Challenges in Appellate Review of Video- and Audio-Recorded Trial Evidence

Video and audio don’t always speak for themselves. In appellate courts, the significance and weight of recorded evidence often will turn on perception as much as precedent.

The proliferation of video and audio recordings has dramatically changed our world.¹ More than 100 million surveillance cameras are continually filming people and places in the United States, and their number will nearly double by 2027.² In addition, smartphones, doorbell cameras, dashboard cameras, police body-worn cameras, and other devices are silently recording our words and deeds each day.

This proliferation has generated an abundance of recorded evidence commonly presented in trials — largely in criminal cases but also in civil matters. Recordings of crime scenes, police chases, arrests, custodial interrogations, car crashes, slips and falls, workplace accidents, and other relevant events are shown to judges and juries with regularity. Between 2009 and 2023, references in court opinions to “video evidence” jumped by a factor of nine.³

The pervasive admission of video and audio evidence at the trial level has also enlarged the record on appeals. In my home state of New Jersey, the intermediate appellate court on which I serve experienced a tenfold increase in video exhibits in criminal appeals in the five years from 2016–2017 to 2021–2022.⁴ Other state appellate courts have likewise reported increases in video submissions.⁵

This explosion of video evidence has greatly altered the work of appellate courts. Drawing on a tradition of admitting images as evidence that dates back to the 19th century, recorded evidence today offers both benefits and operational challenges. Appellate opinions confirm what we sense intuitively: Recorded evidence is often in the eye of the beholder. This article explores those challenges and concludes with a discussion of factors affecting standards of review of recorded evidence, as well as considerations for the future.

Historical Uses of Recorded Evidence in the Courts

As early as before the Civil War, the U.S. Supreme Court and other courts were considering the evidential impact of still photographs.⁶ Such pictures, “worth a thousand words,” could help prove or disprove critical facts. By the early 1900s, another technology — sound recording — began to be used in court cases.⁷ By the 1920s, visual recordings called “motion pictures” had been presented as trial evidence.⁸ As technology developed in the ensuing decades, audio tracks accompanied the video footage.⁹ In some instances, the videos presented in court proved to be historic and had widespread social or political impact.¹⁰

Generally, the same basic principles of evidence law have governed the admission of recordings in the trial courts: authentication (especially now with images and sounds possibly created with artificial intelligence), relevance, undue prejudice, character proof limits, and so forth. Those basic principles extend to the appellate process. Advocates on appeal may argue that video or audio evidence should have been excluded or admitted by the trial court. They may also dispute the extent to which the recorded footage supports or contradicts key facts. In the past, appellate judges have only occasionally deemed it necessary to review recorded evidence to adjudicate an appeal. But the occasional has now become the frequent.¹¹

Pros and Cons of Recorded Evidence

Recorded evidence can have many benefits. Recordings can aid the search for truth by sometimes providing more definitive and objective evidence than the recollections and testimony of witnesses.¹² As a dynamic medium, video — with or without sound — can be more enlightening and impactful than static photos. (Consider how we may be more affected by a video shared with us than by a snapshot.) Furthermore, subject to court-imposed limitations to assure trial fairness, video and audio recordings can be replayed, paused, slowed down, and otherwise displayed in a manner that assists the trier of fact.¹³

On the other hand, video and audio recordings can have many drawbacks, such as poor lighting, poor sound quality, distance from the subject, deficient or biased camera angles, and unidentifiable persons and voices.¹⁴ The footage may have breaks or gaps, or it may omit important events that occurred before or after the recording. The recordings may also be unduly prejudicial and unfairly overshadow witness testimony and other evidence in the case.

Appellate judges must weigh these pros and cons as they sift through video evidence from the trial record, remaining alert to the risks of conscious or unconscious biases that may skew their interpretation.¹⁵

Operational Problems Appellate Courts Have With Recorded Evidence

Apart from handling the growing volume of recorded evidence within trial records, appellate courts are coping with a host of operational problems often presented by such evidence. Since the technology used in creating and storing recorded evidence widely varies, software issues may pose barriers to access for appellate judges and staff, as well as adversaries. Footage may also involve uncertain persons, objects, and sources. Time counters displayed on the video screens may not accurately reflect the actual times. Multiple recordings of the same events from different camera angles may be duplicative and time-consuming to review. The parties’ briefs and trial court decisions may fail to cross-reference the critical spots in recordings. The audio portions of video recordings may not be transcribed by the parties, and, when they are, the transcripts may be replete with “indiscernibles,” misidentified or unknown speakers, and other mistakes.

These sorts of operational snags can cause headaches for both appellate judges and lawyers. To abate the problem, appellate courts should consider adopting uniform rules for the submission of recorded evidence and the manner in which it is cited in briefs.¹⁶ Judges may also wish to delegate to law clerks or staff attorneys the initial review of video evidence, under their careful supervision.

Disagreement Among Appellate Jurists in Assessing the Recordings

Individual appellate jurists may strongly disagree about what recorded evidence does or does not show. An early prominent example of this was the U.S. Court of Appeals for the Second Circuit’s 2-1 opinion in Arnstein v. Porter, a 1946 copyright infringement case in which the judges listened to audio recordings of music written by the plaintiff and compared it with songs by the defendant, famous composer Cole Porter.¹⁷ Judges Jerome Frank and Learned Hand concluded the recordings were sufficiently similar to enable a jury to find infringement.¹⁸ In dissent, Judge Charles Clark noted, “after repeated hearings of the records, I could not find therein what my brothers found.”¹⁹

More recently in 2007, the justices of the U.S. Supreme Court were sharply divided in perceptions of a dashcam video in Scott v. Harris.²⁰ The video footage, which is now posted on the Court’s official website, depicts a lengthy nighttime police chase of a fleeing suspect on a highway, which ended when a squad car rammed the suspect’s vehicle.²¹ The pivotal legal issue in the Section 1983 case was whether the suspect had been driving so dangerously as to justify the police conduct.²² The district and circuit courts viewed the video footage and ruled the dangerousness of the suspect’s driving behavior was sufficiently debatable to raise a material issue to warrant a civil jury trial.²³ Writing for the 8-1 majority, Justice Antonin Scalia reversed the lower courts.²⁴ He found the video resembled “a Hollywood-style car chase of the most frightening sort, placing police officers and innocent bystanders alike at great risk of serious injury.”²⁵ But Justice John Paul Stevens disagreed, finding the video showed that the suspect generally drove in a safe manner, using his turn signal and waiting for cars in the other direction to pass.²⁶ Interestingly, two of the opinions in Scott v. Harris encouraged readers to view the video evidence themselves and draw their own conclusions.²⁷

Arnstein v. Porter and Scott v. Harris aptly illustrate that sometimes the import of video evidence can be in the eye of the appellate beholder. This brings me to my last point: What standards should govern appellate review of recorded evidence?

Standards of Review

Although case law is still developing, many states have adopted an approach that appellate courts should not reverse trial court decisions unless video or audio evidence clearly contradicts the trial court’s factual findings.²⁸ This deferential approach to the “referee’s call on the field” replicates the standards used in many sports: The call should be upheld unless the video review conclusively reveals error.²⁹ This approach promotes stability within the court system and, at the same time, discourages parties from making untenable factual contentions on appeal.

Apart from this conclusiveness factor, I suggest the standard of review should also take into consideration two other factors: (1) the materiality of the apparent error revealed by the recording and (2) the procedural context. For instance, imagine that a trial judge in a suppression hearing mistakenly found that a police officer credibly testified that she approached the passenger side of a stopped car with her gun holstered, while the video provided on appeal shows that the officer actually approached the driver’s side with her gun pointed at the driver. Does that error mandate an automatic reversal or remand? It depends. We must consider how material — if at all — the error was to the legal issue on appeal. If the issue is whether the driver was intimidated by the officer into consenting to a search of the car, the error could be deemed material. Conversely, if the legal issue on appeal was whether the police had a valid basis to stop the car after it was observed speeding and going through red lights, then the post-stop contradiction unearthed by the video is of scant importance.

Context also matters. If a video clearly contradicts a key assumption of a civil motion judge’s decision that granted summary judgment, there’s a strong basis for vacating the decision. On the other hand, if the video was simply one piece of evidence presented to a jury at a trial, and if the other evidence of the defendant’s guilt was overwhelming, the fact that the video contradicted only one aspect of one witness’s testimony may be inconsequential or harmless.

Takeaways for the Future

Appellate courts are being deluged with video and audio recordings that are now routinely part of the trial court record. To cope with that surge, they should be better equipped to handle the operational burdens that come with such evidence — perhaps through the adoption of court rules tightly regulating those submissions. Appellate judges also must be sensitive to the advantages and limitations of such proof, and cautious in overturning trial court decisions based on their personal perceptions upon viewing and listening to the recordings.

By establishing and following appropriate standards of review, appellate courts can make prudent use of recorded evidence, thereby enhancing the quality of their own decisions. Appellate judges today must be more than careful readers and thinkers — they need to be keen watchers and listeners, too.

---

JACK M. SABATINO has served on New Jersey’s Appellate Division since 2005 and was a temporary appointee to the New Jersey Supreme Court. A former law professor, trial judge, and litigator, he holds an LLM in Judicial Studies from Duke Law.

---

¹ This article is largely based upon a longer analysis presented in Judge Jack M. Sabatino, The Appellate Digital Deluge: Addressing Challenges for Appellate Review Posed by the Rising Tide of Video and Audio Recording Evidence, 96 Temp. L. Rev. 11 (2023). The research was performed in connection with the author’s participation in the Masters of Judicial Studies LLM program at Duke Law School.

2 Statista Consumer Market Insights, “Number of households with smart security cameras worldwide from 2016 to 2027,” Statista (Feb. 28, 2025), http://www.statista.com/forecasts/1301193/worldwide-smart-security-camera-homes.

3 Specifically, a Boolean search on Westlaw of all federal and state court cases mentioning “video evidence” for 2009 revealed 140 total; in 2023, that number sharply rose to 1,237. The raw number of cases in which video — or more broadly, digital — evidence was admitted are likely much higher; the search was limited to the specific use of the term “video evidence.”

4 Email from John K. Grant, deputy clerk, N.J. App. Div., to author (July 13, 2022, 14:29 EST) (on file with author). In the author’s experience, the majority of criminal appeals in New Jersey do not yet involve recorded evidence. Even so, the sharp rise in such cases in recent years may be attributed in part to law enforcement policies requiring greater use of police dashcams and bodycams and the recording of station house interrogations. See Seth W. Stoughton, Police Body-Worn Cameras, 96 N.C. L. Rev. 1363, 1375–76 (2018) (noting that “[i]n car cameras were seen as a way to gather critical data” and noting that 70% of police agencies by 2013 used in-car video cameras); James S. Arrabito, Out of Focus: Zooming In on Body Cameras, Privacy, and Medical Emergencies, 69 Rutgers U. L. Rev. 741, 743–44 (2017) (citing surveys noting that at least a quarter of 17,000 law enforcement agencies use body-worn cameras and 80% of them are evaluating their effectiveness); Courtney A. Lawrence, Note, Recent Minnesota Supreme Court Decisions: Criminal Law: Too Much of a Good Thing: Limiting the Scope of the Scales Recording Requirement to Custodial Interrogations Conducted in Minnesota — State v. Sanders, 37 Wm. Mitchell L. Rev. 325, 331–32 (2010) (using State v. Sanders as a “case in point” to argue that as laws develop around police recordings, “courts will continue to face new issues concerning the scope of their application”).

5 See Email from Terrence D. Pricher, first assistant clerk, Mass. App. Ct., to author (July 13, 2022, 09:45 EST) (on file with author). Among other things, Pricher noted in his email that the technological difficulties arising from video submissions include (1) differences in platforms making the downloading and playing of the audiovisual exhibits a challenge; (2) storing the exhibits, particularly larger files; (3) the fact that software licenses expire; (4) court equipment not meeting the minimum technology requirements to play some exhibits; and (5) the fact that many lawyers are not technologically savvy. Id.; Email from Judge Gregory K. Orme, Utah Ct. App., to author (Feb. 14, 2023, 17:53 EST) (on file with author). Most of that increase was attributed to dashcam footage in criminal cases. Email from Pricher to author (July 13, 2022, 09:45 EST) (on file with author). In consultation with the court’s chief clerk and one of his deputies, Judge Christopher M. Murray estimates a roughly 25% increase in video submissions in criminal and civil appeals in the past five to seven years. Email from Judge Christopher M. Murray, Mich. Ct. App., to author (Feb. 8, 2023, 11:52 EST) (on file with author). Judge Pierre Bergeron estimates that some video evidence is in the record in about half of his court’s criminal appeals. Email from Judge Pierre Bergeron, Ohio Ct. App., to author (Feb. 13, 2023, 15:03 EST) (on file with author).

6 See, e.g., United States v. Fossat, 25 F. Cas. 1157 (C.C.N.D. Cal. 1857) (No. 15,137), rev’d on other grounds, 61 U.S. (20 How.) 413 (1858) (considering, at the trial court level, evidence of a photograph of an oak tree to resolve a boundary dispute); Luco v. United States, 64 U.S. (23 How.) 515 (1859) (concluding from photographs that documents purported to be Mexican land grants were forgeries).

7 See, e.g., Peter P. Roper, Sound Recording Devices Used as Evidence, 9 Clev.-Marshall L. Rev. 523, 218 (1960) (citing Boyne City, Gaylord & Alpena R.R. Co. v. Anderson, 109 N.W. 429, 430 (Mich. 1906) as an example where a litigant permissively presented sound recordings of a noise to a jury).

8 See, e.g., 44 Am. Jur. Trials 171, § 50 & n.69–71 (1992 & Supp. 2022) (citing 2 John Henry Wigmore, A Treatise on the System of Evidence in Trials at Common Law § 798 (2d ed. 1923)).

9 See id. § 50 n.71–72.

10 See, e.g., Nicholas R. Mack, The Use of Amateur Videotapes as Evidence in Criminal Prosecutions: Citizen Empowerment or Little Brother’s New Silver Platter?, 15 Hastings Commc’ns & Ent. L.J. 797 (1993) (discussing the Zapruder film of the assassination of President John F. Kennedy and the video recorded beating of Rodney King by police officers in 1991); Janelle Griffith, New Video Shows What Happened Before George Floyd’s Deadly Encounter with Police, NBC News (March 31, 2021, 7:35 PM), https://www.nbcnews.com/news/us-news/new-videos-show-what-happened-george-floyd-s-deadly-encounter-n1262670 [https://perma.cc/2EHB-LUDZ].

11 See Denise K. Berry, Note, Snap Judgment: Recognizing the Propriety and Pitfalls of Direct Judicial Review of Audiovisual Evidence at Summary Judgment, 83 Fordham L. Rev. 3343, 3343–44 (2015) (noting that “appellate judges, including Supreme Court Justices, have listened to audio recordings, scrutinized artwork, . . . watched video footage in order to decide for themselves whether there is a genuine issue of material fact for trial”).

12 See id. (noting “objective components of the record are considered vitally important to the decisions” and “[w]hen no objective evidence is available, appellate judges are left with ‘he said, she said’ testimonial evidence in which demeanor evidence looms larger and are therefore more likely to allow the cases to proceed to trial”).

13 See, e.g., State v. Knight, 327 A.3d 148 (N.J. 2024) (upholding the presentation of a video recording at slower speeds at a jury’s request during deliberations).

14 Sabatino, supra note 1, at 11 (noting that “[r]ecording footage may be incomplete, blurry, poorly lit, inaudible, skewed, or otherwise lacking. Those who watch and hear such digital evidence may be influenced by implicit or confirmation bias in interpreting its contents.”).

15 The risks of confirmation bias were demonstrated by the famous “gorilla suit” experiment, in which many observers of a video who were focused on counting the number of times people on the screen passed around a basketball totally overlooked a person dressed in a gorilla suit who briefly walked across the screen behind the basketball passers. See Christopher Chabris & Daniel Simons, Selective Attention Test, YouTube (Mar. 10, 2010), https://youtu.be/vJG698U2Mvo [https://perma.cc/6MGE-6KHW].

16 See Sabatino, supra note 1, at 31 (proposing a model rule of appellate procedure regarding video- or audio-recorded evidence).

17 Arnstein v. Porter, 154 F.2d 464 (2d Cir. 1946).

18 Id. at 469–73.

19 Id. at 475–76 (Clark, J., dissenting).

20 Scott v. Harris, 550 U.S. 372, 374–76.

21 The video recording is posted on the official website of the U.S. Supreme Court, under a heading labeled “Scott v. Harris — MP4 File.” See Media Files Cited in Opinions, U.S. Sup. Ct., https://www.supremecourt.gov/media/media.aspx [https://perma.cc/8439-7M2L] (last visited Dec. 27, 2025).

22 Scott, 550 U.S. at 374.

23 See id. at 376; see also Harris v. Coweta Cnty., 406 F.3d 1307, 1319 n.12 (11th Cir. 2005) (discussing what the video showed).

24 Scott, 550 U.S. at 374.

25 Id. at 384.

26 Id. at 390 (Stevens, J., dissenting).

27 Id. at 379 n.5 (majority opinion); id. at 387 (Breyer, J., concurring).

28 See Sabatino, supra note 1, at 50; see also Ex Parte City of Montgomery v. Shavers, 272 So. 3d 155, 163–66 (Ala. 2018) (relying on contradictory dashboard camera video that “depict[ed] a different story than the story [the driver] suggests”); Jarrell v. City of Nitro, 244 W. Va. 666, 674 (2021) (overturning circuit court’s reversal of findings of state Civil Service Commission that police had not used excessive force, given that security video footage of the events did not clearly show the commission’s findings were incorrect); Smith v. Wal-Mart Stores E., LP, 330 Ga. App. 340, 348 (2014) (reversing lower court’s dismissal of a plaintiff’s false imprisonment claim because the video evidence “substantially supports” plaintiff’s account of the events).

29 See Sabatino, supra note 1, at 51.