% 2:1 34% 17% 8% 0% 0%

1.33.1 1.66:1 1.85:1 2.2:1 2.4:1 Image 1.33:1 1.66:1 1.85:1 2.2:1 2.4:1 basis of much of the programming of aspect ratio for the extremes of 1.85: I list of the highest grossing films of all such channels as American Movie on the narrow end and 2.2: I on the time may be compared with a listing of Classics, Nick-at-Nite, and Turner wide end. Those are the widest com­ their aspect ratios." Such a comparison Classic Movies will remain 1.33: I, monly projected nonanamorphic 35mm indicates that while many of the top however. theatrical aspect ratio and the normal 100 films were made at a 2.4: I aspect Table 2 also shows that, of common 70mm theatrical projection aspect ratio, ratio, they account for much less the­ existing or proposed display shapes, 4:3 respectively. While more favorable to atrical revenues than do the narrower offers the largest screen for a given cost 2.2: I and 2.4: I image aspect ratios than aspect ratio films on the list. Lists of for technologies with cost related to is 16:9, 2: I is less favorable to the most the top-grossing films of 1994 in both diagonal measurement, such as direct­ common theatrical 1.85: I and 1.66: I domestic and foreign markets in the view picture tubes. A transition of tele­ aspect ratios" and is much less favor­ February 13-19, 1995, issue of Variety vision to any widescreen aspect ratio able to the 1.33: I aspect ratio. It would, yield similar results. will introduce problems relative to the therefore, be an appropriate compro­ While a 2: 1 aspect ratio is more existing 4:3 display and 1.33: I pro­ mise display format only if there is favorable to the 2.4: I theatrical aspect gramming bases: Thus, it has been some reason to favor the 2.2: I and ratio than is 16:9 or any narrower argued that a 4:3 aspect ratio should be 2.4: I aspect ratios over 1.85: 1, 1.66: 1, aspect ratio, it does not match it in the retained for ATV displays:' and 1.33: I. same way that a 4:3 display matches Unfortunately, as Tables 2, 3. and 4 It has been reported in the past that 1.33:1 programming. The 2:1 (16:8) show, a 4:3 display offers the smallest wider aspect ratio films (2.4: 1) cam aspect ratio is, in fact, considerably images, the lowest resolution, and the more theatrical revenues than other closer to 16:9 than it is to 2.4:1. greatest blank screen area when dis­ films." That is definitely 110t the case at Whatever its disadvantages, a hypothet­ playing the widest aspect ratio imagery. the time of this writing. The highest ical 2.4: I display would at least have Since it is unlikely that homes will grossing film of all time, as reported by the advantage of allowing side panels have different ATV displays optimized Variety in its February 20-26, 1995, or drapes to mask unused portions of a for different aspect ratios of program­ issue, is E.T. - The Extraterrestrial screen for all but the few movies that ming, it seems that a compromise dis­ (1982), a movie shot nonanamorphical­ were wider than that aspect ratio. On a play aspect ratio may be desirable. lyon 35mm film and intended for pro­ 2: I display, however, side masking Using the same formula from which jection at an aspect ratio not exceeding fails for the many 2.2: 1 or 2.4: I the 16:9 compromise aspect ratio was 1.85: I. The second highest grossing movies, which would have to be shown derived, 2: I may be seen to be just film of all time, Jurassic Park (1993), in a letterbox format if their aspect ratio under the ideal equal-area compromise was made the same way. The Variety were to be preserved.

SMPTE Journal, August 1996 467

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

Part of the dissatisfaction with 16:9 ratio said to be as narrow as fering by just I%, but, according to the may be related to the fact that the ratio 1.15:I '145.46.47While a linear compromise paper, having no preferred aspect ratio was introduced as a shoot-and-protect hetween the two new extremes remains between them." production format, and the concept of ncar 2: I (1.95: I), an equal-area com­ shoot and protect involves of promise reverts, again, to 16:9 (1.78: I). The nonaction area. It has been suggested One of the most powerful aspect that the use of 16:9 as a display format History of the Perfect Aspect ratios listed was 1.618: I, a rounded precludes the use of letterbox to pre­ Ratio version of a mathematical relationship serve the composition of material shot Since accommodation of different tech nically cal1ed the di vision in in a wider aspect ratio. That is, of aspect ratios necessarily adversely extreme and mean ratio (DEMR) but course, not true. Any of the aspect ratio affects the images involved, perhaps it more commonly referred to as the accommodation techniques described would be worth ignoring issues of com­ Golden Section." It is worth noting in this paper can be used on displays of promise between existing aspect ratios some of the many names used for this any aspect ratio. A 2: I display aspect and searching for an ideal aspect ratio. quantity, because they appear frequent­ ratio will be no more free of accommo­ In addition to references already listed, ly in the history of moving picture dation techniques than is a 16:9 dis­ there is a wealth of literature on the his­ aspect ratios. play, and the 2: I display will have to tory of widescreen movies.":" Most ref­ Names for DEMR can be created by use those techniques to a greater extent erences attribute the impetus behind the combining the adjectives continuous, on 1.33: I, 1.66: I, and 1.85: I program­ current era of widescrecn movies to divine, golden, medial, or sacred with ming than will a 16:9 display. competition with television. In other the nouns cut, mean, number, propor­ The ASC call for a specific 2: I dis­ words, the problem of showing tion, , ratio, rectangle, or sec­ play aspect ratio appears to have origi­ widescreen movies on television was tion. It has also been called simply the nated in a presentation by the cine­ intentional. A study of the literature section, the jewel ofgeometry, the mid­ matographer Vittorio Storaro at a for­ indicates some other unusual facts: dle and two ends, the proportional divi­ mats seminar conducted by the • Widescreen motion pictures arc at sion, the whirling squares, and the Technology Council of the Motion least a century old. more bizarre he who understands, Picture/Television Industries on • The impetus for many widescreen Faratra, phi. and Victoria. Dynamic January 29, 1994. Seeking standardiza­ developments had nothing to do with a symmetry, a term that has been incor­ tion on a single aspect ratio, Storaro preference for a wider aspect ratio. rectly used to identify DEMR, refers to suggested a linear compromise between • The terms wide and widescreen an aspect ratio of a rectangle that can­ HDTV at approximately 1.8: I and have not always indicated a wider not be divided into squares. While a 70mm theatrical projection at 2.2: 1.41 aspect ratio. Golden Rectangle meets the criterion of Using linear averaging rather than • Cinematographers and directors dynamic symmetry, so do rectangles equal area could be one reason to reject have not always favored aspect ratios with aspect ratios of the square roots of a 16:9 aspect ratio, but again, a linear even as wide as 2: I. 2, 3, or 5, all within the range of aspect average between the same limits of It might be useful to start at the ratios from 4:3 to 2.35: I. In contrast, 1.33: I and 2.35: I is just over 1.84: I, beginning, but it is difficult to say 4:3,3:2,5:3, 16:9, 1.85:1,2:1,2.2:1, not 2: I. Another option would be where that beginning is. Motion picture and 2.4: I do not meet the requirements changing the limits. Given the vast antecedents have been traced to ancient of dynamic symmetry. libraries of 1.33: I programming, it Rome55,56 and even curlier." Since The principle of DEMR is quite sim­ seems unreasonable not to consider that aspect ratios are as old as rectangular ple. A line is cut in such a way that the ratio, but the upper limit could be imagery, however, if there is a human ratio of the whole line to the larger sec­ extended to the 2.75: I aspect ratio of preference for a particular aspect ratio, tion is the same as that of the larger such movies as the 1965 epic, The that preference may be considerably section to the smaller section. This may Greatest Story Ever Told (because such older than motion pictures. be expressed mathematically as movies were meant to be seen theatri­ A technical paper in 1931 traced an x/y = (x+y)/x (2) cally on curved screens, it is difficult to indication of aspect ratio preference to attribute a particular aspect ratio to an Egyptian papyrus document dated to where x is the larger section of a line them; the chord and the arc of the 4750 B.C.;5H that paper was referenced and y is the smaller section. If the screen will yield different figures for (indirectly) in a debate about the appro­ whole line is said to have a unit length, width)." An equal-area compromise priate aspect ratio for advanced televi­ then between those aspect ratios would be sion that took place in 1940.59 Another x+y= 1 (3) just over 1.91: I; a linear compromise technical paper, this time referenced would be just over 2.04: I. directly in the same debate, listed 16 and a quadratic equation is the result: If the few movies with 2.75: I aspect especially "powerful aspect ratios" ratios are to be considered, however, between 1.236: I and 3.618: I in addi­ x'+x-J = 0 (4) what about those created in the years tion to some others that were merely If the smaller section becomes the between the advent of the sound track powerful. Both 1.309: I and 1.809: I fell height of a rectangle and the larger the and the institution of the Academy into the most powerful category; so did width, the resulting rectangle has an aperture in 1932, films with an aspect 2.4472: I and 2.472: I. aspect ratios dif- aspect ratio of approximately 1.618: I

468 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

(the absolute value of one solution of seem to show a preference for an aspect used by a number of directors, includ­ the equation); if the opposite is done, ratio in that range for still pictures." For ing and , the resulting rectangle has an aspect a "Wide Film" symposium conducted is not commonly used ratio of approximately 0.618: I (the by the Technicians, Producers, and today. other solution). Both shapes are Golden Directors branches of AMPAS on The previously mentioned 1994 Rectangles. September 17, 1930, the Academy's Technology Council formats seminar The mathematical principle of assistant secretary distributed a memo­ offers anecdotal evidence of aspect DEMR has been known for centuries. randum stating that "Howell and ratio preference for moving pictures. In the 19th century, Gustav Fechner Dubray, Lane, Westerberg, and As reported in International Photo-gra­ conducted experiments to find out Dieterich agree that the most desirable pher, "The votes were consistent. The whether there was a most preferred proportions are those approximating audience always preferred the widest aspect ratio, and his results seemed to 1.618: I, which correspond to those of format with the largest image area.,,71 show a preference in the vicinity of the the so-called 'whirling square' rectan­ The largest image area presented, how­ Golden Ratio." A noted physicist and gle (also known as the Golden Cut), ever, was the 70mm format at 2.2: I, sti mulus/sensation investigator, based on the principles of dynamic while the widest was anamorphic Fechner is considered a pioneer of psy­ symmetry which have predominated in 35mm at 2.4: I. Thus, the largest image chophysics and contributed much to the the arts for centuries." area was not the widest; yet, according technologies of both film and video, The director Sergei Eisenstein to that report and others, the largest was including the principle that, within cer­ responded in a speech at the meeting preferred. tain limits, the intensity of visual stimu­ that '''Predomination in the arts for cen­ The test did not include IMAX, with lation increases as the of the turies' should in itself be a cause for the an image area much larger than any­ stimulus (a principle reiterated fre­ profoundest suspicion when application thing tested but one of the narrowest quently in the technical literature)," is considered to an entirely and basical­ aspect ratios (1.43: I ).72 From IMAX After Fechner's publication of a ly new form of art, such as the and other formats, there is anecdotal seemingly preferred aspect ratio, what youngest art, the art of cinema." evidence that viewers may prefer nar­ appeared to be evidence of that ratio Eisenstein went on to point out that cin­ rower aspect ratios when they are pre­ was said to be found in works of art ema is based on dynamics." sented on screens very much larger dating back to ancient times, and such It is easy to see why a dynamic than those of wider aspect ratios. reports appeared (and continue to image medium may elicit different In a staged event held at Radio City appear) in the literature of aesthetics, aspect ratio preferences from those of a Music Hall in April 1954, Paramount architecture, art, mathematics, percep­ static image medium. A photograph of was able to demonstrate its relatively tion, and psychology, e.g., "Much evi­ a skyscraper may be appropriately narrower aspect ratio VistaVision for­ dence of the conscious use of the pro­ framed in a vertical image format, mat very favorably by comparing it portions of Golden Rectangles can be while one of a python is more appro­ with CinemaScope's wider aspect ratio found in early Greek art and architec­ priately framed horizontally. In a projected on a smaller area.' (It's ture."?' Even one of the ATV systems dynamic medium, however, a horizon­ impossible to assign a specific aspect proposed to the Federal Communi­ tal format can tilt down from the tip of ratio to VistaVision because Para­ cations Commission selected the the skyscraper to its base; the vertical mount allowed "a great deal of latitude Golden Section as its aspect ratio." format can pan the python from tip of with respect to aspect ratio. Our pic­ There appears to be a similarly large tongue to tip of tail. Furthermore, a tures can be played in anything from 4 body of literature debunking the character may walk into or across a to 3 up to 2 to I in aspect ratio.")" Golden Ratio as an aesthetic prefer­ frame or may rise from a chair or Much later, the author's contemporary ence, however. One researcher repeated descend stairs. It would seem impor­ report of a demonstration of the nar­ Fechner's experiments and found that tant, therefore, to study aspect ratio rower aspect ratio FuturVision 360 film the supposed preference appeared to be preferences specifically for moving format at the SMPTE convention on an artifact of the experimental tech­ image media; unfortunately, it is diffi­ October 28, 1986, stated, "As the nique.?" Another found that any shape cult to find such studies. FuturVision screen is lowered after the even vaguely near the Golden Ratio demonstration, the normal, wide theater had been considered to be evidence of Static vs. Dynamic Image Aspect screen behind it looks as tiny as a tele­ its use; he nevertheless acknowledged Ratio Preferences vision set."74 that the raw data "would support It has been stated that there is a pref­ A more formal study found a clear hypotheses that suggest that preferred erence for wider aspect ratios in mov­ preference for 16:9 moving images rectangles often have ratios associated ing image media, even if that means over 4:3, even when the 16:9 images with a spread of values containing sacrificing resolution." A classic case is are smaller." Unfortunately, only those points from the interval between 0.6 the Techniscope , developed two aspect ratios were tested, so while and 0.7" [horizontally oriented ratios by Italiana in 1960, essen­ the study may show a preference for between 1.43: I and 1.67: I].67 tially dividing a standard film frame widescreen imagery, it docs not neces­ Indeed, there have been numerous into two much wider aspect ratio sarily identify the preferred aspect experiments performed with different frames, thereby losing half the avail­ ratio. It is also possible that the pro­ techniques at different locations, and all able vertical resolution.'>"> Although gramming selected affected the out-

SMPTE Journal, August 1996 469

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL come. The movies chosen were all said neers, and technicians in 1930 was held museum artworks, was used by to have been selected partly on the to determine the best action to take on Paramount's Lorenzo del Riccio to jus­ basis of their having been shot with aspect ratio following the introduction tify the creation of a 1.85:1 aspect both theatrical presentation and televi­ of the sound track. The 4:3 35mm ratio." The differences between the two sion in mind. Thus, shoot and protect frame, essentially unchanged since its studies may be related to the artworks was used, with key action likely to be 1889 introduction in the Edison selected and/or the measurement tech­ kept within the confines of the wider Kinetoscope, was suddenly narrowed niques used. The inclusion of picture aspect ratio. The preference shown for by the addition of a sound track. At frames results in a narrower aspect the wider imagery may have been a approximately the same time, numer­ ratio, as shown in Fig. 8. preference for less fluff in the frame; it ous widescreen film techniques were Another paper in the January 1930 is also conceivable that it was a prefer­ being tried. Virtually the entirety of the Journal was from the Bell & Howell ence for something different from ordi­ January 1930 issue of the Journal of Camera Co. and suggested three differ­ nary television. the SMPE was devoted to the topic of ent film widths, all with a 5:3 aspect An unpublished study conducted for aspect ratio. No one, it seemed, liked ratio." That proposal was particularly Philips using moving images found that the newer, squarer ratio formed by the significant coming from an organiza­ aspect ratio viewing preference was sound track, and this seemed an oppor­ tion that previously "had an ironclad influenced slightly by the originally tune time to change it to something company policy to refuse to manufac­ intended aspect ratio. It was also influ­ even wider than 4:3. ture, modify, or repair any cinema­ enced slightly by viewer habit (TV Communication No. 410 from the chine not of the standard 35 mm viewers who saw few movies preferred Kodak Research Laboratories was gauge.'?' narrower aspect ratios; movie goers reprinted in the Journal as "Rectangle That 5:3 ratio was also referred to as who watched little TV preferred wider Proportions In Pictorial Composition." the Golden Rule, a fact explained by aspect ratios) and by viewing angle." The paper came up with yet another the Academy's memorandum: "For The previously mentioned study found term for 1.618:1, "the Golden Rule," simplicity, the ratios 5:3 (which equals no relationship between aspect ratio and it performed statistical analyses on 1.667:1) or 8:5 (equaling 1.6:1) are preference and screen size and contra­ some 250 museum paintings, specifi­ generally advocated instead of dictory preferences based on viewing cally excluding those with vertically 1.618: 1."69 Part of the current aspect distance (screen size and viewing dis­ oriented aspect ratios. A frequency ratio debate seems to involve nomen­ tance are the only factors affecting curve was plotted, similar to that in clature," so it is worth pointing out that viewing angle)." A third study found a Fig. 7. cinematographers (even ASC mem­ correlation between preferred screen The thrust of the paper was to have bers) frequently referred to ratios as 5:3 sizes and viewing distances but one that provided impetus for a change in or 8:5 (or 3:5 and 5:8) at the time of the contradicts the results of the other stud­ motion picture aspect ratio, but the 1930 debates." It is true that a ratio ies." The research for this paper found average of the aspect ratios shown was relating to one provides a more imme­ no clear indication of any particular just over 1.4:I, and by far the greatest diate sense of the shape of an aspect aspect ratio preference for moving frequencies noted were in the range of ratio than does an integer ratio like 4:3, images. the presound-track 4:3 aspect ratio." 16:9, or 64:27; there is a small techni­ The AMPAS meeting of directors, Perhaps curiously, the exact same tech­ cal difference, however, between cinematographers, producers, engi- nique, averaging the aspect ratios of 1.33:I and 4:3 and an even larger dif-

Rectangle Proportions In Composition JSMPE Jan. 1930, p. 37 35 30 \ \ f \ 25 '-

20

15

10 5 ._1 •... ••• ,

',"..., . • ~". . I o . .-=1=-. -+----t---+----+---4-----. +- .0'. ··t -_ ·t- • I '-+ 1 1.05 1.1 1.15 1.2 1.25 1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8 1.85 1.9 1.95 2 Aspect Ratio

Figure 7. Aspect ratio frequency in museum paintings.

470 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

place. William Dickson, in developing for Thomas Edison the first motion pic­ ture camera to use flexible transparent film, ordered film 1.375 in. wide (almost 3 mm), because that could be obtained by slitting existing photo­ graphic down the middle. Inner Aspect Ratio 2: I Running the film vertically and perfo­ rating it dictated an image width of I in. The height of the image was not dic­ tated mechanically, however. Dickson's probable desire for the Golden Section as an aspect ratio in 1889 was tempered by another desire to work in 1/4-in. picture increments. Therefore, the first movie frame, 1 in. by 3/4 in., 4:3, was as close as he could 2 87 Outer Aspect Ratio 5:3 get to 1.618:1. • Unfortunately, some parts of the Figure 8. Picture frame effect on aspect ratio. hypothesis seem weak. Dickson did, indeed, report increasing picture size in ference between 1.66:1 and 5:3. Again, with any aspect ratio from 0.46: I to 1/4-in. increments," but his perfora­ circuit design also commonly requires 2.35: I. A short film sponsored by the tions appeared 64 times/ft, or every integer factors for or British Film Institute, The Door in the 3/16 in., proving that he worked in . Wall (1955), later also made use of increments other than 114-in. As far as Perhaps the most urgent paper in the varying aspect ratios, both horizontal picture size is concerned, Edison's first January 1930 Journal was from and vertical, in a single movie.'>" patent caveat, submitted in 1888, AMPAS. It described a situation in describes images just 1/32-in. in size; which standardization had broken 4:3 the third caveat specifies lI8-in. down, and both theater owners and The amazing result of all the aspect images." movie studios were taking matters into ratio discussions circa 1930 was the Had Dickson felt strongly enough their own hands. Nine different projec­ Academy aperture, standardized by about the Golden Section, he could eas­ tion apertures and 11 different SMPE in 1932, based on the desires of ily have masked the height of the image viewfinder reticles were noted to be in AMPAS. The projection aperture was to that ratio. If that would be consid­ use, none matching any standard, and precisely 11:8 (1.375: 1), very close to ered a waste of film, he could have, as many with different aspect ratios." the presound-track 4:3"3 (it has since has more recently been suggested.f-" The stage seemed to be set for the changed to 1.37:1).39 It was not the first made images three perforations high first major change in motion picture time the roughly 1.33: 1 aspect ratio instead of four, thus saving a great deal aspect ratio. Heads and feet were would survive a challenge, and it of film. Though it would have required sometimes being chopped off by arbi­ wouldn't be the last. It is difficult to a different design, the film could also trary projection apertures that varied explain why AMPAS and SMPE have been moved horizontally through as much as 14% from the standard. reverted to 1.375: 1, especially since the camera aperture (as in Fear's Super There were many proponents of an projection focal lengths and apertures Pictures, Glamorama, VistaVision, aspect ratio approximating the Golden had to be changed anyway. It may have Technirama, and IMAX, for example)" Section and some for aspect ratios been the case that 11:8 was the only instead of vertically, thereby removing even wider. Eisenstein's call at the shape on which agreement could be the I-in. width restriction. Academy symposium for a "Dynamic reached (both AMPAS and SMPE had Belton suggests that Dickson may Square" has been misinterpreted as a previously tentatively decided on 4:3,84 have been influenced by Ouomar lone call for square imagery. Instead, but SMPE wanted a firm decision from Anschutz,' whose 1887 animated pho­ Eisenstein wanted a square frame into producers)," and agreement on some­ tography display system used large which filmmakers could place any thing was certainly necessary, as an transparencies in a 3:4 aspect ratio (the chosen aspect ratio, whether vertical article by an ASC member in the 1931 opposite of 4:3).13 One might then ques­ or horizontal." D. W. Griffith in the Journal noted." That article called for tion why Anschutz selected 3:4. It may U.S. and Germaine Dulac in France a 4:3 aperture, despite its author's pre­ have had to do with the shape of his had previously masked images to viously expressed preference for the apparatus or with the recurring ancient highlight certain areas and, well after Golden Section." Pythagorean 3-4-5 right triangle (a loop the Academy symposium, Eisenstein's The widescreen historian John of flexible material 12 units of length dream was realized in the Soviet Belton suggests that knowledge of the long, with each unit marked, can be Union's Vario film systems, the most Golden Section helped create the 4:3 used to create a perfect right angle flexible of which, Vario-70, could deal motion picture aspect ratio in the first repeatedly, a principle that was used in

SMPTE Journal, August 1996 471

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL the construction of the pyramids). The 4:3 aspect ratio was attributed in 1940 Ahsolute to the ancient Greeks." There was a plethora of different Screen shapes for and Height prior to Dickson's 4:3, and there was a Limit Bakonv similar plethora afterward. An 1899 survey listed 89 different movie projec­ tion systems in its "Present-Day Apparatus" section, many with differ­ ent aspect ratios, then added another 56 announced systems." L------======Viewer Even though they were using 35mm film, Auguste and Louis Lumiere began with a 5:4 aspect ratio frame." For compatibility with Edison's movies, they later adopted both a 4:3 Figure 9. Balcony overhang screen height limitation. aspect ratio and the use of four perfora­ tions per side per frame. In 1898, how­ tion, camera and projector mechanical palaces, the balcony problem became a ever, when they developed the widest design, and, if the theater could accom­ serious issue. Today, however, bal­ film motion picture format (75mm), for modate it, even a larger image." conies are becoming ever more rare, special projections at the Paris Similar benefits remain true today for removing perhaps the major reason for Exhibition of 1900, they retained the larger film formats." Another driving the advent of wider aspect ratios. 4:3 aspect ratio, even though the cam­ factor for wider film width had nothing Belton notes the change in subheadings era, screen, and projector were all to do with pictures; the wider the sound of his last chapter, "The Return of the unique and needed no compatibility track, the higher the sound quality (a Nickelodeon," regarding multiplex the­ with anything else." Similarly, Max position disputed, however, at the 1930 ater complexes with small auditoriums, and Emil Skladanowsky, independent Academy symposium)." and "The Return of the Peepshow," of Edison compatibility and seemingly The term wide film was clearly regarding video.' independent of mechanical require­ defined in an Academy publication: There were also supposed economic ments, adopted a 4:3 aspect ratio for "Wide Film has a width greater than considerations pushing a wider aspect their first Bioskop projection system." the standard 35mm."9<1 By that defini­ ratio. "Though the opinions of cine­ One of the first post-Kinetoscope tion, current scope movies are not wide matographers were not canvassed, art wide-aspect-ratio systems, the Latham film. directors favored the wider frame as it Eidoloscope (1895), was developed by There are also references to wide meant they did not have to build sets as Dickson, who is said to have adopted a screens that indicate simply larger high, and production managers favored wider film, frame, and aspect ratio images, not necessarily with a wider it because it was felt the larger, clearer specifically to avoid infringing Edison aspect ratio.v":" A publication of the images would eliminate the need for patents.' Dickson may welI have sought National Association of Theatre closeups and the additional time to to avoid infringing aspects of Edison's Owners states, '''Wide-screen' became shoot them.?" Even today, it has patent claims, but none of those claims the industry watchword for an array of become necessary to point out that specified any film size or aspect ratio." filmmaking techniques and projection aspect ratio does not determine the Other film systems developed at the systems that delivered high, wide, and height of a scene being shot." time - even those using wider film­ mighty images that dwarfed the typical Similar arguments have been made did not always have an aspect ratio 16-foot by 20-foot theatre screen of the about video production in a high-defin­ wider than 4:3.lJ.~4"J4 day" [emphasis added]." ition, wider aspect ratio format - that There was a strong impetus for wider it will be possible to use fewer cameras film (but not necessarily wider aspect Why Wide and less editing. Like the impetus creat­ ratios) regardless of patent infringe­ There were, however, considerations ed by balconies, the impetus created by ment issues. That impetus was the favoring wider projected aspect ratios. any real or imagined economic benefits requirements of projection versus those Key among those was the architecture associated with widescreen production of the "peep show" Kinetoscope view­ of the auditoriums in which movies has also vanished, as the publicity ers for which Dickson had first devel­ were projected, especially the existence about 1995's record-cost widescreen oped the 4:3 35mm format. of balconies in movie theaters.~2,~3.9~.99.,oo (1.85:1) Waterworld indicates. In an era of nitrate film stock, bright­ The overhanging balconies limited ness could not be increased indefinitely sightlines from the rear of the auditori­ Visual Aspect Ratio without danger of fire; a larger frame, um, placing an absolute limit on picture Periodically, during aspect ratio therefore, meant a brighter image. A height, as shown in Fig. 9. As movie debates, allusions have been made to larger frame also offered benefits relat­ theaters changed from small, single­ the human visual system - that there cd to jitter, resolution, lens magnifica- level nickelodeons to huge, multilevel is something about it that would favor

472 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. I SMPTE TUTORIAL one aspect ratio over another (separate­ no farther than six widths.s That is a ered too wasteful of the surface area of ly from any psychological prefer­ much greater range than the difference then-round picture tubes, however. To ences). As is the case with the Golden in aspect ratios between 4:3 and 2.4: I. cope with the roundness problem, the Section, however, arguments are often Wide angular ranges can also be found committee set itself an aspect ratio limit made on both sides. One researcher in a SMPTE theatrical presentation of 1.4:I. found that the maximum visual field is manual,'?' and common theatrical prac­ In the end, having found no com­ approximately twice as wide as it is tice exceeds both BKSTS and SMPTE pelling physiological or aesthetic rea­ high," while another found it to be recommendations. Television also son to adopt a widescreen format, the only 1.6: I for the range captured by offers widely varying visual angles. NTSC selected a 4:3 aspect ratio and the eyes individually and I: I for both Though the preceding argument ren­ declared that the controlling factor was eyes together."" ders the fact irrelevant, it may be point­ that it "has all advantages found in "A widely spread opinion has it that ed out that the largest motion picture motion picture practice." The other the screen with a horizontal location, screens have always had aspect ratios cited advantage was that it "permits n with an aspect ratio of approximately less than 1.4:l. , I07,1O" motion picture scanning without 1:2 [2: I] constitutes the optimum psy­ Stimulated retinal angle is not the waste." It was a slightly curious choice, chophysiological condition. Some only shape-related aspect of vision. given that the motion picture industry authors believe that such a screen for­ Panel 2 of the National Television had changed to 11:8 (the Soviet TV mat best satisfies the requirements of a System Committee (NTSC) in 1940 set aspect ratio investigated by the NTSC) full field of view for the two stationary itself the following task as its question a decade earlier. eyes. Such a conclusion is incorrect, number 1: "Considering the shape and however, because the field of distinct nature of the binocular visual field of The Eventual Advent of vision of the eye is equal to only 2 or 3 view, can there be deduced any pre­ Widescreen degrees. It is only within this small ferred aspect ratio for television pic­ Today's problems of aspect ratio angle that the acuity of vision is tures? Are there any other theoretical accommodation might be even worse approximately 50-100%."102 bases for the selection of any particular had the NTSC met in 1929 instead of The preceding appeared in the preferred aspect ratio?'?" 1940. A technical paper published that Journal ofthe SMPTE in 1969. Earlier, The panel investigated various art year!" also tried to rationalize an aspect an article in Film Quarterly expressed forms and vision. In retinal isopters ratio for television and came to the similar views but expanded them to (intensity perception contours) an same conclusion as did the NTSC ­ include wider visual fields, all the way "aspect ratio" (a slight favoring of the that motion picture practice should be out to peripheral vision, and found that horizontal versus the vertical) between the deciding factor. Since, at the time, even the widest screens stimulate only I: I and 1.2:I was found. In color fields, sound tracks had eaten into the 4:3 a tiny portion of the visual field.'?' it was 1.3: I. Visual acuity offered the frame, the selected aspect ratio was 6:5. Another paper published in the SMPTE widest "aspect ratio" disparity, between By the time of the sound-track crisis, Journal found important contributions 1.5: I and 1.6: I (a possible reason that circa 1930, wide-aspect-ratio film tech­ to "sensation of reality" from a wide­ the poor vertical resolution due to tele­ nology was relatively advanced. All of field display, however, and that paper, vision's 2:1 interlace has not been as the techniques that would later be used in part, forms the basis for the desire much of a problem as it might other­ in the current widescreen era ­ for a wider aspect ratio for HDTV.'ll-' wise have been). An effect called the anamorphic squeezes and expansions, Whether a sensation of reality is vertical-horizontal illusion was said to wider film, masked frames, multiple valuable or not (a director/film-system favor 1.1: I, and field of fixation (said film strands - had been demonstrated, inventor recently suggested that it can to be related to eye movement) 1.2: I. sometimes used for theatrical release, actually interfere with traditional fic­ No other vision-related differences that and generally found to be technically tional filmmakingj.!" and regardless of would suggest a bias for a particular successful. how we see, the key to arguments aspect ratio were reported. Even before the Academy's stan­ about visual field is the fact that aspect The NTSC also surveyed 31 existing dardization on an 11:8 (1.375: I) aspect ratio has little or no effect on the retinal television systems around the world. ratio, however, the early era of wide angle stimulated by an image. The hori­ There were one with an 11:8 aspect film appeared to be going nowhere. zontal visual field angle is determined ratio, 19 with 4:3, 7 with 5:4, one with The earliest wide-aspect-ratio systems primarily by the display width and the 6:5, 2 with 3:4, and one with an (e.g., Eidoloscope) failed either because viewer's distance from it (there are also unspecified aspect ratio. they were technically flawed or off-axis contributions); the vertical A clear preference for a horizontally because the Motion Picture Patents Co. field is determined by the same dis­ oriented aspect ratio was expressed: dominated the industry.' As early as tance and the height of the screen. The "Since most of man's activities occur in 1913, however, it was suggested to principle is similar to that used to argue a horizontal plane, it is reasonable that exhibitors in Britain to try masking 4:3 that aspect ratio is not a determinant of there should be more freedom of frames to create a wider aspect ratio. scene width during shooting." motion horizontally than vertically." According to the article, "the result is a The BKSTS recommended theatrical For aesthetic reasons, there were propo­ better shaped picture - more artistic. seating plan has the front row no closer nents on the NTSC of an aspect ratio of The portion masked off will never be than twice the screen width and the rear the Golden Section. That was consid- missed."!" There does not appear to be

SMPTE Journal, August 1996 473

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL any evidence of mass defections from After the NTSC's standardization of reaching a low of 15,800,000 in 1971. 4:3 prior to the introduction of the U.S. television (with a 4:3 aspect ratio) Nevertheless, wide-aspect ratios, in at sound track, however. in 1941 and the end of World War II, least some versions (cropping and The I 920s saw a great deal of large­ the movie exhibition situation changed. anamorphic projection, neither of screen experimentation, each new form Average weekly atten­ which was particularly expensive for an of which was supposed to herald a new dance in 1929, when SMPE's exhibitor to implement), endured, or era. Magnascope was simply an enlarg­ Standards Committee met to discuss perhaps more precisely, thrived (more ing lens system. When dropped in front wide film, was 95 million. In 1946, expensive processes, such as three-pro­ of an ordinary projection lens, it caused right after the war, it was about 90 mil­ jector Cinerama and the multichannel the picture to double linearly in size lion, about the same as in 1930, despite sound version of Cinema Scope were both horizontally and vertically (and a growing population. By 1953, howev­ less successful). become much dimmer), retaining a 4:3 er, it had dropped to just 46 million, a Recognizing a need for revenues aspect ratio or changing (through crop­ reduction generally attributed to televi­ beyond a limited market of specially ping) to whatever size the theater archi­ sion.!" The movie industry decided to equipped theaters, producers of movies tecture would allow. It was said that it fight the audience loss by offering sen­ in some of the new systems also shot received a standing ovation when it was sations that could not be experienced the same scenes on ordinary 35mm first used." by watching television at home.!" frames, thereby eliminating aspect-ratio The Fox Grandeur system was very "From an historical point of view (and, in some cases, frame-rate) accom­ much like today's 70mm systems. Henri both the so-called 3-D - stereoscopic modation problems. Producers of ordi­ Chretien's Hypergonar anamorphic films - and wide screen pictures are nary 35mm movies, seeking to cash in lens, used in production in 1927, is, in not new, dating back as they do to the on the attraction of widescreen, faced a fact, the same lens that made earliest days of the art and industry. different problem. CinemaScope possible (it had been used However, 3-D and wide screen pictures Shane (1953), composed and intend­ to create both wider and narrower burst upon the American motion pic­ ed for viewing in a 1.375:I aspect ratio, aspect ratios, the latter by rotating the ture scene in the closing weeks of 1952 was projected instead at 1.66:I when it squeeze axis by 90°). The triptych pre­ with all the suddenness of new-found was premiered at Radio City Music sentation in Abel Gance's Napoleon comets. Each week, indeed, almost Hall, a ratio Paramount found tolerable, (1927) was in some ways a precursor of every day of 1953 was marked with an as it involved cropping just 10% from Cinerama (though it wasn't used the announcement of a new method, the top and bottom of a 4:3 image. same way). In 1929, SMPE's Standards process or scheme."51 One such (Paramount adamantly opposed projec­ Committee considered four large-frame process, Scanoscope, applied tion at any ratio greater than 2: I, even widescreen systems ranging in film CinemaScope's 2:I anamorphic princi­ for VistaVision movies, which were width from 35mm (horizontal film trav­ ples to television.!" 3-D television was composed for wider aspect ratios.)!" el, 10 perforations/frame) to 70mm and also broadcast at the time. I 17 The Band Wagon (1953) fared less well in aspect ratio from 1.84:I to 2.27:1.111 It wasn't only 3-D and widescreen in cropped exhibition, with complaints (As it has been recently suggested that that exhibitors tried. The 19th-century received about the loss of the dancing 16:9 was developed as a linear compro­ Cineorama technique of completely feet of Fred Astaire and Cyd Charisse. mise between the sound-track aperture encircling viewers with synchronized Nevertheless, cropping of existing and 2.35:1 and 1.85:1 as a compromise movie screens was revived at movies became common practice. "The between 4:3 and 2.35: 1,112.113 it is worth Disneyland in 1955. Cinerama and fact that many actors found their heads noting that 1.85: I was proposed as a Todd-AO both used higher frame rates chopped off and many dancers found preferred aspect ratio by two unrelated (26 and 30 frames/sec, respectively). that their feet were not on the screen organizations long before the existence Those systems and others used deeply didn't seem to bother the exhibitor or of 2.35: 1.) curved screens, sometimes extending the theater patron to any degree. The An article called "Wide Film" in The into the seating area. During a rockslide public was fascinated with the wide 1931 Film Daily Yearbook of Motion sequence in It Came From Outer Space screen.'?' Pictures summarized the situation suc­ (1953), some theatrical viewers were Distributors were very flexible about cinctly: "Dormant condition of the sub­ pelted with foam rocks. Vibrators aspect ratio, lest they lose the business ject is attributable to two major reasons. administered "shocks" to some seats of some exhibitors. A Universal­ First, the fact that recent-year experi­ when viewers watched The Tingler International promotional document for ments failed to convince producers that (1959), a technique recently revived in Imitation of Life (1959) informs enlarged pictures exercise a definite one of the motion picture attractions at exhibitors "Aspect ratio: any ratio up to influence at the box office. Second, the Luxor Hotel in Las Vegas (the same 2:1." gigantic costs would be involved in theater's screen has a 0.5: I or 1:2 Acceptance of cropping continues to changing the industry over to accom­ aspect ratio). Behind the Great Wall the present, regardless of the intended modate them."!" There was an econom­ (1959) was exhibited in Aromarama, or displayed aspect ratios. The most ic depression, and the industry had just featuring 72 different smells." commonly noticed form of cropping begun to accommodate sound. Wide None of these techniques was able to occurs when widescreen movies are film, and wider aspect ratios, would restore movie attendance to pre-1950 shown on television screens via the have to wait. levels. In fact, it continued to fall, truncation method. A scope movie con-

474 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL verted to a flat print for theatrical pro­ requires this format, although I don't panned, and boomed often, and the jection at between 1.66: I and 1.85: 1 think the subject gains anything.,,40 resulting shots are intercut, dissolved, also undergoes cropping, however, Lucien Ballard: "I like 1.75, 1.8, almost and inserted; there are even widescreen even though no video is involved. I? the old screen ratio best.,,'2o closeups. Ordinary 4:3 U.S. television coverage Director was per­ of the 1992 World Series baseball haps the most acerbic, referring to the The Perfect Aspect Ratio championship was shown on the 10:3 CinemaScope aspect ratio as "a system It is normal for opinions and tech­ (3.33: I) Jumbotron screen of the of photography that pictures a boa con­ niques to change with time. Standard­ Toronto Skydome to accommodate strictor to better advantage than a man." ization of a particular display shape, fans. Although the uncropped picture He also provided the adage that "no however, especially when that shape is was available free of charge on broad­ screen is larger than its smallest dimen­ imposed upon a large glass bulb, locks cast television, viewers paid to watch sion."'" in a specific preference well into the the cropped version in the stadium (on In 1994, director Stanley Kubrick future. Therefore, it is worth very care­ a giant screen but one with a small released a restored version of Dr. fully considering any proposed display visual angle due to its great distance Strange/ave or: How I Learned to Stop aspect ratio for ATVIHDTV. from viewers). Worrying and Love the Bomb (1964). IMAX was designed originally to Film Forum in New York screened the allow nine 35mm film images to Filmmakers' Acceptance of release in "the squarish 1.66: I ratio appear simultaneously on a single Widescreen Kubrick originally intended, with more screen.!" and it retains its basic non­ It is readily understandable why a detail now visible at the top and bottom widescreen camera aperture" (its pro­ filmmaker would not favor cropping. of the screen."!" As recently as 1995, jector aperture has been variously Even when cropping was not an issue, Lassally wrote, "The adoption of, say, specified, and its screens vary, too, but however, (here were initial objections 1.75: I as a universal new standard... they are usually near 4:3 and are never to wide aspect ratios among cinematog­ would in my opinion greatly benefit the even as wide as 1.66:1).72 It is an raphers and directors. industry as a whole."!" extremely popular film format.?' and Cinematographer Fred Westerberg Except for those in the preceding has recently added feature-length and actively opposed ratios as wide as 2: I paragraph, however, it has been rough­ star-cast fictionall dramatic movies. during the sound-track aspect ratio ly 25 years since the most recent of Does this indicate a trend towards nar­ debates circa 1930. During the same those sentiments was expressed, and, as rower aspect ratios in motion picture debates, cinematographer Karl Struss, the ASC's position on displays indi­ film? Should such a trend be consid­ who favored 5:3, said 2: I would result cates, there has clearly been a shift of ered? in smaller images and its lack of pro­ position. It was Stevens's Shane that HDTV is said to have a need to be portional height was problematic; and had been cropped at the beginning of interoperable with other media. The Joseph Dubray, described as a "motion the current widescreen era; he went on most common computer picture tube picture engineer and erstwhile camera­ to direct (and produce) the very wide display shape is 4:3, though such dis­ man," said that the consensus in aspect ratio (2.75:I) epic The Greatest plays vary between 1:1 and 1.5:1 (and Hollywood was that 2: I was "neither Story Ever To/d (1965). may be rotated 90 0 to create aspect pretty nor desirable.'?" More recently, Some of the unfavorable comments ratios less than I: I). In print, the cinematographer Lee Garmes said, "I may be attributed simply to a change in familiar U.S. 8-1/2 x l l-in. piece of found working in CinemaScope a hor­ traditional methods. In an article called paper has an aspect ratio of 0.77: lor, 0 ror - shallow focus, very wide angles, "New Medium - New Methods," rotated 90 , 1.29: I; its international everyone lining up, awfuL,,119 Director Jean Negulesco wrote of his counterpart, the A4 size, is 210 x 297 Other cinematographers in the same experiences with CinemaScope. mm, with an aspect ratio of 0.71: I, or, 0 1 2:1). period had somewhat more forgiving '''Writing for the new wide screen rotated 90 , 1.41:1 (2 / In a book comments. Walter Lassally: "I think should be easy,' I told my script writer. on the history of papermaking, there is 'scope is all right. I'm not mad about it 'All you have to do is put your paper in no evidence of any aspect ratio of 2: 1 personally, but it is suitable for certain the typewriter sideways.' Well, he or greater.!" Photographic aspect subjects. It's very good for outdoor didn't laugh either." ratios commonly used (ignoring verti­ subjects, Westerns, scenes of epic pro­ Henry Koster, director of the first cal orientations) range from a mini­ portions, but it's no good for intimate CinemaScope movie, The Robe (1953), mum of 1:1 to a maximum of 1.5:1, subjects." Paul Beeson: "I think if said the process made "a director at last except for rarer panoramic formats.!" you've got a very small intimate sub­ free of the camera" without having "to Here is a list of some currently used ject it's crazy doing it in ; worry about 'dolly shots' and 'pan or proposed aspect ratios for moving you're just wasting the process. shots' and 'boom shots' and all other image media displays: Panavision is really for a large canvas. camera movements." Negulesco added • Infinite. This is one way to When you're in close-up all the time that Cinema-Scope freed a director describe the cylindrical surround the­ it's very difficult to compose for from concern about cuts, dissolves, aters such as those found at Disney Panavision. There's a lot of wasted closeups, and inserts." Clearly, even amusement parks. It seems highly space on either side, but these difficul­ such favorable comments have aged; impractical for a home advanced televi­ ties can be overcome if the director today, scope cameras are dollied, sion display.

SMPTE Journal, August 1996 475

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

• 48:9 (5.33:1). This is the ratio of used a 1.78:1 (16:9) screen. In the cur­ • 4:3 (1.33:1). This is the shape of Toshiba's HD Horizon system, using rent edition of the American virtually all television programming three 16:9 projected HDTV images Cinematographer Manual (1993), and display screens, virtually all CRT­ placed end to end. The first use of the 1.8:1 is listed as the aspect ratio of a based computer display screens, and system was documentation of the proposed theatrical anamorphic projec­ many movies. As the narrowest com­ restored Michelangelo-painted ceiling tion system designed to replace the monly used or recommended aspect of the Sistine Chapel. current 2.4:1.3S There is much less dif­ ratio, it is the most efficient for the • 4:1. This ratio is commonly creat­ ference between this ratio and 16:9 manufacture of cathode-ray tubes (1:1 ed when three 4:3 images are com­ than between the Academy aperture of would be even more efficient, if such bined, as at the Geographica video the­ 1.37:1 and 1.33:1. There is also much displays were commonly used). It is ater in Washington, D.C. In the Tokyo less difference between 1.8:1 and 16:9 the longest-lived aspect ratio for mov­ Audio Visual Center Superwide­ than between 2.4:1 and 2.35:1. ing imagery and continues to be cho­ Vision system, the combination is • 16:9 (1.78:1). This is the aspect sen for recent large-format film sys­ internal to a video camera, so a single ratio of the standards SMPTE 240M tems, such as the 70mm IMAX and lens may be used. and SMPTE 260M. It has also been Dynavision systems." • 10:3 (3.33:1). This is the shape of adopted by other countries around the • Narrower than 4:3. This is the the Jumbotron display at the Toronto world for both HDTV and other forms shape of some post-sound-track, pre­ Skydome. of widescreen television. Academy-aperture movies, some com­ • 2.75:1 to 2.55:1. Some anamor­ • 1.75:1. This is a popular projection puter display screens, and some special phic film projection and most anamor­ aspect ratio in some theaters around venue films. Data Check, a manufac­ phic video projection falls within this the world. It was once called "the turer of television monitoring equip­ range, the latter because it is the result widest screen possible without changes ment, in 1995 introduced tiny 1:1 pic­ of applying a common 2:1 anamorphic in camera technique" [from that used ture-tube-based monitors on which expansion to television's 4:3 aspect for nonwidescreen movies]." even 4:3 images are displayed in a let­ ratio, resulting in 8:3 (2.67:1). • 1.66:1 (5:3). This is a popular terbox format. • 2.4:1 to 2.35:1. This is the projec­ widescreen projection aspect ratio in tion range most commonly recom­ many theaters outside the U.S. Some mended for 35mm anamorphic movies. HDTV programming has been shot in Conclusion Theaters do not always abide by rec­ this aspect ratio. This paper began with the statement ommendations. If it is accepted that • 14:9 (1.56:1). This is a very com­ that two aspect ratios are inherently this is the widest commonly found mon aspect ratio used to mitigate the incompatible and has ended with a list aspect ratio, then a display of this effects of letterbox when HDTV is of well over a dozen different aspect shape offers the benefit of allowing downconverted to non-HD TV.127 It is ratios. The techniques of aspect ratio masking for narrower images to be so commonly desired that it exists as a accommodation are equally applicable drawn in from the sides (like theatrical preset function in some aspect ratio to any. There is no clear evidence of an curtains), rather than from the sides, conversion equipment!" aesthetic or physiological reason to top, and bottom. ·16:10.7 (1.5:1). This strangely enu­ choose anyone aspect ratio over • 2.2:1. This is the recommended merated ratio (an integer ratio of 3:2), another. shape of projected 70mm movies; also called Cinema Wide, is offered by For the particular ranges of aspect again, theaters do not always abide by Pioneer in projection television ratios between 4:3 and 2.35:1 (or recommendations. receivers.!" Like 14:9, it is intended as between 1.15:1 and 2.75:1), a display • 2:1. This is the display aspect ratio a compromise ratio between HDTV shape of approximately 16:9 will proposed by the ASC. A few and non-HD TV. The method of num­ require the least aspect ratio accommo­ widescreen movies were shot in this bering the ratio appears intended to dation for both extremes of the range. aspect ratio. For comparison purposes, promote it as having even larger num­ For the specific requirement of dou­ it may be expressed as either 18:9 or bers than 16:9, lending some credence bling ITU-R Rec. 601 (720 active pix­ 16:8 (2:1 is already an integer ratio). to a complaint about the promotional elslline) resolution for HDTV, 16:9 • 1.85:1. This is the projection use of the 16:9 ratio relative to others best matches random access memory aspect ratio most commonly recom­ in press releases." As 1.5:1, this aspect (RAM) capacities. mended in the U.S. for nonanamorphic ratio is also the shape of the If those characteristics and the oth­ 35mm widescreen movies." There is VistaVision frame" and has been sug­ ers listed in this paper are considered less than 4% difference between this gested as a shape for the future.!" insignificant or become outweighed by aspect ratio and 16:9 (there is a compa­ • 1.375:1 (11:8) to 1.37:1. This is other considerations, there may no rable difference between the original the shape of almost all movies shot longer be a strong reason to choose Academy aperture of 1.375:1 and between 1933 and 1953 and many 16:9. The 16:9 aspect ratio has already 1.33:1). thereafter. It is sometimes described as been chosen, however, and is in use • 1.8:1. This ratio was selected by being 4:3 or 1.33:1 even though it dif­ around the world. The research for SMPE in 1930 on the basis on an fers from that aspect ratio by 3.2%, this paper has not found any com­ AMPAS recommendation to be used almost as much as the difference pelling reason to change any existing with wide film. For its tests, SMPE between 1.85:1 and 16:9. choice of aspect ratio.

476 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

References Committee on Advanced Television (ATV) Pictorial Composition," J. SMPE, 14:32-49, Service, Advisory Group on Creative Issues Jan. 1930. I. w. F. Schreiber, "Another Method of Aspect­ of the Planning Subcommittee of the FCC 48. R. E. Carr and R. M. Hayes, Wide Screen Ratio Conversion for Use in Receiver­ Committee on Advanced Television Service, Movies: A History and Filmography ofWide Compatible EDTV Systems," M.LT., Apr. 5, 1988. Gauge Filmmaking, McFarland & Co., Institute of Technology, ATRP-T-9i, July 24. G. S. Dibie, letter from International Jefferson, N.C., 1988. 21,1988. Photographers Guild to Advanced Television 49. M. Z. Wysotsky, Wide-Screen Cinema and 2. J. Belton, Widescreen Cinema, Harvard Univ. System Committee, Dec. 8,1988. Stereophonic Sound, trans. from Russian by Press, Cambridge, Mass., 1992. 25. J. S. Nadan and R. N. Jackson, "Signal A. E. C. York, ed. by R. Spotiswoode, Focal 3. W. Beyer, "TV Safe Action Limits for Wide Processing for Wide-Screen Television: The Press, London, 1971. Screen Films," American Cinematographer, Smart Receiver," SMPTE J.• 93:726-729, 50. "American Widescrccn" (special issue), The 43:366-367,381-384, June 1962. Aug. 1984. Velvet Light Trap. 2/: 1-80, Summer 1985. 4. M. Schubin, "Wide Not?," , 26. C. A. Poynton, "HDTV for Computer 51. J. L. Limbacher, "Width," Four Aspects ofthe 15:18-28, Oct. 1990. Workstations," Television: Merging Multiple Film. Brussel & Brussel, New York, 1969. 5. BKSTS Dictionary ofAudio-visual Terms. Technologies, pp. 197-203, SMPTE, 1990. 52. M. Quigley, Jr. (ed.), New Screen Techniques, Focal Press, London, 1983. 27. C. Rhodes and J. Lowry, "Compatibility of Quigley Publishing, New York, 1953. 6. B. Fox, ''The Digital Dawn In Europe," 1EEE Widescreen and Non-Widescreen Television 53. R. Mitchell, "History of Wide Screen Spectrum, 32:50-53, Apr. 1995. Transmissions," U.S. Patent 4,651,208, Mar. Formats," American Cinematographer, 7. M. Schubin, "On The Other Hand...." 17,1987. 68:36-41, May 1987. Yideography, 17:16-28, June 1992. 28. J. Dickson, "3 Perf Film In The Future," 54. G. Lobban, "Worth a Thousand Words: Wide­ 8. R. Hopkins, "Advanced Television Systems," International Photographer, 55: 19, Dec. screen & 3-D Film Formats,' BKSTS J., 65: I Educational Sessions, IntI. Conf. on 1984. 563-565, Oct. 1983. Consumer Electronics, Rosemont, III., June 8, 29. R. Ericson, "Three Perfs for Four," Image 55. C. F. Jenkins, "History of the Motion Picture," 1988. Technology. 69:100-110, Mar. 1987. Trans. SMPE. no. I, pp. 136-49, 1921 (no 9. K. Katsumata, et aI., "Development of Picture 30. "Film Formats," Panavision Price Catalog: month). Converting System Applying an NTSC May 2, 1994. 56. H. Hecht, Pre-Cinema History, British Film Signal to a Wide Aspect Display," IEEE 3 I. "Report of the Standards and Nomenclature Institute and Bowker-Saur, London, 1993. Trans. on Consumer Electronics. 38:303­ Committee," J. SMPE. 15:818-823, Dec. 57. R. W. Hubbell, 4000 Years of Television. G. 312, Aug. 1992. 1930. P. Putnam's Sons, New York, 1942. 10. H. Johannes, The History of the Eidophor 32. "Standard Ratio: 1.66 to 1.79?" Showmen's 58. M. M. Ditts, "The Golden Section," Bell Large Screen Television Projector, Gretag Trade Review. pp. 10-I I, May 16, 1953. Laboratories Record, 10:97-98, Nov. 1931. Aktiengesellschaft, Zurich, 1989. 33. RCA 34" G34170AT CinemaScreen 16 x 9 59. D. G. Fink, ed., Television Standards and II. G. Hewlett and W. Werner, "Analysis of WideScreen Monitor/Receiver. Thomson Practice: Selected Papers From The Electronic Cinema Projection with the Texas Consumer Electronics Inc., Indianapolis, Proceedings of the National Television Instruments Digital Micromirror Device Form CS-4728, 1994. System Committee and Its Panels, McGraw­ Display System," Moving 1mages: Meeting 34. G. Kennel, 1. Pytlak, R. Sehlin, and R. Uhlig, Hill Book Co., New York, 1943. the Challenges. pp. 435-445, SMPTE, 1995. "Major Motion-Picture Production 60. A. Van Dyck, "Dynamic Symmetry in Radio 12. W. Glenn, G. J. Dixon, and P. Bos, "High Standards," SMPTE J.• 97:985-990, Dec. Design," Proc. 1RE, 20:1481-151 I, Sept. Efficiency Light-Valve Projectors and High 1988. 1932. Efficiency Laser Light Sources," Moving 35. "TV Production Format Policy," Warner Bros. 61. R. Herz-Fischler, A Mathematical History of Images: Meeting the Challenges, pp. 446­ Technical Operations, Feb. 24,1993. Division in Extreme and Mean Ratio, Wilfrid 458, SMPTE, 1995. 36. M. Rotthaler and A. Orring, "Widescreen Laurier University Press, Waterloo, Ont. 13. B. Coe, The History ofMovie Photography, Television Systems," 1mage Technology. Canada, 1987. Eastview Editions, Westfield, N.J., 1981. 74:9-12, Jan./Feb. 1992. 62. G. T. Fechner, Yorschule clerAesthetik, vol. I, 14. L. J. Roberts, "Aspect Ratios and the 37. M. Schubin, "Politics, Numerology, and Breithopfund Hartel, Leipzig, 1876. Cinematographer," International Photogra­ HDTV," Yldeography, 15:27-30, Apr. 1990. 63. C. A. Poynton, "Gamma and Its Disguises," pher, 55:17,22-23,25, Mar. 1984. 38. "Cinematographic Systems," American SMPTEJ.• 102:1099-II08,Dec. 1993. 15. K. Powers, "Framing the Camera Image for Cinemarographer Manual. ASC Press, 64. G. E. Runion, The Golden. Section lIlld Related Aspect-Ratio Conversions," NAB Broadcast Hollywood, 7th ed., pp. I-12, 1993. Curiosa. Scott, Foresman and Co., 1972. Engineering Conf., Las Vegas, March 21, 39. "HDTV Aspect Ratio," Image Technology, 65. "An HDTV Proposal from Faroudja 1994. 77:70-76, Mar. 1995 (Widescreen Issue). Laboratories, lnc.,' Faroudja Labs, submis­ 16. R. 1. Stumpf, "Forging a HDTV System for 40. R. Campbell, ed., Practical Motion Picture sion to FCC, Mar. 10, 1988. Production and Post Production -A Photography, A. Zwemmer Ltd., London, 66. M. Godkewitsch, "The 'Golden Section'; An Working Group Report," Components of the 1970. Artifact of Stimulus Range and Measures of Future. pp. 370-384, SMPTE, 1985. 41. I. Blonder, "Aspect Ratio: Cost vs. Art," Preference," American J. Psychology, 17. G. Lobban, "Guide To 35 mm Release Print Communications Technology, 2:12, Feb. 87:269-277, Mar-June 1974. Formats," supp. to 1mage Technology, 72, 1986. June 1990. 67. R. Fischler, "On the Application of the Golden 42. L. Maltin, ed., Leonard Maltin's Movie & 18. British standard BS 2784, 1956. Ratio in the Visual Arts," pp. 349-351, Video Guide. 1996 Ed.• Signet, New York, Autumn 1981. 19. R. Hummel, "Pros and Cons of 1.85, 2.35 and 1995. 68. B. L. Jones, "Aspect Ratio Preference Study," Film Formats," American 43. D. Heuring, "Formats Seminar: Experts Cinematographer Manual, ASC Press, Report 9/84, CBS Technology Center, Nov. Debate Pros and Cons," American 20,1984. Hollywood, 7th ed., pp. 13-28, 1993. Cinematographer, 75:24-32, June 1994. 20. H. L. Vogel, Entertainment Industry 69. "The Dynamic Square," Close Up. 8:3-14, 44. G. Lobban, "Reflections On Deep Curvature Mar. 1931. Economics, Cambridge University Press, Screens," Cinema Technology. 8:56-59, July Cambridge, England, 3rd ed., 1994. 70. H. Mathias, "Image Quality from a Non­ 1995. 21. R. Mitchell, "The Origins of Ratios," Engineering Viewpoint," SMPTE J., 93:712­ 45. B. Salt, Film Style and Technology: History 716, Aug. 1984. American Cinematographer. 75:8-10, May and Analysis, Starword, London, 2nd ed., 1994. 71. B. Fisher, "Choice of Film Formats is as Old 1992. 22. J. Hora, "Past is Prologue," American as the History of the Movies," International 46. V. Nilsen, trans. by Stephen Garry, The Cillelll£ltographer. 76:110-119, Apr. 1995. Photographer, 66:27, Mar. 1995. Cinema As Graphic Art, Ncwnes, Ltd., 23. J. Hindman, "Creative Issues Affecting 72. W. C. Shaw and J. C. Douglas, "IMAX and London, 1937. Terrestrial Transmission of Advanced OMNIMAX Theatre Design," SMPTE J.• 47. L. A. Jones, "Rectangle Proportions in Television Systems," FCC Advisory 92:284-290, Mar. 1983.

SMPTE Journal, August 1996 477

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply. SMPTE TUTORIAL

73. L. L. Ryder. "Economic Aspects of Utilizing 94. A. Reeves, "Wide Film," International 117. M. Schubin, "The Other Stereo, Part II, in New Engineering Development s," J. SMPTE, Photographer, 1:8-9, June 1929. 3D," Yidcography, 12:104-107, Mar. 1987. 65:80-84, Feb. 1956. 95. K. MacGowan, Behilld The Screen: The 118 . L. L. Ryder, "VistaVi sion, the New 74. M. Schubin, "SMPTE Convention: One Histo ry and Techniques of the Motion Paramount System," Motion Picture Man's View," Yldeography, 12:31-33 , Jan. Picture, Delacortc Press, New York, 1965. Herald: 22, Apr. 10, 1954. 1987. 96. L. Cowan, cd., Recording Sound for Motion 119. C. Higham, Hollywood Cameramen: Sources 75. K. Pills and N. Hurst,"How Much Do People Pictures, McGraw-Hili Book Co., New York, of Light, Indiana Uni versity Press, Prefer Widescreen (16x9) 10 Standard NTSC 1931. Bloomington, 1970. (4x3)?," IEEE Trans. Consumer Electronics, 97. R. H. Heacock, "W ide Screens in Drive- in 120 . L. Maltin, Behind the Camera, The 35:160 -169, Aug. 1989. Theaters," J. SMPTE, 64:86-87, Feb. 1955. Cinematographer's Art, Signet, New York, 76. Personal conversation with W. E. Glenn, 98. B. Stones, America Goes to the Movies: 100 1971. Communications Technology Center, Florida Years ofMotion Picture Exhibition, National 121. "Still Scary, Now In Mint Condition," The Atlantic University, Boca Raton, Aug. 17, Association of Theatre Owners, N. New York Times, CI4, Nov. 4, 1994. 1995. Hollywood,1993. 122. W. Lassally, "Letters to the Editor," BKSTS 77. M. Ardito and M. Gunetti, "The Impact of 99. F. M. Falge, "Motion Picture Screens - Their Images, 4:2, Sept. 1995. Display Parameters on the Quality Perceived Selection and Use for Best Picture 123. G. G. Graham, Canadian Film Technology, by the Viewers," IEEE 1995 lntl. Conf. 011 Presentation," J. SMPE, 17:343-362, Sept. Associated University Presses, London, Consumer Electronics, Digest of Technical 1931. 1989. Papers, pp. 112-113, June 1995. 100.1. lzod, Hollywood ami the Box Office 1895­ 124. M. Schubin, "Commonplace vs. Common 78. J. Lasky with D. Weldon, I Blow My OWII 1986, Macmillan Press, New York, 1988. Place." Yideography, 20:22-28,122, Feb. Hom, Doubleday & Co., Garden City, N.Y., 101. W. Szabo, "Some Comments on the Design 1995. 1957. of Large-Screen Motion-Picture Theate rs,' 125. D. Hunter, Papcrmaking: The History and 79. A. S. Howell and J. A. Dubray,"So me SMPTEJ., 85: 159-163, Mar. 1976. Technique of all Ancient Craft, Alfred A. Practical Aspect s of and Recommendations 102. V. G. Komar, "Recent Work on Varios-copic Knopf, New York, 2nd ed., 1967. on Wide Film Standards," J. SMPE, 14:59­ ," J. SMPTE, 78: 851-857, 126. J. Hedgecoc, The Book of Photography, 84, Jan. 1930. Oct. 1969. Alfred A. Knopf, New York, 1980. 80. M. Ehrenberg and L. J. Roberts, "Seventy­ 103. E. Callcnbach, "Optometrical Criticism," 127. C. R. Caillouet and C. Pantuso, "Woodstock Five Years of Motion Picture Standards: Film Quarterly, 16:25, Summer 1963. ' 94 - A Live Multi-Format Production," Contributions of the Bell & Howell ce.. 104. T. Harada, H. Sakata, and H. Ku sata, NAB/ITS Advanced Teleproduction SMPTE J.. 92:1058-1065, Oct. 1983. "Psychophysical Analysis of the 'Sensation Conference, LasVegas, Apr. 10, 1995. 8 t. F. Westerberg, "The Academy's Sym­ of Reality' Induced by a Visual Wide-Field 128. Snell & Wilcox ProductGuide 1995/6, Snell posium," International Photographer, 2: 14­ Display," SMPTE J., 89:560-569, Aug. & Wilcox, Hampshire, U.K., 1995. 15, Oct. 1930. 1980. 129. Pioneer 1995 Audio/video Catalog, 82. L. Cowan, "Camera and Projector Apertures 105. B. Fisher and M. Rhea, "Interview: Doug Pioneer Electronics (USA) Inc., Long in Relation to Sound-on-Film Pictures," J. Trumbull and Richard Yu ricich, ASC," Beach, Calif., Dec. 1994. SMPE,14:109-12I,Jan.1930. American Cinematographer, 75:55-66, 130. R. A. Strain,"The Shape of Screens to Come," SMPTE J., 97:560-567, July 1988. 83. "Report of the Committee on Standards and Aug. 1994. Nomenclature," J. SMPE, 19:477-490. Nov. 106. D. V. Kloepfel, cd., Motion-Picture 1932. Projection and Theatre Presentation 84. P. N. Sands, A Historical SlIIdy of the Manual, SMPTE, 1969. Academy of Motion Picture Arts and 107. K. MacGowan, "T he Wide Screen of Arno Press. New York, 1973. THE AUTHOR Sciences, Yesterday and Tomorrow," Quarterly of 85. F. Westerberg. "Standardization of the Picture Film, Radio, and Television, II:217-241, Aperture and the Camera Motor -A Spring 1957. Needed Development," J. SMPE, 17:395­ 108. P. Robertson, The Gulnness Book ofMovie 397, Sept. 1931. Facts & Feats, Abbeville Press, New York, 86. F. Westerberg. "Wide Film and Its 5th cd., 1993. Possibilities," International Photo grapher, 109.1. Weinberger, T. A. Smith, and G. Rodwin, 1:12-13,Aug.1929. "The Selection of Standards for 87. J. Belton, "The Origin s of 35mm Film as a Commere ial Radio Television," Proc. IRE. Standard," SMPTE J., 99 :652-661 , Aug. /7:1584-1587, Sept. 1929. 1990. 110. "Th nt Show's Different. Why?" Klnemato­ 88. W. K. L. Dickson, "A Brief History of the graph and Lantern Weekly, 12: 1675, Feb. Kinetograph, the Kinetoscope and the 20,1913. Kinetophonograph,' J. SMPE, 21:435-455, Ill. "Report of the Standards and Nomenclature Dec. 1933. Committee," J. SMPE, 14:122-137, Jan. 89. G.Hendrieks, The Edison Motion Picture Myth, University of California Press, 1930. 112. J. Horn, "The Chicken and the Egg," Berkeley, 1961. Mark Schubin is a technological con­ 90 . H. V. Hopwood, Living Pictures: Their American Cinematographer, 75:22, Jan . sultant in New York City. He is a mul­ 1994. History, Photo-Production and Practical tiple Emmy Award-winning produc­ Workillg, Gutenberg Press, Ltd ., London, 113 . G. Sutor Vuille, "This Is Sutoramu," 1899. American Cinematographer, 75:10-12, July tion engineer, a forensic video analyst, 91. L. J. Roberts, "Historical Review of Motion 1994. a teacher, a historian , and an interna­ Picture Standards," lntemational Photo-gra­ 114. A. W. Eddy, "Wide Film," The 1931 Film tional award-winning writer who has pher, 54:18, Nov. 1983. Dail y Yearbook of Motion Pictures, The Film Daily, New York, 13th ed., 1931. had hundreds of articles about video 92. F. P. Liesegang, trans. by H. Hecht, Dates and 115. H. L. Vogel, technology published. Schubin is a Sources, A Contribution to the History ofthe Entertainment Industry Economics, Cambridge University Press, Fellow of the SMPTE and serves on Art of Projection and 10 Ctnemato-graphy, Cambridge, Eng., Ist ed., 1986. The Magic Lantern Society of Great Britain, the Board of Editors of the SMPTE 116. S. Rosin and M. Cawein, "Wide-Screen London, 1986. Journal. Television," J. SMPTE, 66:404-406, July 93. T. A. Edison. "Kinetographic Camera," U.S. 1957. Patent 589,168, Aug. 31, 1897.

478 SMPTE Journal, August 1996

Authorized licensed use limited to: Mark Schubin. Downloaded on August 22,2016 at 17:03:04 UTC from IEEE Xplore. Restrictions apply.