Admissibility of Computer-Generated Animations and Simulations
New York Law Journal
This article revisits the issue of admissibility of computer-generated animations purporting to reconstruct a series of events or an accident. Our discussion is prompted by the Aug. 6 decision of the California Supreme Court in People v. Duenas, upholding the admission of an animation re-enacting the sequence of events in the shooting of a deputy sheriff and offering guidance on why the animation was properly received by the trial judge. The appeal raised the interesting question whether slick animations can so beguile a jury that the demonstrative imagery is unduly prejudicial for that reason.
In 2004, U.S. Eastern District of New York Judge Jack Weinstein issued his visionary opinion in Verizon Directories v. Yellow Book USA,1 favoring admissibility of computer-generated demonstrative exhibits or techniques referred to as “pedagogical devices.” Weinstein was confident that judges could control against abuses or inequities in resources and tailor proceedings so that such demonstrative materials would help jurors find the truth. This writer discussed the opinion and admissibility issues in two columns.2
Weinstein defined four categories of such demonstrative exhibits: (1) “static images” (common in courtrooms, such as tables, graphs, maps and diagrams); (2) “animations” (moving pictures in which static images are shown in rapid succession to create the illusion of motion); (3) “simulations” or “re-creations” (detailed and realistic depictions which, in the opinion of the creator, simulate the nature of the events); and (4) “computer models” (compilations of mathematical formulae and expressions integrated into a sophisticated program or series of programs which, then, are translated into graphics explicating the results). To these he added a fifth: “enhanced images,” a classification somewhere between static images and animations. While the enhanced image is for the most part static, it can be manipulated by highlighting, enlarging particular areas, presenting side-by-side, split-screen images, printed commentary or transcripts moving in tandem with the image and/or audio.
Courts had recognized them as useful trial aids but the common view was that such pedagogical devices are not evidence themselves. Weinstein forcefully advocated that pedagogical devices be admitted as evidence, subject, of course, to court control against prejudice and manipulation of the viewer’s subconscious. Since 2004, computer-generated technology has advanced by huge leaps and bounds. Advances in computer-generated graphics now portray imagery so sophisticated, so authoritative, so evocative that the pictorial details can dwarf the testimonial foundation that justifies its admissibility in the first place. Could jurors be so smitten by the informative power of the pedagogical medium that the graphics could be overvalued? Consider, for example, the special effects presented in many successful movies. Without the visual impact of such well-staged imagery, the dialogue itself often would be flat and ineffective. The viewer’s emotions are rabidly excited by the visual cues and it is the imagery punctuating the dialogue, in large part, that moves the viewer.
On Aug. 6, 2012, the California Supreme Court, in People v. Duenas,3 seems to have rejected the “beguilement” argument when it comes to the admissibility of animations illustrating what an expert said had happened. The convicted defendant argued on appeal that the animation gave the prosecution’s case an unjustified “air of technical and scientific certainty,” citing court decisions from other states that juries find such visual evidence uniquely persuasive. This part of defendant’s argument thus took issue with the “form” of the animation as a visual reenactment. That form, defendant contended, was “likely to beguile the jurors into uncritically accepting the version of events depicted in the animation.” Thus, he argued that the animation tended to give the evidence a “posture of mystic infallibility in the eyes of a jury.”4
In a nutshell, here is what happened regarding admissibility of the animation. Defendant used methamphetamine and, carrying a loaded pistol, bicycled past a Los Angeles County Sheriff’s Deputy who asked him to stop. Uttering an obscenity and gesturing with his finger, defendant continued riding. The deputy pursued in his patrol car and pulled in front of the bicycle with lights flashing. Defendant fired a shot that shattered the car’s rear window and hit the Deputy’s right hand. The officer drew his gun and began to exit the patrol car. Defendant walked around to the driver’s side and shot the deputy three times. He then fired more shots and fled. The deputy was dead. While fleeing, defendant turned and fired three more shots. Defendant was tried, convicted and sentenced to death.
The prosecution presented Carly Ward (an expert in biomechanics) and her son Parris Ward (who creates computer graphics) to reconstruct the events and the movements of the victim and shooter, and tie these to physical evidence. Following a preliminary hearing and over objection, the jury was shown a four-minute computer animation as to how the shooting occurred. Parris Ward has a background in photography, law, computer graphics and creating computer animations for use in court. He relied on a variety of sources, including police and coroner reports, photographic records, precise measurements he took at the scene, examinations of the patrol car and bulletproof vest, and personal consultations with the coroner.
Before the animation was played to the jurors, Parris Ward told them that the animation “doesn’t tell you, because it’s from computer, that this had to happen this way”; instead, it is “an illustrative tool for explaining concepts.” The trial court gave the jury the following cautionary instruction: “What you’re going to see is an animation based on a compilation of different expert opinions. This is similar to the expert using charts or diagrams to demonstrate their respective opinion. This is not a film of what actually occurred or an exact re-creation. It is only an aid to giving you a view as to the prosecution version of the events based upon particular viewpoints and based upon interpretation of the evidence.”
The animation was a series of mostly still images drawn to give the impression of three-dimensional space. Although the objects and figures in the images rarely move, the viewer’s perspective moves within the images, allowing the viewer to see the objects and figures from different angles. The figures were drawn in generic fashion, their facial features indistinct and expressionless. The viewer’s perspective shifts from where the principals were in relation to the patrol car at the various stages in the sequence. Red lines were used as the bullets’ trajectories with a number of descriptive words or headlines superimposed over the animation. Interspersed with the animation were actual autopsy photos.
The California Supreme Court observed that courts and commentators draw a distinction between computer animations and computer simulations, citing an annotation titled, “Admissibility of Computer-Generated Animation.”5 Animation is merely used “to illustrate an expert’s testimony while simulations contain scientific or physical principles requiring validation. Animations do not draw conclusions; they attempt to recreate a scene or process, thus they are treated like demonstrative aids. Computer simulations are created by entering data into computer models which analyze the data and reach a conclusion.” In other words, a computer animation is demonstrative evidence offered to help a jury understand expert testimony or other substantive evidence. A computer simulation, by contrast, is itself substantive evidence.6
Courts have compared computer animations to classic forms of demonstrative evidence such as charts and diagrams that illustrate expert testimony. The animation is admissible if it is a fair and accurate representation of the evidence to which it relates. A trial court’s decision to admit such evidence is reviewed for abuse of discretion. A computer simulation, by contrast, is admissible only after a preliminary showing that any “new scientific technique” used to develop the simulation has gained general acceptance in the relevant scientific community. In Duenas both sides agreed the contested evidence was an animation, not a simulation.
The animated reenactment was relevant to the question of defendant’s premeditation and deliberation and illustrated the theory of the prosecution’s biomechanics expert, Carly Ward, that defendant fired a series of shots from different locations, including one at close range. The California Supreme Court rejected the defense argument that the animation was speculative. Whatever uncertainty might exist as to the actual facts, the animation accurately illustrated the opinions of the prosecution’s experts. The trial court’s instruction to the jury that this was not a film “of what actually occurred or an exact re-creation” but only a “view as to the prosecution’s version of the events,” clarified the point to the jury.
The court also rejected the argument that the animation gave the prosecution’s case an unjustified “air of technical and scientific certainty.” The animation was not likely to beguile the jurors into uncritically accepting the prosecution’s version of the events because the animation creator (Parris Ward), the court and the prosecutor all made clear to the jury that the animated film was not an exact re-creation of what actually occurred. Accordingly, the trial court did not abuse its discretion in allowing the demonstrative evidence. An Oklahoma appellate case, relied on by defendant, that held four computer animations to be prejudicial,7 was distinguished because the evidence in that case did not adequately support the conclusions reached by the expert and depicted in the animations. Further, the Oklahoma trial court had failed to give an instruction informing the jury about how it should understand and evaluate the animations.
In complex civil litigation, such as a high-stakes products liability trial, both sides often are armed to the teeth with a variety of demonstrative pedagogicals, animations, simulations or accident reconstructions. Frequently, the computer-generated simulation is depicted in some animated format. In such cases, an arbitrary distinction between a simulation and an animation may be misleading. Such a work product is in reality a computer simulation but is depicted, or perhaps masquerades, as a mere animation. The end product to be shown to the jury is not merely an illustration of the expert’s opinion (akin to a diagram). Computer-generated simulations are subject to the GIGO (“garbage-in, garbage-out”) maxim. Faulty factual inputs can skew the results and yield “garbage” and “junk science” opinions, no matter how distinguished-looking or sounding the expert may be, and no matter how impressive the computer program may appear to be.
Indeed, some computer-generating reconstruction experts may start out with an opinion they were hired to profess and then perform “trial and error” computer runs, alternating factual inputs until they achieve the desired result. They may take one witness’ version of events, for example, and input that data, discounting or ignoring other countervailing versions of the facts observed by others. That type of selective use of limited facts is not only subject to cross-examination and impeachment. The gambit may make the simulation, the computer-generated result and the animation purporting to depict the re-enactment “unreliable.” That could justify a “Daubert” threshold inquiry or a motion in limine on whether the computer-generated animation should ever see the light of day.8
Thus, counsel need to be alert to what is being peddled as a mere animation. Diligent pre-trial discovery of all participants in the production of a computer-generated reconstruction (whether labeled a simulation or animation) is necessary. Minimally, such information is needed to cross-examine, to prepare one’s own version of an animation and, even, to frame a proposed curative instruction that the court should give to the jury if it is disposed to allow the evidence. Do the computer “runs” generating the imagery properly and reliably reflect the actual facts? Does the imagery in the animation depart from the facts? What critical factual inputs have been neglected or ignored? How would omitted facts alter the result and the image depictions? A slew of other “checklist” questions ought to be developed.
The impressive power of computer-generated pedagogicals to win the hearts and minds of the modern, “techie”-oriented jury is inexorable and undeniable. The dramatic influence of pictorial imagery in our lives outside court mirrors the potential influence that such types of evidence can have inside the courtroom. We are not speaking of mere “trial by cartoon.” Some modern animations don’t even move the figures, as in Duenas; rather, they move the viewers’ perspective within and around the images, allowing views from different angles. The juror is being put in the driver’s seat, so to speak. That may not just amount to depiction. Subconsciously, to some extent, the juror may “experience” the reconstructed event. The California decision shows that courts will not obstruct such technology so long as jurors are properly instructed. Thus, it falls upon counsel to perfect his or her advocacy, pro or con, regarding such mesmerizing evidence.
Michael Hoenig is a member of Herzfeld & Rubin.
- 331 F.Supp.2d 136 (E.D.N.Y. 2004).
- Hoenig, “Computer-Generated ‘Pedagogical’ Devices; Admissible or Not, ,” New York Law Journal, Nov. 8, 2004, p. 3; “More on Computer-Generated ‘Pedagogical’ Devices,” NYLJ, Dec. 13, 2004, p. 3.
- 2012 Cal. LEXIS 7251; 55 Cal. 4th 1; 2012 WL 3155944 (Cal. Sup. Ct. Aug. 6, 2012).
- People v. Duenas, 2012 Cal. LEXIS 7251, at *43—*44.
- People v. Duenas, 2012 Cal. LEXIS 7251, at *35—*36 (citing, Annot. 111 A.L.R. 5th 529, 538-539, §2[a] (2003)).
- People v. Duenas, 2012 Cal. LEXIS 7251, at *36 (citing cases).
- Dunkle v. State,139 P. 3d 228, 249—251 (Okla. CR. 2006).
- Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993).