ONTARIO SUPERIOR COURT OF JUSTICE
COURT FILE NO.: CR-12-00000841-40
DATE: 20140814
BETWEEN:
HER MAJESTY THE QUEEN
Applicant
– and –
Babak Andalib-Goortani
Respondent
Philip Perlmutter and Molly Flanagan, for the Crown
Harry G. Black, Q.C. and Amelia Martin for the Respondent, Mr. Andalib-Goortani
HEARD: May 14 and 15, 2014
Trotter J.
INTRODUCTION
[1] A photograph can be a powerful piece of evidence, especially when it purports to show a crime in progress. Photographs are introduced into evidence on a daily basis in our courts. Most of the time, no objection is taken to the admission of this type of evidence.
[2] This case is different. Babak Andalib-Goortani is a Toronto Police Service (TPS) officer who is charged with assault with a weapon. The Crown seeks to rely upon a photograph that purports to record the alleged assault. The problem is that no one knows who took the photo because it was posted anonymously to a website.
[3] The defence opposes the admission of the photo, arguing that the Crown is unable to establish its authenticity. The Crown contends that it has authenticated the image and that the issue is one of weight for the trier of fact.[^1]
THE FACTS
(a) Photographs at the Scene
[4] The alleged assault occurred on June 26, 2010 during the G20 Summit in Toronto. I need not say much about the G20 Summit, except to mention that it was marked with violent protests and clashes between protesters and the police. It is alleged that, during a protest at the Ontario Legislature at Queen’s Park, P.C. Andalib-Goortani struck Ms. Wyndham Bettencourt-McCarthy (the complainant) with a baton, causing a nasty bruise to her hip.
[5] Many photographs were taken at the protest. Mr. Vincenzo D’Alto, an accredited photojournalist from Quebec, took photos that day. He testified at the preliminary inquiry and authenticated many images. Mr. D’Alto did not take the photo in question, which depicts a police officer winding up and about to hit a female with a baton. There is no doubt that the female image in the photo is that of the complainant. The question is whether P.C. Andalib-Goortani is the police officer and whether the image accurately reflects what occurred that day.
[6] At the preliminary inquiry, the complainant said that she saw the photograph in question several days after the event. She testified that her friend told her he found it on a site called www.g20justice.com. There is no evidence as to who uploaded the photograph to the site. The complainant was shown the photograph at the preliminary inquiry and said:
Ah, this is a photograph of the officer who struck me, and this is a photograph of him with his arm raised and his baton out. Ah, about the moment, the moment before he struck me and in the background, there are a group of officers who are engaged in…forcibly detaining…the man who is, who is lying on the ground.
Asked what she remembered about the officer, the complainant said:
… I was not able to see his face…very clearly as, as you can see in the photo, um, he has a helmet and a face shield that is...partially concealing his face. Um, but I remember seeing him come towards me and, ah, seeing a man who was…a bit shorter than me and somewhat stalky and who had facial hair, goatee style of facial hair.
The complainant is unable to identify the person who struck her.
[7] The Crown intends to rely on other photographs of P.C. Andalib-Goortani at the scene, as well as other photos of him, out of uniform, on a different day. These photos are included as part of this application because the Crown intends to call another TPS officer, Sgt. Angelo Costa, to identify the face of P.C. Andalib-Goortani in some of those photographs as a step towards identifying him in the photograph in question.
[8] Sgt. Costa, who has been with the TPS since 1982, has known P.C. Andalib-Goortani since 2010. At the preliminary inquiry, in November of 2012, he identified P.C. Andalib-Goortani in many photographs, but was unable to recognize him in the photograph in question. I am advised by counsel that Sgt. Costa has since given a statement saying that he was mistaken in his evidence at the preliminary inquiry and he is now able to identify P.C. Andalib-Goortani in the impugned photograph. This revelation occurred during an interview on May 9, 2013, almost three years after the alleged assault.
(b) Expert Evidence
[9] Both parties adduced expert evidence concerning the photograph. The Crown called Tracy D. Peloquin, a civilian member of the Ontario Provincial Police. Among other things, she is a Certified Forensic Video Analyst and a Forensic Identification Analyst. Her extensive curriculum vitae demonstrated that she is well qualified to provide an opinion on the authenticity of the photograph. She has attended many courses in her area of expertise.
[10] Ms. Peloquin located the image on the website noted above. She “captured” the image and examined it with a number of different types of software. This is explained in her comprehensive report, which was filed as an exhibit.
[11] Ms. Peloquin stated that the image was largely stripped of its metadata (i.e., data concerning the type of camera, focal length, lens type, etc.). She said that this likely happened when it was uploaded to the site, a common feature of many websites. She was also certain that the image had been cropped (i.e., reduced in size) when it was uploaded. As Ms. Peloquin concluded in her report:
The JPEG evaluation software (JPEGsnoop) concluded that the image had been processed and/or edited. There are a number of reasons why this result may have occurred:
The image had been altered or manipulated using some form of editing software.
The resizing/cropping of the image in the internet posting stage or at another time was read as a manipulation which indeed it would be.
Due to the lack of metadata, exif data and hex information due to the image’s origin, the evaluation software was not able to come to a reliable and robust conclusion.
Additional information would be of assistance to further authenticate the questioned images such as an image audit trail, original recording camera/recorded media, camera operator or any witnesses to the image being taken. [Emphasis added.]
[12] After reviewing her findings dealing with a comparative analysis (i.e., a comparison of the photograph in question with other known images of P.C. Andalib-Goortani), which is not relevant to this application, Ms. Peloquin concluded:
It is the writer’s opinion that there are no clear indications of manipulation in the suspect image…around the persons of interest or otherwise. The images show no artifacts that can be a result of being added or manipulated with an image-editing program. Depth of field, lighting, shadowing and digital compression remain consistent throughout the image. No visual evidence of image alteration and/or changes to the image structure was found. [Emphasis added.]
[13] Ms. Peloquin confirmed her opinion during her testimony before me. Upon a visual examination of the image, Ms. Peloquin did not notice anything that would raise alarms that it had been tampered with. She noticed no problems with shadowing, nor did she observe any sharp edges that might be indicative of an image being superimposed onto another image. Ms. Peloquin could only find evidence that the size of the image had been altered.
[14] Towards the end of her examination-in-chief, Ms. Peloquin said that, if someone did alter the image, they would have to be possessed of strong knowledge of the relevant software to do so without being detected. Mr. Black picked up on this cross-examination. Ms. Peloquin agreed that she was unable to say that the image had not been altered; all she could say was that there was no evidence it had been altered. For example, she could not say that the face of the police officer in the photo had not been changed or that the distance between the officer and the complainant had not been adjusted. Ms. Peloquin agreed that it is impossible to determine whether the image originated from a digital or analog device, or whether it is a still shot taken from a video. Moreover, Ms. Peloquin testified that there is no way to determine how many computers or websites the image has been uploaded to and downloaded from. She was able to locate the image on ten different websites.
[15] The defence relied on the evidence of Martin Musters, a person who I qualified as an expert in computer forensics. Mr. Musters does not have the same degree of formal training as Ms. Peloquin. However, he has a wealth of practical experience over 35 years, sufficient to provide an opinion on the matter at hand.
[16] Mr. Musters was asked to provide an opinion on whether the image has been altered, and to comment on Ms. Peloquin’s report. Similar to Ms. Peloquin, Mr. Musters was concerned about the lack of metadata. As he said in his report, “[t]he removal of the metadata results in many uncertainties about the provenance of this photograph and it makes it difficult to draw conclusions.”
[17] As in Ms. Peloquin’s report, Mr. Musters’ report discussed the easy availability of Photoshop, a program used to edit photos. Mr. Musters noted that, if a photo has been modified with Photoshop, and then uploaded to a website, evidence of the alteration would be removed when the image is stripped of its metadata.
[18] Mr. Musters testified that the dpi (dots per inch) of the photo is 300. At the time of the G20 Summit in 2010, a typical camera would take a photo at a value of 72 dpi. Based on this observation, and running the image through a forensic program called FOURMatch (a program developed by a company in the field of image validation), Mr. Musters concluded:
In our opinion the absence of metadata is indicative that the photograph was uploaded to Facebook at some point in its life cycle. The fact that the dpi of [the image] is 300 on the www.G20justice.com site points to it being altered manually with a photo editing software before it was posted to that website.
Based on our analysis, we can conclude that this image…was processed/edited with a software program, likely one of the ones listed by FOURMatch.
We support the “Verdict” of FOURMatch which states “that the file has been changed, though the amount of change is unknown.
[19] Mr. Musters elaborated on his opinion during his testimony. In addition to noting the unusual dpi feature of the image, he said that the image taken from the website was 61.2 kb in size. Mr. Musters testified that, in the unlikely event that the photo was taken with a camera set to 300 dpi, the size of the photo would have been far larger than 61.2 kb. This led him to conclude that the image taken from the website is not the original photo from the camera. Furthermore, Mr. Musters testified that it is possible to alter facial hair on an image and not see any forensic abnormalities. In short, Mr. Musters said that one could make changes to facial hair with Photoshop and not leave any evidence of having done so because Photoshop does a “really good job” of editing without detection because it blends colours to make it look as if no alternations have been made.
[20] Mr. Musters was challenged on his credentials and on his opinion during cross-examination. Mr. Musters acknowledged that he had only testified on six occasions, and never on the issue of digital imaging. However, he said he has been consulted many times about digital imaging, within the broader framework of his work in computer forensics. Mr. Musters admitted to not having had the benefit of the formal training achieved by Ms. Peloquin, but relied on his extensive 35 years of practical experience.
[21] Mr. Musters acknowledged that, at the time the photo was taken, there were some cameras available, high-end cameras made by Nikon and Canon, which could take pictures at 300 dpi. Mr. Musters agreed with the suggestion that a private website, such as www.g20justice.com, could reduce the size of a 300 dpi photo to 72 dpi. However, he pointed out that most private websites generally do not strip images of metadata like some commercial sites (e.g., Facebook) do. Indeed, Mr. Musters thought the image had originally been obtained from Facebook because of its reduced sized and because it was stripped of its metadata, a common feature of images uploaded to that site. Moreover, Mr. Musters saw the image on Facebook. In the end, Mr. Musters agreed that he did not know what www.g20justice.com did to the sizing of images. He agreed with Ms. Peloquin’s ultimate conclusion that there was no clear evidence of manipulation.
[22] Lastly, the defence also called Mr. John Sandeman to provide an opinion on the possibility of tampering. However, I need not deal with the substance of his evidence. During an impressive cross-examination by Mr. Perlmutter on Mr. Sandeman’s background and training, it became readily apparent that Mr. Sandeman was not qualified to provide an expert opinion in this case. During submissions, Mr. Black did not stress the point. I place no weight on his evidence.
ANALYSIS
(a) Framing the Issue
[23] As part of its application, the Crown seeks a ruling on the admissibility of the photograph and on the admissibility of Det. Costa’s evidence. The Crown prefers to deal with the two matters together because, unlike the complainant, Det. Costa is able to identify P.C. Andalib-Goortani in a number of photographs including, apparently, the image in question. However, the admissibility of Det. Costa’s evidence is governed by different principles and requires a separate voir dire to determine whether the proposed evidence meets the “prior acquaintance/better position” test: see R. v. Leaney, 1989 28 (SCC), [1989] 2 S.C.R. 393, R. v. Brown (2006), 2006 42683 (ON CA), 215 C.C.C. (3d) 330 (Ont. C.A.) and R. v. Berhe (2012), 2012 ONCA 716, 292 C.C.C. (3d) 456 (Ont. C.A.). These cases provide that, before the admissibility of that identification evidence is determined, the image(s) in question must first be authenticated. This decision deals only with the authentication issue.
(b) Onus and Threshold
[24] The Crown argues that, because photographic evidence is not presumptively inadmissible, it bears no burden in establishing the admissibility of the image in question. The Crown submits that the onus is on the defence to bring a motion to exclude the evidence. This is incorrect.
[25] Some types of evidence are considered presumptively inadmissible, such as similar fact evidence (R. v. Handy (2002), 2002 SCC 56, 164 C.C.C. (3d) 481 (S.C.C.)), hearsay (R. v. Baldree (2013), 2013 SCC 35, 298 C.C.C. (3d) 425 (S.C.C.)), prior consistent statements (R. v. Ellard (2009), 2009 SCC 27, 245 C.C.C. (3d) 183 (S.C.C.)) and statements of an accused made during a “Mr. Big” operation (R. v. Hart, 2014 SCC 52). Photographic evidence has not been so categorized. However, this does not mean that it is automatically admissible; instead, photographs are conditionally admissible. Certain pre-conditions must be “established” on the basis of “some evidence” before a photograph is admissible and made available to a witness[^2] and the trier of fact: see S.C. Hill, D.M. Tanovich and L.P. Strezos, McWilliams’ Canadian Criminal Evidence, 5th ed. (Toronto: Canada Law Book, 2013) (looseleaf), at pp. 23-5 to 23-9.
[26] This proposition is demonstrated in R. v. Nikolovski (1996), 1996 158 (SCC), 111 C.C.C. (3d) 403 (S.C.C.), in which the Court considered the admissibility of videotape evidence. Writing many years ago, Cory J. said the following at p. 416:
Once it is established that a videotape has not been altered or changed, and it depicts the scene of a crime, then it becomes admissible and relevant evidence. Not only is the tape (or photograph) real evidence in the sense that that term has been used in earlier cases, but is to a certain extent, testimonial evidence as well. [Emphasis added.]
See also R. v. Penney (2002), 2002 NFCA 15, 163 C.C.C. (3d) 329 (Nfld. & Lab. C.A.), at pp. 335 and 342.
[27] In a criminal trial, it would be improper for counsel to wave a photograph around in front of the jury, or thrust it under the nose of a witness, without first addressing the issue of authenticity. The potential for unfairness is obvious. In this case it is very real. The party wishing to make use of a photograph bears the burden of authentication, not the other way around.
(c) Authentication
[28] The leading Canadian case on authenticating images is R. v. Creemer and Cormier, 1967 711 (NS CA), [1968] 1 C.C.C. 14 (N.S.S.C. App. Div.). McKinnon J.A. noted the following requirements for authentication at p. 22:
All the cases dealing with the admissibility of photographs go to show that such admissibility depends upon (1) their accuracy in truly representing the facts; (2) their fairness and absence of any intention to mislead; and (3) their verification on oath by a person capable of doing so.
This formulation has been widely accepted in many subsequent decisions and by numerous commentators: see, for example, R. v. Maloney (No. 2) (1976), 1976 1372 (ON SC), 29 C.C.C. (2d) 431 (Ont. Co. Ct.), R. v. Penney, supra, R. v. Schaffner (1988), 1988 7108 (NS CA), 44 C.C.C. (3d) 507 (N.S.C.A.), at pp. 509-511, R. v. J.S.C., 2013 ABCA 157, [2013] A.J. No. 455 (C.A.), R. v. Adams (2011), 2011 NSCA 54, 274 C.C.C. (3d) 502 (N.S.C.A.), Sydney N. Lederman, Alan W. Bryant and Michelle K. Fuerst, The Law of Evidence in Canada, 4th edition (Toronto: LexisNexis, 2014), at pp. 44-45 and pp. 1294-1296, David Watt, Watt’s Manual of Criminal Evidence, 2013 (Toronto: Thomson Reuters, 2014), at p. 88 and David Paciocco, The Law of Evidence, 6th ed. (Toronto: Irwin Law, 2011), at p. 462.
[29] The Crown’s attempt to authenticate the image in this case may be addressed by the application of the second criterion -- fairness and absence of an intention to mislead. I conclude that the Crown has failed to establish that the image was not tampered with or altered before it came into its possession.
[30] Both Ms. Peloquin and Mr. Musters agree that some properties of the image have been altered through the process of being uploaded. However, neither could say to which site it was initially uploaded. Neither is able to discern whether the image was automatically stripped of its metadata during this process, or whether it was intentionally removed (and by whom). Nor can the experts determine how many websites and/or computers this image has been uploaded to or downloaded from.
[31] I appreciate that Ms. Peloquin has more formal training than Mr. Musters in this area. However, Mr. Musters’ practical experience over the years equally qualifies him to provide an opinion on this matter. In his report and his testimony, Mr. Musters expressed more concern than Ms. Peloquin that the image had been tampered with or altered, and in more serious ways. He had the image analyzed by FOURMatch, which Ms. Peloquin did not. He refused to use so-called “freeware” (i.e., free software available from the internet) because of his concerns about its reliability, whereas Ms. Peloquin relied on at least one of these free products (“JPEGsnoop”). In the end, I am not prepared to discount the weight of Mr. Musters’ opinion in the manner that the Crown submits.
[32] Some of the observed changes to the image may, at first blush, appear innocuous. However, given that the image has been changed in these ways (for reasons and by persons unknown), there is a lingering concern that it has been manipulated in other ways, ways that are intended to distort the true state of affairs that the image purports to capture. Neither expert could say that the image has not been altered in this manner. Ms. Peloquin said that it would take considerable skill for someone to alter an image and elude forensic testing. Mr. Musters suggested it would take less skill to pull off such a feat, especially with the widely available Photoshop software.
[33] Materials taken from websites and offered as evidence in court must be approached with caution, especially in a case such as this where no one is prepared to step forward to say, “I took that photo and it has not been altered or changed in anyway.” Several U.S. cases warn about the possibility of tampering in this context. In People v. Beckley, 110 Cal. Rptr. (3d) 362 (Ct. App. 2010), the Court expressed concern about the dangers of unauthenticated digital images at pp. 515-516:
Recent experience shows that digital photographs can be changed to produce false images .… Indeed, with the advent of computer software programs such as Adobe Photoshop “it does not always take skill, experience, or even cognizance to alter a digital photo.” (Parry, Digital Manipulation and Photographic Evidence: Defrauding The Courts One Thousand Words At A Time” (2009), 2009 J.L. Tech. & Pol’y 175, 183).
Similar concerns were expressed in St. Clair v. Johnny’s Oyster & Shrimp, Inc., 76 F. Supp. (2d) 773 (Tex. Dist. Ct. 1999), Griffin v. State, 19 A. (3d) 415 (Ct. App. Md. 2010) and People v. Lenihan, 911 N.Y.S. (2d) 588 (Sup. Ct. N.Y. 2010). These common sense warnings take on special significance when eyewitness identification is critical, as it is in this case.
[34] The Crown has failed to satisfy me that the image has not been tampered with or altered in some material way. For this reason alone, the image cannot be authenticated. In the circumstances, it is not necessary to addresses the other authentication criteria mentioned above.
CONCLUSION
[35] The photograph has not been properly authenticated. It is not admissible at P.C. Andalib-Goortani’s trial. The Crown’s application is dismissed.
___________________________ Trotter J.
Released: August 14, 2014
COURT FILE NO.: CR-12-00000841-40
DATE: 20140814
ONTARIO
SUPERIOR COURT OF JUSTICE
BETWEEN:
HER MAJESTY THE QUEEN
Applicant
– and –
Babak Andalib-Goortani
Respondent
REASONS FOR PRE-TRIAL RULING
Trotter J.
Released: August 14, 2014
[^1]: The matter was argued before me as a pre-trial motion. A jury trial is expected to take place later this year.
[^2]: Assuming the witness is not involved in authenticating the image.