An interesting surveillance video authentication case involving opposing experts found its way to the Appellate Division of the Superior Court of New Jersey. This article reviews the facts that were considered by the trial judge, the expert evidence presented, the authentication ruling of the trial judge, and the appellate review of that ruling.

The procedural history of this case is somewhat complicated. The defendant was convicted of murder and related offences at his first trial. Surveillance video was an important feature of the prosecution’s case. The Appellate Division overturned the convictions and ordered a retrial. In advance of the retrial, an evidentiary hearing was held before the second trial judge to determine the admissibility of surveillance video evidence that the prosecution wished to tender. The video evidence was ruled admissible, and the defence appealed this ruling to the Superior Court of New Jersey, Appellate Division. This article is based on the 2023 ruling of the Appellate Division (State v. Puskas, WL 3486271).

The Evidence Presented at the First Trial

The victim’s body was found in the backyard of a residence during the mid-morning. The cause of death was determined to be “multiple blunt force injuries to the head delivered by blows from a two-pronged instrument like a crowbar or wrench”. The police investigation determined that the victim had attended two frat parties the night before and was heavily intoxicated when he left the last party. The police obtained and reviewed approximately 400 hours of surveillance video from area businesses recorded that night. The state’s forensic video analyst prepared a forty-five minute composite video which combined many separate videos from various locations and perspectives. Though the video quality was less than ideal, the state argued that the videos showed the victim travelling from the last party to the street where his body was ultimately located in the residential backyard. The state further contended that the video recording also showed a second person who was alleged to be the defendant. None of the video evidence showed any interaction between the two people depicted in the video.

The state’s forensic video analyst testified in a pre-trial evidential hearing that the video evidence was obtained from eleven different locations. The analyst personally recovered the surveillance video from several of those locations. He testified as to the recovery process, noted that the video evidence fairly and accurately depicted the events that were recorded, and explained the infrared aspects of the some of the recordings. He clarified the video evidence but did not alter its content. He tracked the victim and the suspect from 0202-0248 hours. There were minor time discrepancies on some of the recovered videos and these offsets were identified and applied when presenting the timeline evidence in court. The analyst also testified as to the description of clothing worn by the victim and tracked him through various recordings. Similar testimony was given regarding the tracking of the person believed to be the defendant.

The defendant’s forensic video analyst testified that some of the video recordings offered very limited visual information and that clothing details could not be determined due to the distance of the person from the camera. He also cautioned that linking a person seen in one video to a person shown in another video required considerable analysis, which the defendant’s analyst did not undertake. The defence argued that the video evidence should not be admitted because the video quality was so poor that it could not be authenticated. The trial judge disagreed and admitted the video evidence finding that the authentication threshold had been met. The defendant was ultimately convicted at the first trial.

The First Appeal

The appellate court reversed the convictions and ordered a retrial on the basis that the trial judge erred in admitting certain (non-video related) evidence. The appellate court did not address the defence complaints about video authentication but directed that in the re-trial, those concerns should be addressed by the trial judge.

The Second Trial (Re-trial)

The trial judge hearing the retrial conducted a pre-trial hearing to determine the admissibility of the state’s video evidence. The video evidence was presented to the trial judge to view, but no live expert testimony was presented by either side. Instead, both the state and the defence relied upon the expert testimony presented at the first trial and offered to present the experts if any further testimony was required.

The trial judge found the video evidence to be highly relevant as it supported the state’s theory as to the movement of the victim and the alleged defendant. Even though some of the video recordings were “not the clearest” the court found that the state’s interpretation of them was plausible, and that the jury should be tasked with deciding whether the state’s interpretation had merit. The trial judge rejected the defence’s argument regarding authentication and found that the prior testimony of the state’s analyst adequately established authenticity given his knowledge of the recordings, the locations depicted, and the fact that he recovered several of the video recordings in question. Further, the time discrepancies were adequately explained.

The trial judge admitted the state’s video evidence but ruled that the state’s analyst could not testify as an expert and could not offer a lay opinion as to the identification of the people shown on the video. The appellate ruling does not state why the trial judge ruled that the analyst would not be qualified as an expert, but it likely relates to a desire to forestall any expert opinion evidence on the identification of the defendant. Curiously, the ban on lay opinion identification appears to be contrary to an earlier appellate ruling (State v. Nieves, which is the subject of an earlier article on this website), which would have otherwise permitted such an opinion given that the state’s analyst knew the defendant personally.

The Second Appeal

The defence appealed this ruling in advance of the second trial and argued that the trial judge erred in ruling that the video evidence could be presented at trial. The defence argued that the video evidence was of such poor quality that it was incapable of being authenticated. The Appellate Division ruled that the state provided a sufficient basis upon which the jury could find that the individuals shown in the video recordings were the victim and the defendant. The trial judge correctly ruled that these were matters of weight, not admissibility. Further, the time stamp discrepancies did not lead to a lack of authentication given the evidence presented by the state’s analyst. Video evidence shown to have the correct time could be used to authenticate other videos with the incorrect time. These too were arguments for the jury. Accordingly, the Appellate Division found no error in the trial judge’s pre-trial authentication ruling. The retrial is currently scheduled for January 2024.

Commentary

The Puskas ruling does not create new law, but rather reaffirms long held evidential principles regarding video authentication, as follows:

a) Poor quality video is almost always a question of weight, not admissibility.

b) Poor image quality does not usually impact authentication as the trier of fact will determine what value, if any, can be gained from the images.

c) Authenticated video can be used to authenticate other video recordings (ie. address time stamp variances).

d) Date and time stamp variations, once explained, are a question of weight, not admissibility.

This ruling is also helpful because it sets out a good example of proper forensic practice. In this case, the state’s analyst undertook important technical and analytical steps that are worthy of note. These steps reinforce good forensic practices including the following:

a) Surveillance video recordings should be recovered by someone who is trained to do so in a forensically sound manner.

b) Video recordings should be clarified in a non-destructive manner to achieve maximum visual detail.

c) All video recordings should be preliminarily examined to assess time and date discrepancies. These issues can be addressed relatively easily if done promptly.

d) Look outside the box. Use authenticated video recordings to bookend and authenticate other recordings.

e) Examine video recordings in minute detail to excise relevant portions.

f) Track people of interest as between different recordings from different locations. This will help to tell the story that the video recordings have to offer to the informed viewer.

g) Prepare a composite video that contains all relevant video segments.

A forensic video analyst should interrogate, analyze, and interpret video evidence and be prepared to offer the benefit of that knowledge to the parties and the court in furtherance of the search for the truth.