Commonwealth v. Caruso (Lawyers Weekly No. 11-019-14)

NOTICE: All slip opinions and orders are subject to formal revision and are superseded by the advance sheets and bound volumes of the Official Reports.  If you find a typographical error or other formal error, please notify the Reporter of Decisions, Supreme Judicial Court, John Adams Courthouse, 1 Pemberton Square, Suite 2500, Boston, MA 02108-1750; (617) 557-1030;




12‑P‑1096                                       Appeals Court



No. 12‑P‑1096.

Middlesex.     November 15, 2013.  ‑  February 28, 2014.

Present:  Kafker, Milkey, & Hines, JJ.


Practice, Criminal, New trial.  Evidence, Computer simulation.



Complaint received and sworn to in the Malden Division of the District Court Department on November 5, 1998.


A motion for a new trial, filed on March 18, 2008, was heard by Gregory C. Flynn, J., and motions for expert funds were also heard by him.



David A.F. Lewis for the defendant.

Jessica Langsam, Assistant District Attorney, for the Commonwealth.




MILKEY, J.  In 1999, following a jury trial in District Court, the defendant was convicted of two counts of malicious destruction of property (a car), G. L. c. 266, § 127.  In 2001, this court affirmed those convictions in an unpublished memorandum and order.  Commonwealth v. Caruso, 52 Mass. App. Ct. 1101 (2001).  Seven years later, after the defendant was convicted of murdering the owner of the damaged car, he filed a motion for new trial.  He argued that newly discovered evidence definitively proved that he could not have been the person who damaged the car.  That motion was heard and denied by the trial judge, who issued a forty-one page decision that explained his reasoning in thoughtful detail.  For the reasons set forth below, we agree with the judge that the material on which the defendant relies does not constitute “newly discovered evidence.”  We also agree that this “evidence” — while of some superficial force — does not cast any “real doubt on the justice of the conviction.”  Commonwealth v. Grace, 397 Mass. 303, 305 (1986).  We therefore affirm.

Background.  We summarize the key trial evidence as follows.  The victim was Sandra Berfield, who lived in Everett.  On four separate occasions during the fall of 1998, someone caused serious damage to her car while it was parked on the street next to her home.  In the first two incidents, all four of Berfield’s tires were slashed; in the second two, battery acid was poured into her gas tank.  The dispute at trial was over the identity of the perpetrator.


Berfield worked at a Bickford’s restaurant in Medford.  The defendant patronized the restaurant with pronounced regularity, eating there once or twice virtually every day.  Beginning in 1996, the defendant insisted on sitting in Berfield’s section; he stayed there for lengthy periods while staring at her.  Until August of 1998, the interactions between the defendant and Berfield were otherwise unremarkable.  The two engaged in what Berfield described as “light conversations,” and she characterized their relationship as “sort of friendly.”  In late August of 1998, the defendant asked Berfield if she wanted to go to a movie with him, which Berfield declined.  After that, the defendant continued to insist on sitting in Berfield’s section and to stare at her continually while smirking.  Berfield became increasingly uncomfortable in his presence, and she avoided engaging him in conversation.  Eventually, at her repeated insistence, the restaurant manager asked the defendant to sit in a different section of the restaurant.  The incidents with Berfield’s car began in late September.

The first incident.  On September 27, 1998, Berfield found that someone had slashed her tires overnight.  There were no eyewitnesses to the incident, and the defendant was never charged with this offense.


The second incident.  In the early morning of September 30, 1998, while Berfield was lying on her couch, she heard a second tire-slashing episode in progress.  From her second floor window, she observed the apparent perpetrator walk away from her car.  He was dressed all in black and wore a hood, and Berfield could not see his facial features.  However, according to her trial testimony, she was able to recognize the defendant from his distinctive gait (which she had observed at the restaurant).  She described the gait as “almost like a bounce as he walks with his arm swinging back and forth, his left arm.”  When she reported the second incident to the police, Berfield stated her belief that the defendant was the perpetrator.  She acknowledged that she did not mention the defendant’s gait to the police at that time, although she stated that she did mention his “figure.”[1]  Fearful that the vandalism of her car would continue, Berfield set up a surveillance camera on the second floor of her home to observe the streetscape below.

The third incident.  On October 4, 1998, Berfield’s car malfunctioned while she was driving to work.  A repair man discovered that someone had poured battery acid into the gas tank.  When she returned home, Berfield learned that her surveillance camera had captured the apparent perpetrator walking next to her car at 3:15 A.M. that morning.  The person in the video appeared to Berfield to be the same person she had observed after the second tire slashing, clothed as before, all in black and wearing a hood.  Berfield testified that she once again recognized the defendant from his distinctive gait.


The fourth incident.  Battery acid was again poured into Berfield’s gas tank on October 25, 1998, at approximately 2:15 A.M.  Berfield was home at the time, and she watched the incident on a video monitor as it unfolded.  This was also recorded.  The video shows a partially obstructed view of a red car pass by Berfield’s home at approximately 2:11 A.M. traveling from right to left.  Berfield testified that the car looked “very similar” to the one that she knew the defendant to drive.[2]  Approximately four minutes later, what appears to be the same hooded, dark-clothed figure enters the scene, walking from the left.  Again, the person’s facial features cannot be discerned.  The figure walks alongside Berfield’s car while carrying something in his hand, and then attends to the area around the gas cap before returning in the direction from which he came.

Berfield telephoned 911 during the incident, and an excerpted recording of her call was admitted.[3]  During the portion of the call played to the jury, she twice — without any hesitation or expression of doubt — identified the defendant as the person who was vandalizing her car.  She told police after the incident that she recognized the defendant by his “shape, build and walk.”


While Berfield was on the 911 call, she ran downstairs where she was able to observe the perpetrator from the first floor, front hall window.  According to her testimony, the person turned his face slightly as he was leaving, and she thereby was able to confirm that it was the defendant from his facial features.  Berfield acknowledged that she never told the officer who took her statement that she saw the perpetrator’s face, although her specific answer suggested that she might have told a different officer.


The verdict.  Based on the second tire slashing and the two battery acid incidents, the defendant was charged with three counts of malicious destruction of property.  The Commonwealth’s claim that the defendant was the hooded man who vandalized Berfield’s car was based principally on her testimony.  The defendant argued to the jury that any objective measures by which Berfield could have identified the perpetrator were thin, that she merely surmised it was him based on a misperception of his odd but innocent behavior, and that once having settled on the defendant, she embellished her reasons for identifying him significantly beyond the contemporaneous accounts she had given to police.  The defendant also highlighted that Berfield referred to the perpetrator as “tall,” while he was in fact relatively short.[4]  Finally, the defendant sought to suggest other possible suspects, including, most prominently, a former boyfriend (himself six feet tall) whom Berfield mentioned to police as a possible suspect in the two tire slashings.

On May 10, 1999, the jury found the defendant guilty of the two battery acid incidents, but acquitted him of the second tire slashing.  The judge sentenced him to concurrent terms of eighteen months in the house of correction (six months to serve, the balance suspended), and ordered him to pay $ 3,000 in restitution, to undergo a psychiatric evaluation, and to stay away from Berfield and her restaurant.


Posttrial developments.  In 2000, Berfield was murdered when a bomb delivered to her home exploded.  The defendant eventually was convicted of murder in the first degree for her death after a twenty-two day trial.  In preparing a defense of the murder charge, his legal team began to reexamine the evidence related to the incidents involving Berfield’s car.  As part of that effort, the defendant hired Rampion Visual Productions LLC (Rampion) to examine the video of the October 25, 1998, incident.  In 2003, relying on “computer aided design” (CAD) software, as well as on data gleaned from a survey the defendant had done of the area around Berfield’s home, Rampion produced an analysis (Rampion analysis) purporting to show that the person seen in the video must have been at least five feet, ten inches tall.  According to the defendant, the Rampion analysis definitively proves that he could not have been the person who damaged Berfield’s car, because he is some four inches shorter.

In 2007, the defendant moved for a new trial in the malicious destruction of property case, based on the Rampion analysis.[5]  The Commonwealth countered with an affidavit submitted by a “Supervisory Photographic Technologist in the Forensic Audio, Video & Image Analysis Unit of the Federal Bureau of Investigation (FBI)” (FBI affidavit) and a report completed by the same individual (FBI report).  After holding a nonevidentiary hearing, the District Court judge who presided at trial denied the motion.  He also denied the defendant’s separate motions seeking funds for experts to respond to the FBI report.


Discussion.  A defendant seeking a new trial based on “newly discovered evidence” faces formidable burdens.  To overcome the important societal interests of finality and judicial efficiency, he “must establish both that the evidence is newly discovered and that it casts real doubt on the justice of the conviction.”  Commonwealth v. Grace, 397 Mass. at 305.  To grant the motion, “the judge must find that there is a substantial risk that the jury would have reached a different conclusion had the evidence been admitted at trial.”  Id. at 306.  On appeal of the grant or denial of such a motion, we review only for “a significant error of law or other abuse of discretion.”  Commonwealth v. Weichell, 446 Mass. 785, 799 (2006), quoting from Commonwealth v. Grace, supra at 307.  Where the motion judge was also the trial judge, as is the case here, “special deference” is due.  Commonwealth v. Fappiano, 69 Mass. App. Ct. 727, 730 (2007).


1.  Whether the 2003 Rampion analysis was “newly discovered evidence.”  The judge concluded that the defendant failed to meet his threshold burden of showing that the Rampion analysis was “newly discovered evidence.”  To meet that standard, the evidence must “have been unknown to the defendant or his counsel and not reasonably discoverable by them at the time of trial.”  Grace, supra at 306.  The defendant acknowledges that CAD technology has existed “since at least the 1980′s.”  In addition, the FBI affidavit established that CAD technology had progressed by 1999 to allow the type of analysis that the defendant subsequently had performed just four years later.  Faced with that fact, the defendant attempts a more subtle argument.  According to him, even if a CAD-based analysis technically could have been done in 1999, it was not readily available at that time, and it became so only after post-1999 technological advances made it easier and less costly to use.  In this manner, he draws an analogy to the use of deoxyribonucleic acid testing, which became generally available long after it became technologically possible.  See Kaye, The Double Helix and the Law of Evidence 261-263 (2010).

Without deciding the specific contours of what a defendant must show to prove that evidence based on developing technology was not “reasonably discoverable” prior to trial, we conclude that the defendant has failed to make an adequate showing here.  He submitted two affidavits toward that end.  In one, his trial counsel stated that a company that specializes in analyzing videotapes reported to him prior to trial that it was “unable to make any determination regarding the videotape.”  As the judge pointed out, no explanatory context was provided for this vague statement from one unnamed consultant.[6]  We agree with the judge that such a statement hardly establishes that the sort of analysis done in 2003 was not reasonably available in 1999.


Similarly unhelpful are vague statements attributed to the analysts who performed the Rampion analysis.  According to an affidavit from the defendant’s appellate counsel, the Rampion analysts stated that they were able to perform their analysis in 2003, “in part, because of the technological advances since the time of the defendant’s 1999 trial.”  Again, such a statement simply does not establish that the practical unavailability of CAD technology in 1999 prevented the defendant from obtaining the kind of analysis that he had performed a mere four years later.[7]


Even had the defendant been able to show that CAD technology was not reasonably available to him in 1999, he would face a separate problem.  As the judge observed, “there is nothing new about . . . the geometry of estimating heights and distances [depicted in photographs] from known points of reference.”[8]  The height of the person shown in the video still shot could have been calculated using a technique known as photogrammetry.[9]  According to the FBI affidavit, photogrammetry has been in use for over a century, and in any event, the defendant conceded at the motion hearing that a photogrammetric analysis of the height of the person in the video could have been performed prior to trial.  Because the defendant had available to him a different means of achieving the same end, whether CAD technology had sufficiently progressed by the time of trial is ultimately beside the point.  The posttrial development of an alternative technological method of proving the same evidentiary point does not, without more, create “newly discovered evidence.”[10]

2.  Whether the Rampion analysis cast “real doubt.”  Having concluded that the Rampion analysis did not constitute “newly discovered evidence,” the judge went on to consider whether it provided the definitive showing that the defendant claimed.  We agree with the judge’s assessment of that issue as well.


a.  The force of the “new” evidence.  According to the defendant, the Rampion analysis warrants a new trial because it amounts to “unshaken documentary proof” supporting his innocence.  Commonwealth v. Woods, 382 Mass. 1, 9 (1980) (recognizing that appellate scrutiny of denial of new trial motions should intensify “in the degree that unshaken documentary proof exists”).  However, Woods does not help the defendant, because it involved a very different sort of documentary proof.  There, the court concluded that a new trial was warranted primarily because Department of Correction records supported the defendant’s claim that he was incarcerated when the crime most likely occurred.  Id. at 5-7.  Thus, at the heart of Woods is the fact that there was no apparent reason to question the accuracy of the documentary evidence that supported the defendant’s innocence.  Here, by contrast, the documentary proof that the defendant claims establishes his innocence was generated by the defendant’s own expert.  Although the defendant makes extravagant claims about the accuracy of the CAD methods used by his experts,[11] even a limited examination of such claims reveals that the judge’s skepticism was well-founded.


Whatever else can be said about the capabilities that CAD technology presents, it remains by definition merely a tool whose accuracy can be no greater than that of the data and assumptions on which it relies.  See Commercial Union Ins. Co. v. Boston Edison Co., 412 Mass. 545, 549 (1992) (discussing admissibility of computer-generated models).  While the Rampion analysts began by using underlying data whose accuracy the Commonwealth does not question (the survey of the surrounding area and the still shot taken from the video), they then went through multiple steps to produce the final work product that they claim definitively establishes the perpetrator’s height.[12]  The choices that the analysts made during those steps affect the ultimate conclusions drawn about the perpetrator’s height and therefore call into serious question the accuracy of the defendant’s claims.


For present purposes, one example will suffice.  Although the resolution and lighting of the video make it challenging to discern much about the clothes the perpetrator was wearing, it is undisputed that he was wearing a hood.  As explained in their transmittal letter, the Rampion analysts used an outline of the person shown in the video still shot to create “[a] figure with generic build, shoes, and clothing.”  That “[f]igure was ‘dressed’ in a hooded jacket supplied by in-house library.”  In other words, the Rampion analysts made choices about the nature of the hood the person was assumed to have been wearing.  The FBI analyst pointed out that such choices obviously can affect how tall that person would appear:  to the extent that the hood was sticking above the top of the person’s head, he would look taller than he actually was.

As revealed in the slides that Rampion created, the particular hood that the analysts chose was one that fit snugly to the head, as opposed to one with a more distinct peak.  In fact, Rampion’s cover letter reveals that the analysts changed their assumption about the type of hood the person was wearing during the course of their analysis.  Specifically, Rampion described one of the “refinements” made after the initial round of analysis as follows:  “[t]he figure’s jacket was made to look less cumbersome and the hood was fitted closer to the head [while] allowing enough room for hair.”  Changing that assumption presumably resulted in an increase in the person’s calculated height, because a close fitting hood would account for less of the over-all height of the figure seen in the video still shot.  Thus, a careful reading of Rampion’s own explanation of what its analysts did reinforces the FBI’s point regarding the hood.[13]



By focusing on the assumptions the Rampion analysts made about the person’s hood, we do not mean to suggest that these factors are the only ones that undercut the purported exactitude of their analysis.  The FBI report highlighted that serious potential questions that can also be raised about other specific steps in the process that Rampion used.  For example, the Rampion analysts had to place the human figures into the virtual landscape that was created.  As the FBI affidavit pointed out — relying on elementary principles of geometry — it matters significantly how close the person is assumed to be standing in relation to the viewer and to the items against which his height is being compared.[14]  The defendant counters that some of the specific criticisms the FBI affidavit aimed at the Rampion analysis miss their mark.[15]  On this basis, he argues that the judge erred in relying on the FBI report, at least without his being given a chance to rebut that report with additional expert proof.  Resolution of any remaining technical debate is unnecessary to our disposition of this appeal because we can confidently state that the Rampion analysis is not the “unshaken documentary proof” that the defendant claims.[16]  Commonwealth v. Wood, 382 Mass. at 9.

b.  The strength of the Commonwealth’s case.  In concluding that the “new” evidence did not create “real doubt” about the jury’s verdict, the judge also appropriately examined the strength of the Commonwealth’s case.  See Commonwealth v. Lykus, 451 Mass. 310, 326 (2008) (“[In] decid[ing] whether the new evidence probably would have been a real factor in the jury’s deliberations, . . . the judge must consider the strength of the case against the defendant”).  The defendant argues that in making that assessment, the judge exaggerated the strengths of the Commonwealth’s evidence, and ignored obvious problems with it.  We disagree.


If the Commonwealth’s proof was not overwhelming, neither was it thin.  Notably, this is not a case involving stranger identification.  Berfield was able to recognize the defendant from his facial features and distinctive gait after having observed him for literally hundreds of hours and having had reason to have his features etched in her memory.  Contrast Commonwealth v. Woods, 382 Mass. at 7 (new trial warranted where documentary evidence supported defendant’s alibi and where there were specific reasons to question victim’s identification of him as stranger who raped her).


Moreover, the defendant significantly exaggerates the potential problems with Berfield’s testimony.[17]  Although that testimony presented some fodder for cross-examination, defense counsel did not miss any such opportunity.[18]  The jury credited Berfield’s account of the battery acid incidents in spite of the defendant’s pointed arguments why they should not do so.  Having seen exactly what the jury saw, the judge was in a unique position to assess the extent to which the Rampion analysis might have made a difference.  His assessment of the trial evidence is entitled to “special deference.”  Commonwealth v. Grace, 397 Mass. at 307.  See Commonwealth v. Santiago, 458 Mass. 405, 414 (2010) (trial judge “is in the best position . . . to determine [new evidence’s] probable impact on a jury hearing it with all the other evidence”).

Conclusion.  For all the reasons set forth above, we conclude that the judge did not abuse his discretion in denying the defendant’s motions for new trial and for expert funds.

Order denying motion for new

trial affirmed.


Order denying motions seeking                                     expert funds affirmed.




     [1] The officer with whom Berfield spoke remembered that Berfield based her identification on the fact that the defendant was harassing her, and that she “was bringing him to Court.”

     [2] The defendant claims that other trial evidence definitively proved that his car was equipped with a luggage rack, while the car captured on the videotape lacked one.  However, his proof was less than compelling.  During cross-examination, defense counsel handed Berfield a photograph of a car that was equipped with a luggage rack.  That photograph was never authenticated or entered in evidence.  Berfield’s statement that the car depicted in the photograph “looks like [the defendant’s] car” hardly establishes definitive proof that the car was in fact the defendant’s car and that it had a luggage rack on it at 2:12 A.M. on October 25, 1998.  Of similar inconclusive import is an acknowledgement made by an officer who went to the defendant’s home later that morning that he was “not able to identify from the video any of those cars as being the car that [he] saw at [the defendant’s] house when [he] got [there] in the wee hours of the morning.”

     [3] At the defendant’s urging, the jury did not hear Berfield’s statement that the defendant had been “stalking” her.

     [4] At trial, Berfield acknowledged that the defendant’s height appeared to be between five feet, six inches and five feet, seven inches.  Department of Correction records submitted with his motion for new trial indicate that he is five feet, six inches tall.  Berfield herself was only five feet tall, and she offered that she viewed the defendant as “tall” even if others

would not.

     [5] The defendant sought to have the Rampion analysis admitted in the murder trial, but a Superior Court judge in that case ruled it inadmissible as lacking a sufficient nexus to the crime charged.  The Superior Court judge stated his view that the defendant inappropriately was seeking “to relitigate his identity as the perpetrator of other unrelated crimes, in separate prosecutions, in which he had a full opportunity to present evidence on his behalf, and in which he was found guilty beyond a reasonable doubt by a jury of his peers.”  The defendant’s appeal of his murder conviction was stayed and remains pending before the Supreme Judicial Court.

     [6] As the judge noted, the affidavit fails to “indicate what type of analysis [the lawyer] sought at the video business, i.e., whether he wanted to discover whether the tape had been tampered with or was in some way fraudulent, or if he sought an enlargement of a single frame or wanted to see if he could slow the speed of the tape to use at trial.”

     [7] The defendant’s appellate counsel’s affidavit also stated that the CAD tools that Rampion used in 2003 “were not being used by the average computer user or designer” at that time, a standard of little seeming relevance.  In addition, the affidavit — executed in March of 2008 — focused on technological developments “over the last eight to ten years” rather than in the four-year period from 1999 to 2003.

     [8] At the motion hearing, the judge trenchantly noted that “the Pythagorean Theorem didn’t just get invented.”

     [9] Combining optics and geometry, photogrammetry allows refined estimates to be done of distances shown in photographs.  See Moffitt & Mikhail, Photogrammetry 1-7 (2d ed. 1967).

     [10] It does not follow, as the defendant maintains, that his trial counsel must have been ineffective for failing to conduct a height analysis prior to trial.  As this case well illustrates, the stakes of the prosecution at issue are relevant to that question:  in preparing a defense to a charge of murder in the first degree, counsel naturally would be expected to conduct a more intensive investigation than in a case involving low-level property crimes.  Especially where trial counsel did make some effort to consult a video expert, the defendant has not shown that his counsel’s conduct fell “measurably below that which might be expected from an ordinary fallible lawyer.”  Commonwealth v. Saferian, 366 Mass. 89, 96 (1974).  In addition, because the Rampion analysis did not create any “real doubt” about his conviction, see discussion, infra, the defendant is also unable to show that the failure of his trial counsel to have such an analysis done “deprived [him] of an otherwise available, substantial ground of defence.”  Ibid.

     [11] The defendant argues that the Rampion analysis is based on “an exact replica of what the video camera shot because the three-dimensional model does not add or alter the specifications given to it from the survey [and] nothing is represented in the model that was not all of the following:  (1) there in reality, (2) measured in the defendant’s survey and (3) given to the computer.”

     [12] The main steps that the Rampion analysts took can be summarized as follows:


1.  relying on the survey data, the analysts used the CAD software to create a three-dimensional virtual model of the area portrayed in the videos,


2.  using an overlay technique, the analysts manipulated the vantage point used to view this virtual landscape until the perspective shown matched the one in a still shot taken from the October 25, 1998, video,


3.  relying on manufacturer specifications, the analysts inserted a scale model of Berfield’s car into the scene,


4.  based on an outline of the person shown in the video, the analysts created an image of a person, which they “dressed” in a particular manner discussed, infra,


5.  the figure was scaled to represent different heights and the scaled figures were inserted one by one into the scene, and


6.  the apparent height of the scaled figures relative to, for example, the roof of the car, were then compared to the real person shown in the video still shot to derive the best match.

     [13] One additional point bears mention.  The defendant has urged us to apply our own scrutiny to the video-based evidence, relying on precedent that appellate courts are in as good a position to review documentary evidence as trial courts.  See, e.g., Commonwealth v. Woods, 382 Mass. at 9.  This does not aid his cause, because our own review of the video from which the still shot was taken reveals something that apparently escaped the parties’ notice.  At the point the person is attending to the

area of the gas cap, the outline of his hood can be seen more clearly against the background of the sidewalk.  That view reveals that the hood in fact was in the nature of an oversized, peaked cowl (like that of a monk), and not the snug fitting, rounded hood that Rampion selected.

     [14] Indeed, the FBI affidavit noted that, based on the angle of view from the second-story window, erroneously placing the figure approximately one horizontal foot closer to the viewer (and away from the car) than he actually was, by itself could account for a four-inch overestimate of his apparent height.   Although Rampion relied on uncontested survey data in creating its virtual landscape, the surveyor obviously could not have known the precise location of either Berfield’s parked car or of the person shown walking next to it (in relation to each other or to the fixed landmarks).

     [15] The defendant maintains that the FBI affidavit failed to take into account the difference between a CAD-based height analysis and one based on photogrammetry.  For example, while acknowledging that the accuracy of a photogrammetric analysis depends in part on knowing certain data (such as the exact

location and focal length of the camera lens that took the photograph), he contends that such knowledge is unnecessary — or at least less vital — when CAD software is utilized.  He appears to be suggesting that, using CAD, an analyst can apply overlay techniques to create an accurate three-dimensional rendition of what was seen in the video still shot without having to know all the data that would be needed using photogrammetry.

     [16] Especially in that context, the judge did not abuse his discretion in denying the defendant funds to obtain additional expert evidence.

     [17] For example, the defendant highlights that Berfield testified to having been only fifteen feet away from the perpetrator when she saw his face, while on cross-examination she suggested it might have been twice as far.  There is little reason to doubt Berfield’s ability to recognize the defendant from either distance.

     [18] For example, counsel vigorously questioned Berfield about the potential discrepancies between how she claimed at trial to have recognized the defendant, and what she contemporaneously had told police.

Full-text Opinions