Neville v The Queen

[2004] WASCA 62

NEVILLE -v- THE QUEEN [2004] WASCA 62

Jurisdiction:	SUPREME COURT OF WESTERN AUSTRALIA	Citation No:	[2004] WASCA 62
Court:	COURT OF CRIMINAL APPEAL
Case No:	CCA:150/2003	Heard:	18 FEBRUARY 2004
Coram:	MALCOLM CJ MILLER J EM HEENAN J	Delivered:	2/04/04
No of pages:	44	Judgment Part:	1 of 1
Result:	Extension of time to appeal granted, Appeal allowed, Retrial ordered
Category:	A
PDF Version

Parties:	JACQUELINE MARGARET NEVILLE THE QUEEN
Catchwords:	Criminal law and procedure Application for extension of time to appeal against sentence Lengthy delay Test to be applied Criminal law and procedure Evidence Identification Identification based on knowledge of suspect Voice recognition Whether opinion evidence Admissibility of Whether jury in equal position as witnesses to make comparison Adequacy of trial Judge's direction to jury Whether the ultimate issue of identification taken away from jury
Legislation:	Nil
Case References:	Alexander v The Queen (1981) 145 CLR 395 Bulejcik v The Queen (1996) 185 CLR 375 Cameron v The Queen [2004] WASCA 16 Chamberlain v The Queen (1984) 153 CLR 521 Domican v The Queen (1992) 173 CLR 555 Festa v The Queen (2001) 208 CLR 593 Greaves v Aikman (1994) 74 A Crim R 370 Li v The Queen [2003] NSWCCA 290 Miladinovic v The Queen (1993) 47 FCR 190 Mraz v The Queen (1955) 93 CLR 493 Nguyen v The Queen (2002) 26 WAR 59 Nguyen v The Queen [2002] WASCA 181 R v Brotherton (1992) 29 NSWLR 95 R v Brownlowe (1986) 7 NSWLR 461 R v Bueti (1997) 70 SASR 370 R v Callaghan (2001) 4 VR 79 R v Callaghan [2001] 4 VR 79 R v Corke (1989) 41 A Crim R 292 R v Harris [No 3] [1990] VR 310 R v Hentschel [1988] VR 362 R v Jones and Harris (1989) 41 A Crim R 1 R v Kelly (2002) 129 A Crim R 363 R v Leung (1999) 47 NSWLR 405 R v McHardie [1983] 2 NSWLR 733 R v Miladinovic (1992) 109 ACTR 11 R v Omar [1991] 58 A Crim R 139 R v Smith [1984] 1 NSWLR 462 Smith v The Queen (2001) 206 CLR 650 R v Brotherton [1992] 65 A Crim R 301 R v Walker (1998) 101 A Crim R 152

JURISDICTION : SUPREME COURT OF WESTERN AUSTRALIA TITLE OF COURT : COURT OF CRIMINAL APPEAL CITATION : NEVILLE -v- THE QUEEN [2004] WASCA 62 CORAM : MALCOLM CJ

MILLER J
EM HEENAN J

HEARD : 18 FEBRUARY 2004 DELIVERED : 2 APRIL 2004 FILE NO/S : CCA 150 of 2003 BETWEEN : JACQUELINE MARGARET NEVILLE

Appellant

AND

THE QUEEN
Respondent

ON APPEAL FROM:

Jurisdiction : SUPREME COURT OF WESTERN AUSTRALIA

Coram : McKECHNIE J

File Number : INS 169 OF 2000

Catchwords:

Criminal law and procedure - Application for extension of time to appeal against sentence - Lengthy delay - Test to be applied

(Page 2)

Criminal law and procedure - Evidence - Identification - Identification based on knowledge of suspect - Voice recognition - Whether opinion evidence - Admissibility of - Whether jury in equal position as witnesses to make comparison - Adequacy of trial Judge's direction to jury - Whether the ultimate issue of identification taken away from jury

Legislation:

Nil

Result:

Extension of time to appeal granted

Appeal allowed
Retrial ordered

Category: A

Representation:

Counsel:

Appellant : Mr D Grace QC
Respondent : Mr K P Bates & Mr T B L Scutt

Solicitors:

Appellant : Amidzic & Associates
Respondent : State Director of Public Prosecutions

Case(s) referred to in judgment(s):

Alexander v The Queen (1981) 145 CLR 395
Bulejcik v The Queen (1996) 185 CLR 375
Cameron v The Queen [2004] WASCA 16
Chamberlain v The Queen (1984) 153 CLR 521
Domican v The Queen (1992) 173 CLR 555
Festa v The Queen (2001) 208 CLR 593
Greaves v Aikman (1994) 74 A Crim R 370
Li v The Queen [2003] NSWCCA 290


(Page 3)

Miladinovic v The Queen (1993) 47 FCR 190
Mraz v The Queen (1955) 93 CLR 493
Nguyen v The Queen (2002) 26 WAR 59
Nguyen v The Queen [2002] WASCA 181
R v Brotherton (1992) 29 NSWLR 95
R v Brownlowe (1986) 7 NSWLR 461
R v Bueti (1997) 70 SASR 370
R v Callaghan (2001) 4 VR 79
R v Corke (1989) 41 A Crim R 292
R v Harris [No 3] [1990] VR 310
R v Hentschel [1988] VR 362
R v Jones and Harris (1989) 41 A Crim R 1
R v Kelly (2002) 129 A Crim R 363
R v Leung (1999) 47 NSWLR 405
R v McHardie [1983] 2 NSWLR 733
R v Miladinovic (1992) 109 ACTR 11
R v Omar [1991] 58 A Crim R 139
R v Smith [1984] 1 NSWLR 462
Smith v The Queen (2001) 206 CLR 650

Case(s) also cited:

R v Brotherton [1992] 65 A Crim R 301
R v Walker (1998) 101 A Crim R 152


(Page 4)

1 MALCOLM CJ: This was an application for an extension of time within which to appeal against conviction. In my opinion, the extension of time should be granted, the appeal allowed, the conviction quashed and an order made that there be a retrial. I have reached this conclusion for the reasons to be published by Miller J and Heenan J. There is nothing I wish to add.

2 MILLER J: The appellant was indicted in the Supreme Court in the name of Jacqueline Margaret Hinchliffe on a charge that on 25 February 2000 at Mirrabooka she wilfully murdered Michael Ian Wright. To this charge the appellant pleaded not guilty and she was tried before McKechnie J and a jury between 14 and 26 June 2001. She was found guilty by the jury of the charge of wilful murder and sentenced to life imprisonment, with a minimum term to be served before eligibility for parole of 15 years. The sentence dated from 27 February 2000.

3 On 16 September 2003, more than two years after the conviction, the appellant purported to file a notice of appeal against conviction. By reason of the fact that the notice of appeal was so far out of time, an application for extension of time within which to appeal was filed on 25 November 2003. It was supported by an affidavit of the appellant's solicitor, in which it was explained that the delay in the filing of the notice was due to difficulties with funding of the appeal, for which ultimately legal aid was granted on 7 July 2003, but only on a very limited basis. It transpires that on 16 September 2003 Mr Grace QC, counsel for the appellant at this hearing, agreed to appear on her behalf and argue her appeal on limited legal aid funding.

Extension of time

4 There is no doubt that there has been a very lengthy delay between the appellant's conviction and the filing of a notice of appeal against conviction. The principles upon which an extension of time within which to appeal will be granted were recently restated by this Court in Cameron v The Queen [2004] WASCA 16 by McKechnie J (with whom Steytler J and Wallwork AJ agreed) at [28]:

"The principles governing the grant of leave when there has been a long delay have been settled: Narkle v The Queen, unreported; CCA SCt of WA; Library No 6108; 2 December 1985. If there has been a long and unexplained delay exceptional circumstances must be shown before an extension of time will be granted unless it can be demonstrated that there

(Page 5)

will be a miscarriage of justice if an extension is not be granted."

It is thus necessary to consider whether the conviction of the appellant constituted a miscarriage of justice.

Grounds of appeal

5 The appellant seeks to appeal against her conviction on two grounds. As amended by leave during the hearing, they are:

"1. The Learned Trial Judge erred in law in allowing into evidence the testimony of three police witnesses namely Detective Sergeant Darryl Johns Noye, Detective Senior Constable Giles Timothy Pargin and Senior Constable Joanne Lee McCabe concerning the purported identification of the appellant on a surveillance video recording taken at the front door of the deceased's premises on the 25th February 2000, on the basis that such evidence was irrelevant and/or inadmissible pursuant to the principles enunciated in Smith v The Queen (2001) 181 ALR 354, or otherwise and/or should have been excluded in the exercise of the learned trial Judge's discretion.

2. Further and/or in the alternative, the Learned Trial Judge erred in his directions to the jury concerning the identification evidence of the said three police witnesses in that he should have directed the jury that such evidence was irrelevant and should be excluded from its consideration in its deliberations and/or to have directed the jury that the determination of the issue of the identification was the ultimate issue."

The evidence at trial

6 The evidence at the trial established clearly that at about 8.20pm on 25 February 2000 Michael Ian Wright (the deceased) was living at the residence of his parents at 113 Apple Blossom Drive, Mirrabooka. He answered the front door of the house in response to a ring on the doorbell and when he did so, he was shot in the chest at close range. The person at the front door ("the shooter") fired a total of seven shots. The first of these hit the deceased in the left side of the chest and the injury sustained by the deceased was fatal. He died within minutes of being shot and prior

(Page 6)

to medical help arriving. The cause of death was found by a doctor on post-mortem examination to be death from a gunshot wound to the chest. The bullet which entered the deceased's chest was a 0.22 calibre round.

7 The jury saw a graphic video of the shots fired at the deceased. It was a surveillance video from a surveillance camera which had been installed at the house where the deceased lived. The video itself was extremely short. It showed a person walk to the front door of the premises where the deceased was residing and either knock at the door or ring the bell. After a short period of time somebody opened the door but not the locked security screen door. There was a dispute between the Crown and the defence as to what words can be heard on the video, but the Crown case was that the exchange was as follows:

"Deceased: Hello.

Shooter: How ya goin' mate, not too bad?

Deceased: What do you want?

Shooter: You."

The video footage showed the shooter produce a short firearm, raise it and discharge numerous shots. The shooter then turned and ran from the scene.

8 At the hearing of the appeal counsel for the appellant conceded that at least the words "What do you want" and "You" were spoken, but there was no agreement as to what the preceding words were. In my view nothing turns on whether or not they were accurately transcribed in the manner in which I have set out.

9 The shooting of the deceased was, on any view of it, a cold-blooded and ruthless execution. Whoever did it was clearly guilty of the crime of wilful murder. The question was whether it was the appellant who was the shooter. The Crown case was that the jury could be satisfied beyond reasonable doubt that it was the appellant because three police officers were able to identify the shooter on the surveillance video as one and the same person as the appellant.

Evidence of identification

10 The first police officer to testify that he could identify the appellant as the shooter in the surveillance video was Detective Sergeant Darryl John Noye. He said that he had watched the video of the shooting shortly

(Page 7)

after midnight on 26 February 2000 and upon viewing it was able to identify the shooter as the appellant. However, the identification that he made was dependant upon the audio track on the video and he quite frankly conceded that he did not believe he could have identified the shooter as the appellant without that sound. He said it was a combination of the visual and auditory factors that enabled him to reach the conclusion he did. Noye testified that he listened to the appellant's voice over an extensive period of time whilst monitoring intercepted telephone calls. It transpired that he had listened to a total of 78 intercepted phone calls between the appellant and her then husband Mark Hinchliffe. He had in fact listened to more calls but was only able to say definitely that he had listened to 78 because he had printed off a hard copy of the summary of each of those 78 calls. Twenty-one of those 78 calls were played to the jury. They took 3 hours, 11 minutes and 17 seconds to play.

11 In addition to listening to these calls, Noye had spoken to the appellant for a period of 40 minutes on 17 November 1999. This was at the Nollamara Women's Refuge in Stirling and at that time, Noye noted that the appellant had her head shaved and she was wearing a beanie. The shooter depicted on the video was wearing a beanie.

12 It is important to note that Noye was responsible for transcribing many of the intercepted calls upon which the appellant's voice was heard. He testified that he spent two weeks transcribing calls between the appellant, Mark Hinchliffe and the deceased, which meant that he had spent much more time listening to the voice of the appellant than was available to the jury in the context of the tapes being played in open Court. When the jury retired to consider its verdict in the case, they had several hours within which to do so, but it is unknown how much time was spent by the jury listening to those tapes.

13 Subsequently, on 15 and 16 December 1999 Noye saw the appellant at a Court hearing when her then husband, Mark Hinchliffe, was applying for bail. He spoke to her on that occasion.

14 Detective Senior Constable Giles Timothy Pargin was the second identification witness called. He testified that he was asked by another Detective to look at a video which he knew related to the investigation into the death of the deceased. He was asked by an officer to give an objective view as to who he thought the person on the video was. At that time, Pargin knew that the appellant had been charged with wilful murder and that the investigating officers were looking for police officers who had some familiarity with the appellant. He said that he was able to

(Page 8)

identify the voice of the appellant on the video, and together with the physical shape or general characteristics and body movements of the shooter, he had no doubt that it was the appellant. He explained that he had spoken to the appellant on two occasions and he described her voice as having a strong Australian accent. He conceded that he would not have been able to identify the shooter on the video of the appellant without hearing the audio track on the video. His evidence on the point was:

"Were you able to identify anyone on that video? --- Yes, I watched that videotape several times and combining - I identified the voice and I believe that the voice is the voice of Jacqueline Margaret Hinchliffe, and together with the physical shape of that person, combined with the voice, I have no doubt that that's who it is."

15 In cross-examination of Pargin counsel for the appellant extracted from him more detail as to how he was able to identify the physical shape of the shooter on the video as that of the appellant. He said:

"The shape of the legs, what was it about the shape of her leg in that video that tells you anything - suggesting it was Jacqui Hinchliffe? --- It's - I recall though the legs on the video as being the same shape. I mean, there's different types of legs - the same shape of legs that I'd seen on ---"

16 When Pargin was shown a video which was admitted to show the appellant talking with the deceased at the same doorway some months earlier, he gave the following answer:

"That comparative tape, the last tape that you have just watched here in court. Do you say that the legs of Jackie Hinchliffe on the left-hand side - that's a video of the same doorway obviously a couple of months earlier, 3 months earlier. You say that the legs looked the same as that of the assailant? --- Yes, I do."

17 Pargin added:

"All right. Now, one of the things you said about visual similarities was the way she turns or the way she turned. Right? You mean when she turned and ran? --- Yeah, turned and ran.

Assuming it was a female, the way the person turned and ran looked to you like the way Jackie Hinchliffe, what, turns and runs? --- Sorry, what's the question?

(Page 9)

Well, is the way that the assailant turned and ran in that tape - do you say that's the same way that Jackie Hinchliffe turns and runs? --- I say that ---

Yes, go on? --- As she turns towards the camera and runs, I still consider that that's her shape, yes."

18 Pargin conceded that he was unable to point to any other occasion upon which he had seen the appellant turn and start running, but he explained what association he had had in the past with the appellant:

"In the course of your police duties did you come across the accused Jacqueline Margaret Hinchliffe in connection with the towing business? --- Yes, I did.

Are you able to tell the ladies and gentlemen between what period of time you came into contact with Ms Hinchliffe? --- I first became aware of Mrs Hinchliffe in about 1996.

Over what period of time did you have contact with her? --- Up until about 1998, late 98.

And between 1996 and 1998 was that contact that you had in connection with her association with the towing business? --- Yes, in that period."

19 Pargin had also seen the appellant on three specific occasions in 1999. On 24 September 1999 he spoke to her for a period of about 10 minutes when she was in hospital; on 4 November 1999 he spoke to her face to face between five and 10 minutes on an occasion where she was wearing a baseball cap and her hair was extremely short; and on 24 November 1999 he observed her for a period of between five and 10 minutes when she attended the Rockingham police station.

20 Senior Constable Joanne Lee McCabe gave evidence that she became aware of the fact that investigating officers wished to speak to people who had previous dealings with the appellant, and she contacted a Detective Wright who arranged for her to view the video depicting the shooting. McCabe identified the shooter on the video as the appellant. She said in evidence that she knew the appellant was a suspect at the time, but did not know that she had actually been charged. She was able to recognise the voice of the appellant, which she described as a nasally female voice, and she said that the word "mate" was a common word used by the appellant. She also observed visual similarities between the

(Page 10)

shooter and the appellant in terms of weight, size and height. She conceded, however, that she could not have made the identification without hearing the voice of the shooter.

21 McCabe gave evidence that like Pargin, she was familiar with the appellant from her attendance at traffic accidents where the appellant also attended as a tow-truck operator. She said:

"Are you able to estimate for the ladies and gentlemen of the jury, in the period that you were at the Rockingham traffic branch from December 1993 through until January 1997, on average how many occasions you would see Jacqueline Margaret Hinchliffe at traffic crashes that you attended in your duties as a traffic officer attached to Rockingham traffic branch? --- Approximately, it would be approximately four times a month.

And just tell us, on some of those occasions, would you speak to her on all of those occasions or only some? --- No, only some. The tow truck - if an accident was called over the radio for us to attend there would be more than one tow truck there, so even if the vehicle wasn't towed by Wizard Towing or Rockingham Towing they would always be there because it was always a race to get there and see who could get the tow, so on all those occasions you would see the tow truck drivers, often greet them, say hello. You might not be engaged in conversation with them long because they might not have the tow, but you would also see them and always say hello and talk to them, and so I'd say on most occasions, yes, you would say something."

22 McCabe also testified of her knowledge of the appellant from seeing her around the Rockingham area:

"Thank you. You have mentioned also that you have lived in Rockingham for some considerable period of time and also worked there for some considerable period of time. As well as seeing her at crash scenes in your capacity as an investigator with the traffic branch, did you also see her around the Rockingham area from time to time? --- From time to time, yes.

Are you able to give us some estimation of the period over which you saw her? --- Maybe four, five occasions. You could pass her driving the car. You could pass her in the shop."

(Page 11)

23 Each of the witnesses Noye, Pargin and McCabe was strenuously cross-examined by counsel for the appellant. It is unnecessary to refer to the detail of the cross-examination, but their identification evidence was, in each case, the subject of particular scrutiny by cross-examining counsel. Inconsistencies were revealed. For example, Noye conceded in one passage in cross-examination that the person at the door of the time of the shooting looked quite different to the "slim, slender (appellant) shown on 27 November".

Other evidence

24 There was evidence other than identification evidence led by the Crown to implicate the appellant in the commission of the crime alleged. Most of it was circumstantial evidence.

25 The Crown led evidence that the deceased and the appellant had had an affair at a time when the appellant was married to Mark Hinchliffe, who was a member of the Coffin Cheaters motorcycle gang. Mark Hinchliffe had assaulted the appellant, leading to her hospitalisation in September 1999 and in November 1999 March Hinchliffe visited the deceased at his place of work and assaulted him. He then sought payment of financial compensation from the deceased for the affair he had had with the appellant. Threats were made to the deceased's father by a friend of Mark Hinchliffe. This had led the deceased to move to his father's house in Mirrabooka.

26 In November 1999 the appellant was admitted to the Stirling Women's Refuge where she remained for a period of some nine days. She had been badly beaten by her husband but apparently declined to implicate him in any statement.

27 By reason of the threats that had been made by Mark Hinchliffe and/or persons on his behalf to the deceased, a camera and audiotape were installed at the house in which the deceased was living at Mirrabooka. The camera had later been replaced by the deceased's father.

28 On 27 November 1999 the appellant attended at the house in Mirrabooka. There was a long video tape with audio showing her outside the door and her voice was clearly identifiable. She was wearing a baseball cap at the time and no hair was visible.

29 On 28 November 1999 the appellant attended at the Women's Refuge and told a worker that "she was going to be in a lot of trouble because she

(Page 12)

had been sent to kill somebody but she couldn't kill them". At that time she was observed to have a shaven head.

30 Telephone intercepts of calls between the appellant and Mark Hinchliffe revealed some damning admissions against interest made by the appellant. For example, on 28 November 1999 she said to Mark Hinchliffe "I failed tonight … I felt I was getting nowhere and I thought I would rather disappoint you than sit in a fucking gaol by killing that cunt who thought he could fucking walk around … your fucking town".

31 There was another video of the appellant's attendance at the deceased's home. This was on 26 December 1999 but it was a very short visit. At that time the appellant was wearing a baseball cap.

32 On 24 February 2000 the appellant hired a Corolla Seca vehicle from Quality Auto Rentals in Rockingham. She returned it on 26 February. It was seized by police that day. Swabs were taken for DNA and gunshot residue from the interior of the vehicle and a speck of possible gunshot residue was located on the steering wheel cover of the car, with further residue located in the glovebox. A forensic chemist conceded that particles of gunshot residue could remain in situ for a very long time if they were not subject to disturbance and when the appellant's hands were swabbed for gunshot residue on 26 February 2000 there was no residue found. Nor was there any residue found on clothing seized from her house.

33 There was evidence that the appellant had attended a firearm training course in 1996 but the last attendance at that course was on 30 November 1996.

34 There was other evidence which it is unnecessary to recount, because its probative value was limited.

Ground 1 - Admissibility of the identification evidence

35 Although the evidence of the three police officers was described as identification evidence, it seems to me that in reality the evidence of Noye was primarily of voice recognition and the evidence of Pargin and McCabe was primarily evidence of recognition rather than identification. Evidence of identification generally relates to a situation in which a stranger has observed a suspect at a crime scene and later identified that suspect from an identification parade, photograph or even (in some cases) dock identification.

(Page 13)

36 The distinction between various forms of identification evidence was made clear by McKechnie J in R v Kelly (2002) 129 A Crim R 363 at [33] et seq. McKechnie J was in dissent in relation to the ultimate outcome of the case, but his Honour's treatment of the principles in relation to identification evidence constitute a valuable analysis of the issues concerned. His Honour said:

"Introduction

Identification evidence has troubled courts for more than a century because identification evidence often appears cogent and is generally given by a witness who is creditworthy and thus often persuasive. This can make the evidence dangerous because a believable witness may nevertheless be honestly mistaken.

A number of special rules have been developed dealing with different aspects of the problems arising from identification evidence. It is important to remember that the rules deal with different aspects. Each case must be examined in the light of the particular issue with which it is concerned.

Issues in identification evidence

In dealing with identification evidence at trial, and also its subsequent examination on appeal, it is important to isolate the particular issue or issues raised in the evidence. Different issues have given rise to different rules.

(a) Initial observation at the crime scene

Brief observation by stranger

Brief observation of a suspect at a crime scene who is a stranger to the observer: This evidence is generally admissible as part of the circumstances surrounding the alleged criminal transaction even if the observer is unable to positively identify the suspect on a later occasion. The witness may give evidence of matters of similarity or dissimilarity between the suspect and the accused person. The exercise of a judicial discretion to exclude the evidence is enlivened. However, generally, the evidence is regarded as probative and admissible.

(Page 14)

Observation of a suspect who is known to the observer

This is not evidence of identification but of recognition.

Recognition and identification

There is a difference between recognition and identification. A suspect known to a witness may be recognised by that witness at a place which, for convenience, I will refer to as the crime scene. While examination is undertaken by the court as to the circumstances under which the recognition at the crime scene was made - sometimes very similar to issues of identification - other problems associated with identification are absent. The witness had had a prior opportunity to be familiar with the suspect.

Identification occurs when the witness does not know or recognise the suspect who is observed by the witness at the crime scene. In these circumstances there is a later opportunity to see the suspect and for the witness to compare the features of the suspect on that later occasion with their (the witness') recollection of the person they observed at the crime scene.

This evidence is generally admissible. The conditions surrounding the observations at the crime scene may bear on the reliability of the evidence and may require a judge to give both general and specific warnings to the jury on the dangers of recognition.

Good circumstances of observation of a suspect by a stranger

Where an observer has a good opportunity to observe the suspect at a crime scene, the evidence is generally admissible, subject to a warning by a judge to a jury as to general and particular dangers in observation of the actions of strangers.

Observation by police officers

This is subject to the same general rules applicable to other witnesses though a trial judge in directing a jury may make a factual observation to the jury that honest police officers are likely to be more reliable than the general public, being trained and less likely to have their observations and recollections affected by the excitement of the situation.

(Page 15)

(b) The subsequent identification

The reliability of the identification

This includes questions as to the time elapsed between the actual observation and subsequent identification and the effect the time might have had on the observer's ability to accurately bring the features of the suspect to mind.

If the initial observation was short and the time elapsed was long, the judicial discretion to exclude the evidence may be enlivened. In any event, a trial judge is bound to direct a jury as to general and particular matters which may affect the identification.

The reliability of the identification of the suspect from a group

A properly conducted identity parade probably gives rise to the most confidence in an identification. A dock identification gives rise to the least. In between are a variety of identification methods which have received judicial consideration, including identification in a crowd, such as at a train station or shopping centre, identification outside or close to a court which the suspect may be attending, video recorded identification, photoboard identification, identification where the suspect is in company with police officers, identification where the suspect is alone, subsequent identification after the observer's prior exposure to news items on television, newspapers, photographs or artist's impressions, or photoboards.

Some of these methods when employed in a particular case may produce a result so lacking in probative value as to be inadmissible. Alternatively, the probative value may be so slight as to enliven the judicial discretion. Where there is some probative value in the evidence, however, the evidence is generally admissible."

37 Applying the principles enunciated by McKechnie J in Kelly v The Queen, it seems to me that the evidence of Pargin and McCabe was closer to evidence of recognition than it was to anything else. Both Pargin and McCabe knew the appellant from previous dealings with her. As it happened, they were not identifying her by seeing her in person at a crime scene, but rather by seeing her on the video record of a crime scene in which the appearance of the person responsible for the crime was

(Page 16)

combined with voice - although only a few words were spoken. It was a combination of visual and voice recognition. It was, however, based upon past recognition rather than what is sometimes termed "fleeting glimpse" identification. The evidence was, in my view, entirely admissible on this basis, and I do not consider anything said by the High Court in Smith v The Queen (2001) 206 CLR 650 to affect the position.

38 The evidence of Noye I would classify as evidence of "ad hoc expertise" and therefore opinion evidence. This is because Noye's evidence was primarily based upon his careful consideration of telephone intercepts on which the appellant's voice was recorded and by his comparison of what he had heard in numerous replays of those intercepts with what he heard on the video of the crime scene. It is true that Noye combined with voice recognition identification from the physical characteristics of the person shown on the video, but essentially his evidence boiled down to voice recognition evidence.

Admissibility of voice recognition evidence

39 Voice recognition was admissible in the circumstances of this case. This was made clear in Bulejcik v The Queen (1996) 185 CLR 375 per Brennan CJ at 381:

"… there is no general rule that precludes a jury from taking account of an accused's voice heard at the trial when it tends to prove a fact to be found. In R v Smith, Ashford and Schevella, the Court of Criminal Appeal of Victoria rejected a submission that the trial judge should not have directed the jury that they were entitled to make a comparison between the voice of an accused in making an unsworn statement and the voice heard on certain tapes, in order to assist in deciding whether the jury were satisfied as to the identity of the speaker whose voice could be heard on the tapes. The Court rejected the notion that the jury's making of that comparison might amount to an impermissible experiment.

Recognition of a speaker by the sound of the speaker's voice is a commonplace of human experience. To recognise the voice of a particular speaker some familiarity with that speaker's voice is ordinarily needed. A person who is not familiar with the voice of a putative speaker may be able nevertheless to recognise the speaker's voice by comparison with an established example of that voice if the speaker's voice exhibits sufficiently distinct

(Page 17)

features to permit an ordinary person to identify the speaker or if the person possesses an appropriate expertise."

40 At 382 his Honour added:

"In some cases, judges have treated prior familiarity or distinctiveness as conditions of admissibility of voice identification by non-experts in the absence of other means of identification; in other cases, familiarity and distinctiveness have been treated as factors relevant to the weight of the witness' evidence but not its admissibility. Evidence of identification by voice recognition is not a distinct category of evidence, though its probative value may oftentimes be dubious and will vary according to the circumstances of each case. The test of its admissibility must be, in my opinion, one of degree. The prescription of particular conditions of admissibility is not supported by any principle of the law of evidence. Provided a reasonable jury could find, or be assisted in finding, a relevant fact upon consideration of evidence of voice identification that is admissible under the ordinary rules of evidence, there is no reason why the tender should be rejected. The evaluation of evidence on which a reasonable jury could act is a matter for the jury."

41 In Nguyen v The Queen [2002] WASCA 181, Anderson J (with whom Malcolm CJ and Steytler J agreed) made it clear (at [138]) that voice recognition evidence is not necessarily an expert process. It is permissible for a jury to make its own comparison from the material before it. Indeed, in Nguyen v The Queen the Court held that there is no basis for a submission that a jury should be warned not to embark upon a process of comparison themselves. This point will be of relevance to the second ground of appeal.

42 There have been a number of cases in other States dealing with voice identification evidence. In R v Callaghan (2001) 4 VR 79, Winneke P (with whom Booking JA and O'Bryan AJ agreed) made reference to a possible difference between the approach of the Court of Criminal Appeal in New South Wales and the Court of Appeal in Victoria, on the question of the need for a witness identifying the voice of a suspect to be able to point to some distinctive feature about that voice. At [25] - [27] et seq, Winneke P highlighted the distinction in the two States as follows:

(Page 18)

"Mr Croucher submitted that, in the absence of evidence that the witnesses were familiar with the voice of the applicant and in the absence of material which disclosed that voice to have some very distinctive feature about it, the judge was obliged to exclude the evidence in the exercise of a 'Christie discretion'. In support of these submissions Mr Croucher relied upon, and urged this court to follow, a line of authority in New South Wales which appears to have commenced with R v Smith, in which O'Brien CJ Dr Div said at 477:

[E]vidence of the voice of a person present at a crime as being the same as the voice of the accused can only amount to positive identification where the witness is very familiar with the voice before hearing it at the crime, or that the voice heard at the crime was very distinctive, which means that the witness need not have heard the voice before the crime but heard it as the voice of the accused for the first time after the crime and then noted it to have the same very distinctive features as had the voice at the crime.

The case of Smith went to the Court of Criminal Appeal. Although the court quashed the conviction on the ground of misdirection, it approved the statement of law, to which I have referred, concerning the admissibility of voice identification evidence. It has since been followed in a series of cases in New South Wales.

It should be observed that the statement of law referred to in the preceding paragraph was made in the context of a case in which the evidence of voice identification was crucial to the proof of guilt of the accused. So it was in the case of Brownlowe, above, where Hunt J, who gave the leading judgment in the Court of Criminal Appeal, left open the question of the admissibility of voice identification or voice similarity evidence, where that evidence was led as some evidence in a case based upon circumstantial evidence. His Honour said at 476:


It should be said in parenthesis that this appeal is not concerned with any question of the admissibility of such evidence as to the similarity of the voices in support of a case based simply upon circumstantial evidence. That was not the Crown case which was left to the jury in this trial.

(Page 19)

In any event, the law applied in New South Wales as to the admissibility of voice identification or similarity evidence is not the law which has been consistently applied in this State, where questions of 'familiarity' and 'distinctiveness' have not been regarded as conditions precedent to admissibility, but rather as matters going to the weight of the evidence; matters which are - subject to the judge's discretion - for the jury to decide in the light of adequate directions given to them by the judge, with the authority of his office, pointing to the dangers which the law recognises as lurking in any form of identification evidence. In R v Harris (No 3), Ormiston J (in a ruling which was in fact given on 26 March 1987) referred to the cases of Smith and Brownlowe, above, and said at 316:

There is, however, real difficulty, as a matter of the law of evidence, in ruling as inadmissible a statement by a witness that he or she identified a person or recognised a voice. If there is little reliability in the purported identification or recognition, then that goes to the weight of the evidence only, though various powers exercisable by the judge may be invoked in cases of prejudice or unfairness. Strangely, this is one area of the law where the often-praised commonsense of juries seems to be distrusted, even though the judge is required to add his warning to that of counsel.

His Honour went on to state that he was unable to accept that 'any authority binding on me requires me to rule that the present evidence is admissible', and that he did 'not propose to follow Smith's case'. His Honour 's view of the law of evidence relating to the admissibility of identification evidence was not only consistent with views previously stated by the Court of Criminal Appeal in this State, but it is a view which has been consistently followed thereafter. (In each of these cases, the court declined to follow the law as applied in R v Smith and R v Brownlowe, above.) The view taken by courts in this State appears to have been followed in the Australian Capital Territory and in Tasmania. Although the question of 'voice identification' was raised before the High Court in Bulejcik v R, the appeal was disposed of without having to determine the question whether admissibility of voice identification evidence is or should be conditional upon the distinctiveness of the speaker's voice or upon the witness's familiarity with that voice. Nevertheless it appears to me that the observations of the

(Page 20)

majority of the court favoured the view adopted in this State, without finally deciding the matter (see Brennan CJ at 382-3; McHugh and Gummow JJ at 4-6-7). However, as McHugh and Gummow JJ observed at 407:

… the correctness of Smith and the cases that follow it should await a case where a decision on the point is essential. It is not necessary to decide its correctness in this case.

This review of the authorities on the question of the admissibility of voice identification evidence leads me to conclude that courts in this State should continue to follow the principles which have been established by previous decisions of the Court of Criminal Appeal; namely that there is no rule of law which obliges the trial judge to exclude such evidence in the absence of evidence of prior familiarity or distinctiveness, although he may, in the exercise of his discretion, exclude it on grounds of prejudice or unfairness."

The Victorian view has been accepted in Queensland: R v Corke (1989) 41 A Crim R 292.

43 In the present case the witnesses Pargin and McCabe were able to testify to prior knowledge of the voice of the appellant, although Noye's voice identification was based primarily upon his comparison of voice intercept tapes with what he heard on the video at the scene of the crime. To the extent that he had no prior association with the appellant, it was, in my view, unnecessary for him to have done so for his evidence to be admissible. I consider the Victorian position to be that which should apply in this State.

Was the evidence of Noye evidence of opinion or fact?

44 The line between opinion evidence and evidence of fact is not always clearly defined. This was made clear in R v Leung (1999) 47 NSWLR 405 by Simpson J at [43]):

"The line between opinion evidence and evidence of fact is not always clearly defined. Evidence of physical identification illustrates the point. One the one hand such evidence may be characterised as evidence of fact; but, depending on the circumstances, it may more properly be characterised as evidence of opinion. The ordinary observer would regard

(Page 21)

evidence given by a man identifying his wife of thirty years as evidence of fact; but a witness who identifies a suspect in a police lineup would be perceived as giving evidence more closely allied to opinion evidence. Where the evidence is more properly seen as opinion evidence, it is frequently opinion evidence permitted to be given by a person without specialist qualifications, but whose specialised experience of the person of whom the identification is made qualifies him or her to give the evidence. It may be that this is a familiar example of an ad hoc expert."

45 Adopting this distinction, it seems to me that in the present case the evidence of Noye was probably closer to opinion evidence than it was to anything else, and the evidence of Pargin and McCabe closer to recognition evidence as a category of identification evidence. Perhaps the distinction is not vital, but I think it puts in context the basis upon which the evidence was led before the jury in this case.

46 I would only add that in Li v The Queen [2003] NSWCCA 290 at [42] Ipp JA (with whom Whealy and Howie JJ agreed) considered that an experienced interpreter and translator who was skilled in interpreting and translating the Cantonese and Mandarin languages into English, and also English into those languages, could give evidence of voice recognition from what he heard on a number of audio tapes when compared with an audio recording comprising part of a video tape of a police interview with the appellant. At [42] Ipp JA said:

"Butera and Leung, however, demonstrate that evidence of ad hoc expertise is admissible. The very many hours that Mr Chan spent listening to and working on the five tapes with a view to identifying the words spoken by M1 qualified him as an ad hoc expert in the characteristics of M1's voice."

In my opinion, the evidence of Noye fell within this categorisation.

The effect of the decision in Smith v The Queen (supra).

47 In Smith v The Queen the accused person had been indicted on a charge of bank robbery. Two police officers who had previously dealt with him gave evidence at trial that one of the robbers depicted in security camera photographs was the accused. The High Court held that the evidence of the police officers was inadmissible. Gleeson CJ, Gaudron, Gummow and Hayne JJ considered it inadmissible on the basis that it was irrelevant and Kirby J on the ground that though relevant, it was evidence

(Page 22)

of an opinion which by virtue of s 76 of the Evidence Act 1995 (NSW), was inadmissible to prove the existence of the fact about the existence of which the opinion was expressed.

48 The extent to which the police officers were familiar with the accused person is apparent from the reasons for judgment of Kirby J. The evidence of the officers was tendered as "recognition evidence" and it was contended by the prosecutor that the evidence was founded not only upon viewing the photographs but also upon prior familiarity by reason of the police officers having dealt with the accused person in the Redfern and Waterloo suburbs of Sydney. They said he had been seen there on numerous occasions and interviewed and on one occasion had been arrested and seen over an extended period of time.

49 Gleeson CJ, Gaudron, Gummow and Hayne JJ summarised the evidence of the police officers at [9] as follows:

"The only evidence led against the appellant in relation to [identification] was the evidence of the two police officers and the evidence that demonstrated that the photographs which were tendered in evidence had been taken by the bank's security cameras during the robbery. Neither police officer suggested that he had any basis for concluding that it is the appellant depicted in the bank photographs other than the knowledge that he had gained of the appellant's physical appearance during those earlier encounters and whatever was revealed to a person who looked at the photographs that were produced in evidence. There was no suggestion that the physical appearance of the appellant had changed materially between the time when the photographs were taken and the time of the trial, or that the police, by reason of their previous observations of the appellant, were at some advantage in recognising the person in the photographs. It was acknowledged by counsel, in the course of argument in this Court, that, by the time the evidence had concluded, the jurors had probably spent more time in the presence of the appellant than had the police witnesses before they gave their evidence. The police witnesses were in no better position to make a comparison between the appellant and the person in the photographs than the jurors or, for that matter, some member of the public who had been sitting in court observing the proceedings. If such a member of the public had been called as a witness, the same question of relevance would have arisen. Thus, not only was the issue that was raised a very

(Page 23)

narrow issue, the data available to the jury for its resolution was no different in any significant way from the date upon which the police officers based their asserted conclusion. The police officers' conclusions and the jury's conclusion both depended upon combining their observation of the photographs. (Having regard to the quality of the photographs we saw, it is not clear that the jury could not have compared them with the accused.)"

50 At [11] their Honours explained why the evidence of the police officers was irrelevant and should not have been received. As they put it, no question of admissibility had to be considered. Their Honours said:

"Because the witness's assertion of identity was founded on material no different from the material available to the jury from its own observation, the witness's assertion that he recognised the appellant is not evidence that could rationally affect the assessment by the jury of the question we have identified. The fact that someone else has reached a conclusion about the identity of the accused and the person in the picture does not provide any logical basis for affecting the jury's assessment of the probability of the existence of that fact when the conclusion is based only on material that is not different in any substantial way from what is available to the jury. The process of reasoning from one fact (the depiction of a man in the security photographs) taken with another fact (the observed appearance of the accused) to the conclusion (that one is the depiction of the other) is neither assisted, nor hindered, by knowing that some other person has, or has not, arrived at that conclusion. Indeed, if the assessment of probability is affected by that knowledge, it is not by any process of reasoning, but by the decision maker permitting substitution of the view of another, for the decision-maker's own conclusion."

51 However, at [15] their Honours pointed out that there will be cases in which evidence of identification will be relevant because it goes to issue about the presence or absence of some identifying feature other than one apparent from observing the accused on trial:

"In other cases, the evidence of identification will be relevant because it goes to an issue about the presence or absence of some identifying feature other than one apparent from observing the accused on trial and the photograph which is said to depict the accused. Thus, if it is suggested that the

(Page 24)

appearance of the accused, at trial, differs in some significant way from the accused's appearance at the time of the offence, evidence from someone who knew how the accused looked at the time of the offence, that the picture depicted the accused as he or she appeared at that time, would not be irrelevant. Or if it is suggested that there is some distinctive feature revealed by the photographs (as, for example), a manner of walking) which would not be apparent to the jury in court, evidence both of that fact and the witness's conclusion of identity would not be irrelevant. Similarly, if, as was the case in R v Tipene, there is an issue whether photographs of different incidents depict the same person, evidence given about the identity of the person depicted may not be irrelevant."

52 Thus it was that in Li v The Queen (supra), evidence of a detective was admitted in circumstances where he had made extensive observations of the accused person during surveillance operations between March and May 1988. From knowledge then gained he was able to testify as to the accused person's posture and manner of walking, both of which were distinctive. The witness highlighted the swinging of the accused's left arm, the holding of his head to the right and the manner of his walk in general. Ipp JA (at [106]) pointed out that the jury were not able to see the accused walking in Court and thus the evidence of the detective was not irrelevant and was properly admissible.

53 In the present case, there was, in my view, evidence of identification (or recognition) from the three police officers which was relevant because it went to an issue about the presence or absence of some identifying feature, other than one apparent from observing the accused on trial and the video of the shooting. That evidence consisted of the opinion evidence of Detective Sergeant Noye as to his voice recognition based as it was upon extensive consideration of telephone intercepts in which the appellant's voice was known, and a comparison with the voice of the appellant on the critical video. In the case of the other two police officers, there were important independent factors. Both identified the appellant's voice on the video from their prior dealings with the appellant, details of which I have summarised. Further, in the case of Pargin, he gave evidence about particular physical aspects of the appellant which would not have been readily apparent to the jury in the court room.

54 There is one additional factor. That is, that the appearance of the appellant had changed. At the time of trial her hair was shoulder length. Her appearance was quite different from what it was in November 1999

(Page 25)

when Noye observed at the women's refuge that the appellant had a shaved head. Photographs tendered in evidence and which had been taken on 26 February 2000 showed that the appellant had closely-cropped short hair.

55 When Pargin saw the appellant in November 1999 she was wearing a baseball cap and had extremely short hair. The person who is depicted on the shooting video was wearing a beanie. There was no visible hair. I accept the Crown's submission that Noye in particular, who had spoken at length to the appellant when her head was shaven, was able to make a comparison between the appellant with her shaven head and the person in the video whose hair was clearly covered by a beanie. The Crown's submission that the data available directly to the jury was by contrast that of a face framed by shoulder length hair is, in my view, a highly relevant factor. In terms of what the High Court said in Smith v The Queen (supra) at [15], there was a significant difference from the appearance of the accused shortly before the alleged commission of the crime by her and her appearance to the jury at her trial, some 16 months later.

56 The other distinguishing point about Smith v The Queen is that the evidence which the Crown led in that case was limited to the opinion of police officers that a person with whom they had previously dealt was one of the robbers depicted in a security camera photograph at the scene of the robbery. There was no question of voice identification. That factor is, in my view, a telling distinguishing factor.

57 I am therefore of the opinion that the evidence of the three police officers was relevant and admissible. I an unable to see any basis upon which, had it been objected to, the evidence should have been excluded in the exercise of the learned trial Judge's discretion. It was relevant evidence, probative and based on a combination of recognition of the appellant (varying from witness to witness) with voice identification from knowledge of the voice of the appellant from either past dealings with her or (in the case of Noye) listening for many hours to telephone intercepts on which her voice was played.

58 I would dismiss the first ground of appeal.

Ground 2 - Was the jury precluded from itself determining the issue of identification?

59 The first part of this ground repeats the contention that the identification evidence of the three police witnesses was irrelevant and should have been excluded from the jury's consideration. I have already

(Page 26)

dealt with this aspect of the ground. However, the alternative to the ground (added at the hearing) contends that the learned trial Judge ought to have directed the jury that the determination of the issue of identification was the ultimate issue. So framed, the ground does not quite do justice to the argument which was put. It was that the learned trial Judge had erred in effectively taking away from the jury their entitlement to make their own comparison of what they saw and heard on the video of the crime with (a) what they heard on the voice intercept tapes which were played to them; (b) what they saw and heard on the two videos of the appellant outside the deceased's home on 27 November and 26 December 1999, and (c) what they saw of the appellant as she sat in Court over the period 14 - 26 June 2001.

60 In the course of his comprehensive directions to the jury the learned trial Judge indicated to the jury that the two "big issues" were identification and intention. Identification, he said, was a key issue because there was no doubt that the deceased had been shot and killed and the question was whether the jury were satisfied beyond reasonable doubt that it was the appellant who had shot and killed him.

61 The learned trial Judge pointed out that the Crown had called three witnesses, each of whom gave evidence in varying ways that they knew the appellant and in particular, knew her voice, had watched the video and a combination of her features and the voice taken together satisfied them that it was the appellant on that video tape.

62 Proper directions were given by the learned trial Judge in relation to the care with which the jury should consider identification evidence and no objection has been taken to anything that was said in that regard.

63 The objection taken by counsel for the appellant is to the following passage in the learned trial Judge's directions to the jury:

"The question of identification, of course, is ultimately not for Detective Noye's opinion or Detective Pargin's opinion or Constable McCabe's opinion. It is for you to decide on the whole of the evidence - not just that but all of the evidence - whether the crown has satisfied you beyond reasonable doubt that it was the accused who was the shooter on the night.

You have the CD of many telephone conversations. You have the Douglas video which is the one which is enhanced and uses that split screen for a time. You have observed the accused in court, no doubt, for the last week but I caution you about all of

(Page 27)

that. She has largely been sitting. These are artificial circumstances and I would caution really to be very careful about using your own observations or playing detectives because your exposure to her has not been of the same order as the three witnesses who came before you. Use the Douglas video and the other matter but use it in order to assess the reliability and the honesty of the witnesses who have given their evidence, rather than playing detectives yourselves.

That's a caution which I give you but all of the evidence, having said that, is there available and ultimately how you use it is your affair but I do caution you about playing detectives. The preferred way really is to see whether you can be satisfied from the evidence of the witnesses as they have given it, rather than your own observations of the accused in court which, of course, is not actually sworn evidence."

64 The first paragraph of the directions cannot be criticised. Nor can the first six lines of the second paragraph. However, when the learned trial Judge told the jury to be careful about using their own observations or playing detectives, as he did, in the balance of the second paragraph and in the third paragraph of the directions, a very real question which arises is whether he taking away from the jury the entitlement that they had to themselves compare the materials I have identified with what they could observe on the critical video tape of the shooting.

65 The learned trial Judge unfortunately told the jury three times that they should avoid using their own observations or playing detectives themselves. However, in my view, this was a clear entitlement of the jury.

66 In Nguyen v The Queen (supra) it had been argued that a trial Judge should have given the jury a warning that they should not embark upon a process of voice recognition themselves, or alternatively, if they did so, to do so with extreme care. Anderson J said of this submission at [138]:

"I cannot accept the submission that the jury should have been warned not to embark upon a process of comparison themselves. I see no reason why the jury are not entitled to compare voice recordings in order to come to their own conclusions. Voice recognition is not, of itself, an expert process. As Brennan CJ said in Bulejcik v The Queen (supra) at 381:

(Page 28)

'Recognition of a speaker by the sound of the speaker's voice is a commonplace of human experience.'

It is clear that it is permissible for the jury to make their own comparison: R v Smith, Ashford and Schevella (1990) 50 A Crim R 434 at 453 - 454. See also Barker v The Queen [2002] WASCA 127 as to how the jury may use exhibits."

67 In the present case, the jury had sufficient materials from which they, themselves, could have embarked upon a process of voice identification. They did not have the appellant's voice in Court (save for the few words she spoke when she pleaded), as she gave no evidence at trial. They did, however, have what was clearly established to be her voice on a number of telephone intercepts and on the two videos of 27 November and 26 December 1999 respectively. To the extent that they had the appellant sitting in the dock before them, they also had her appearance (although different from what it was immediately prior to and after the crime allegedly committed by her) and they were entitled if they wished to embark upon a process of physical comparison between what they saw on the critical video and what they saw of her in the dock.

68 It is implicit in the judgments of Malcolm CJ and Anderson J in Nguyen v The Queen (supra) that the jury was entitled to enter into the exercise of voice recognition (see particularly Malcolm CJ at [78] and Anderson J at [138] - [140]). The judgment of the High Court in Smith v The Queen underlines the importance of the jury being entitled to reach a conclusion as to whether or not an accused person is guilty or not guilty from their own observation in cases in which the critical issue is the appearance of the person who committed a crime (whether depicted on photograph or video with audio) with other independent material available to it. In Smith v The Queen that was simply the appearance of the accused at trial. In this case, it was the appearance of the accused at trial but supplemented by the two video recordings of 27 November and 26 December 1999 and the many voice intercept tapes on which it was established that the appellant was speaking.

69 The evidence of the police officers was admissible and a matter for the consideration of the jury. To the extent that it was opinion evidence (in which category I would place the evidence of Noye) the learned trial Judge properly directed the jury that the ultimate issue was not the opinion of a detective but the decision of the jury itself on the whole of the evidence. It was evidence which may or may not have been helpful to the jury. However, the learned trial Judge erred in my view in telling the

(Page 29)

jury that it was the evidence of the three police officers which was the important evidence and that they should avoid playing detectives themselves. In the final paragraph of the direction that I have quoted, the learned trial Judge described it as "the preferred way" to see whether the jury could be satisfied beyond reasonable doubt from the evidence of the three police witnesses, rather than on their own observations of the accused in Court.

70 For the reasons that I have given, I consider that this was an erroneous direction. The jurors were always entitled to reach their own view on all of the evidence before them as to whether or not the Crown had proven beyond reasonable doubt that the person who carried out the shooting depicted in the video was one and the same person as the appellant.

71 In written submissions filed with the Court by its consent after the hearing of the appeal, the respondent argued that there is no authority for the proposition that a trial Judge must permit a jury itself to enter into the issues of voice and physical appearance comparison. It is said that the authorities stand only for the principle that it is not an error for a trial Judge to permit the jury to make the comparison. However, I disagree. In my opinion it is implicit in the decisions in Smith v The Queen and Nguyen v The Queen that the jury must, themselves, be allowed, to make the comparison. Indeed, in my view, the cases stand for the proposition that a jury must be informed that although there is evidence to assist them on the issue, it remains ultimately their decision and a decision which they can take, having regard to their own views on the matter from the material available in the Court, irrespective of the opinion or identification evidence which may have been adduced by the prosecution.

72 Although the learned trial Judge's direction that the jury should "avoid playing detectives themselves" was the only blemish in an otherwise model direction to a jury in a case of wilful murder, because of it, the appellant was denied a trial in which the relevant law was correctly explained to the jury and the rules of procedure and evidence strictly followed. It led, in my view, to the loss of a chance which was fairly open to the appellant of being acquitted and thus there was a miscarriage of justice: Mraz v The Queen (1955) 93 CLR 493 per Fullagar J at 514.

73 I would therefore grant an extension of time within which to appeal, and allow the appeal on the alternative to the second ground of appeal. I would quash the conviction of the appellant and order that she be retried.

(Page 30)
74 EM HEENAN J: I have had the advantage of reading in draft the reasons for decision of Miller J. These fully set out the background, the details of the trial and conviction of the applicant, the reasons for the delay in instituting the appeal and the submissions at the hearing of the application to extend time and in support of the appeal if time were to be extended.

75 For the reasons given by Miller J I too would extend the time within which to appeal to allow the appeal to be heard and determined on its merits. I can therefore pass directly to the two grounds of appeal raised. In summary, these are:

1. there was an error of law in admitting into evidence the testimony of three police witnesses concerning their identification of the appellant from the surveillance video recording of the offence, because this evidence was irrelevant or should have been excluded in the exercise of the learned trial Judge's discretion; and

2. the learned trial Judge erred in directing the jury about the identification evidence of those three police witnesses by failing to direct the jury that their evidence was irrelevant and should be excluded from consideration and/or because his Honour should have directed the jury that the determination of the issue of the identification [of the accused] was the ultimate issue [for the jury's decision].

The real issue at the trial was whether or not it had been proved that the appellant was the person who shot and killed Michael Ian Wright. The evidence in the case was overwhelming that, whoever the killer was, it was a case of wilful murder. In practical terms the issue became whether or not the prosecution had established that the person, depicted in the video surveillance film, standing at the front door of the deceased's father's home firing the pistol, was the appellant.

76 The prosecution set out to establish its case that the appellant was the killer in several ways. There was evidence from three police officers who, to varying degrees, had each had past dealings with or concerning the appellant. Each officer said that he or she identified the appellant as the killer from viewing the surveillance film and that this identification was a result of recognition of the image of the killer on the surveillance film and from recognising the voice of the killer, as recorded on the surveillance film. In this regard, however, the evidence of these officers was in effect that identification of the appellant as the person shown on the video was largely dependent upon recognition of her voice, rather than her image, on the surveillance film. Detective Sergeant Noye said that he

(Page 31)

did not believe that he could have identified the killer as the appellant without the voice recording and that it was a combination of the image and the voice recording which enabled him to make the identification. Detective S/C Pargin said that having watched the video several times the combination of the voice and the physical appearance of the killer allowed him to identify the appellant. As his cross-examination proceeded it became evident that Detective Constable Pargin must have relied more heavily on voice recognition than visual recognition to make the identification. Senior Constable McCabe also viewed the videotape and as a result identified the appellant as the killer because she was able to identify her voice and because she also observed similarities in appearance between the killer and the appellant. She readily accepted that she could not make the identification without recognising the appellant's voice.

77 In the case of Det Sgt Noye, his ability to recognise the voice of the appellant, as the voice recorded on the videotape, was because of his experience over a lengthy period of monitoring intercepted telephone calls which included calls to or from the appellant and the knowledge of her voice which he had accumulated by this process. He had listened to many such calls and could vouch for listening to 78 at least, but probably more. Those 78 calls had been recorded and 21 of them, with a total duration of over 3 hours and 11 minutes, were played to the jury. Noye had also spoken to the appellant on one occasion in November 1999 for a period of about 40 minutes when he was interviewing her at a women's refuge in Stirling. Not only had Noye listened to each of these 28 recorded calls and more but he had actually transcribed the calls from the telephone intercepts, spending some two weeks in doing so, so that he had become familiar with the voices on the recordings.

78 There was also other circumstantial evidence adduced by the prosecution at the trial which pointed towards the appellant being the killer. She had been involved in an affair with the deceased which was later discovered by her husband leading her husband to beat her viciously and later to assault and threaten the deceased and also to threaten his father. It was the deceased's father who owned the house at which the killing occurred where the deceased was staying temporarily in the hope of avoiding the angry husband. There was a lengthy videotape showing a visit to the house by the appellant, plainly identifiable, on an earlier occasion during which she attempted to persuade the deceased to come out of the house to speak to her and in the course of which it became apparent that she had some grievance of an unusual kind against him. There was evidence that the appellant told a worker at the women's refuge

(Page 32)

on 28 November 1999 "that she was going to be in a lot of trouble because she had been sent to kill somebody but she couldn't kill them". There were also telephone intercepts between the appellant and her husband in November 1999 in which she said "I failed tonight --- I felt I was getting nowhere and I thought I would rather disappoint you than sit in a --- jail by killing that ---- who thought he could --- walk around --- your --- town".

79 In addition there was evidence that, on the day before the killing, the appellant had hired a motor vehicle in Rockingham which was returned on the day after the killing. A police examination of the vehicle showed a small trace of possible gunshot residue on the steering wheel and in the glove box. There was evidence that the appellant had previous training in the use of firearms. All this suggested that, possibly at her husband's behest, and as a result of beatings which she had suffered from him because of the affair with the deceased, the appellant had planned killing the deceased. She had persisted in visiting him after the affair had been broken off, after both had been assaulted by the appellant's husband and when the deceased wished to have nothing to do with the appellant. Hence, there was some motive for a killing, an opportunity for that to be done and evidence which linked the appellant with the use of firearms and which suggested that there may have been a recently discharged firearm in the vehicle which she had on hire at the time this killing occurred.

80 The videotape which plainly identified the appellant as visiting the deceased's father's house during November 1999 and the audio recordings of the 78 intercepted telephone calls recording conversations with the appellant were all in evidence and were, respectively, played and showed to the jury. The members of the jury, therefore, clearly had an opportunity to watch the videotape of the killing themselves, to hear to the voice of the killer, to listen to the voice of the appellant as recorded in the earlier visit to the deceased's father's home and in the course of the intercepted telephone conversations. As a result the jury was in a position to evaluate whether or not those materials allowed the conclusion to be drawn that it was the appellant who fired the shots at the front door which killed the deceased. It seems highly likely that a decision by the jury about whether or not such a conclusion could be drawn, beyond reasonable doubt, from those materials would depend on the ability to make an identification of the appellant from the voice recorded on the videotape at the time of the shooting and, for that matter, upon the appearance of the person depicted in that videotape.

(Page 33)

81 All the indications are to the effect that the identification made from the videotape must have been based more on voice recognition than visual recognition.

82 It is important, however, that the other circumstantial evidence should not be overlooked because the verdict of the jury may be justified on the whole of the evidence, including the circumstantial evidence, or on such parts of it as the jury itself regarded as reliable and sufficient to persuade them beyond reasonable doubt on the issue of identity – Chamberlain v The Queen (1984) 153 CLR 521. If the evidence of the police officers by which they identified the appellant as the person shown on the videotape shooting the deceased had been wrongly admitted into evidence, or if a misdirection had been made by the learned trial Judge when instructing the jury about the significance of this evidence or how they may use it, then plainly an error would have occurred which would necessitate the quashing of the conviction and an order for retrial. The reason why a retrial should be ordered in that eventuality is because there was other evidence upon which a properly instructed jury could reach, if it was so satisfied, a verdict of guilty in this case.

First ground of appeal – the admission of the voice recognition evidence from the three police officers

83 In Festa v The Queen (2001) 208 CLR 593 McHugh J said, at 619:

"The risk of mistake in identifying a voice is at least as great as in identifying a person R v O'Sullivan (unreported, Qld Court of Appeal 21 July 1995 at 4). The reliability of voice identification varies with such factors as the length and volume of speech heard, the witness's familiarity with the accused's voice and the time elapsing between the occasions when the witness heard the voice of the perpetrator and the voice of the accused – cf Bulejcik v The Queen (1996) 185 CLR 375 at 381-2; 394-5; 406-7."

There can be no doubt about the admissibility of evidence of identification of the accused based upon recognition by a person of the accused's voice, either of the words actually spoken by the accused, or of some recording of them so long as it is established that the witness is able to make that recognition as a result of his or her own experience of the accused, whether before or after the event in issue, and so long as a direction suitable to the needs of the occasion is given to the jury. In Bulejcik v The Queen (1996) 185 CLR 375 at 381 Brennan CJ said:

(Page 34)

"Recognition of a speaker by the sound of the speaker's voice is a common place of human experience. To recognise the voice of a particular speaker some familiarity with that speaker's voice is ordinarily needed – R v Harris [No 3] [1990] VR 310 per Ormiston J at 318. A person who is not familiar with the voice of a putative speaker may be able nevertheless to recognise the speaker's voice by comparison with an established example of that voice if the speaker's voice exhibits sufficiently distinct features to permit an ordinary person to identify the speaker or if the person possesses an appropriate expertise."

However, there is a divergence of authority between the courts in New South Wales, on the one hand, and other courts in Australia, notably of Victoria, South Australia and Tasmania about whether or not the admissibility of voice identification evidence is conditional upon the witness being very familiar with the voice before hearing it at the crime scene, or at the time when identification is required, or conditional upon the voice which is heard on the occasion when identification is required being very distinctive.

84 The need for significant prior familiarity with the voice or some distinctive quality about the voice heard is regarded as a test for the admissibility of the voice identification in New South Wales and can be traced to the decision in R v Smith [1984] 1 NSWLR 462, a decision of O'Brien CJ of Cr Div. That approach has been approved in New South Wales in R v Brownlowe (1986) 7 NSWLR 461; and R v Brotherton (1992) 29 NSWLR 95. On the other hand, familiarity and distinctiveness have been treated as factors relevant to the weight of the witness's evidence but not to its admissibility in Victoria in R v Hentschel [1988] VR 362 at 364, 369; R v Jones and Harris (1989) 41 A Crim R 1 at 6 and 7; R v Harris [No 3] [1990] VR 310; R v Omar [1991] 58 A Crim R 139 and 146 – 147; R v Miladinovic (1992) 109 ACTR 11 at 15 – 16; Greaves v Aikman (1994) 74 A Crim R 370 and R v Callaghan [2001] 4 VR 79. That approach has also been taken in Tasmania in Greaves v Aikman (supra) at 378 – 379 and, to some extent, in Queensland in R v Corke (1989) 41 A Crim R 292. These differences, while noted by the High Court in Bulejcik v The Queen (supra) were not resolved by that decision because the use of the voice recording in the particular circumstances of that case was held to be improper for different reasons. However, when speaking of this controversy Brennan CJ said, at 382:

"Evidence of identification by voice recognition is not a distinct category of evidence, though its probative value may often

(Page 35)

times be dubious and will vary according to the circumstances of each case. The test of its admissibility must be, in my opinion, one of degree. The prescription of particular conditions of admissibility is not supported by any principle of the law of evidence. Provided a reasonable jury could find, or be assisted in finding, a relevant fact upon consideration of evidence of voice identification that is admissible under the ordinary rules of evidence, there is no reason why the tender should be rejected. The evaluation of evidence on which a reasonable jury could act is a matter for the jury. It exceeds a judge's function to withhold evidence from a jury merely because, on that evidence, the judge would not reach and thinks a jury should not reach a conclusion adverse to the accused beyond reasonable doubt. However, the ordinary rules of evidence confer on a judge a discretion to exclude evidence that is unduly prejudicial, albeit the evidence is otherwise admissible. The exercise of that discretion is designed to avoid a significant risk that the evidence will be misused by the jury in a way that cannot be guarded against by an appropriate warning."

If I may say so, with respect, this appears to be a firm starting point from which an examination of the issue of whether or not the evidence of the three police officers in the present case by which they identified the accused, largely as a result of voice recognition, can be commenced.

85 Also in Bulejcik v The Queen (supra) Toohey and Gaudron JJ, at 397 - 398, drew attention to the hazards associated with all forms of identification evidence, whether visual identification, a comparison of handwriting or voice identification. Their Honours treated the well-known passage of Mason CJ, Deane, Dawson, Toohey, Gaudron and McHugh JJ in Domican v The Queen (1992) 173 CLR 555 at 561 - 562 about the need for a jury to be warned of the dangers of convicting upon disputed evidence of identification as applying as well to voice identification evidence. The need for a warning of the dangers involved in acting upon such evidence does not, however, mean that particular types of voice identification evidence are or should be inadmissible. A full account of the dangers of using voice identification and recognition evidence can be found in two articles by David Ormerod: "Sounds Familiar? - Voice Identification Evidence" [2001] Crim LR 595 and "Sounding Out Expert Voice Identification" [2002] Crim LR 771 where the author observes that everyone can recount situations in which he has mis-identified a voice, even when the person he thought he was hearing

(Page 36)

was well-known to him. This author points out that the hazards of voice identification can be traced to biblical writings where Isaac mis-identified Jacob's voice: Genesis Ch 27 v 1-22. While this long experience may emphasise the need for caution in handling such evidence it does not, of itself, mean that it is inadmissible.

86 In this case it is necessary to decide whether any special requirements, such as substantial familiarity with the voice in question, or a distinctive nature of the voice, should be regarded as essential prerequisites for the admission of voice identification evidence in this State. Next it will be necessary to consider whether or not the prior experience of each of the three police officers with the voice of the appellant permitted them to give evidence identifying her as the person shown on the videotape firing the fatal shots. This involves determining whether or not each of those witnesses was in a position, through his or her experience, to identify or to attempt to identify the voice of the accused and, further, whether in giving evidence about the significance of the videotape upon which the voice of the killer was recorded, those witnesses were doing any more than merely interpreting the significance of that evidence in a manner which the jury alone could and should have done themselves. Two separate steps are involved in this process of evaluation, the second of which requires an examination of whether or not the principles established by the High Court in Smith v The Queen (2001) 206 CLR 650 apply. It is convenient to deal first with the issue of whether or not the witness must have some special familiarity with the voice of the person to be identified or whether that voice must be very distinctive.

87 In R v E J Smith (supra) O'Brien CJ of Cr Div regarded evidence of recognition of a voice as one species of identification evidence and, therefore, subject to the general rules relating to identification evidence established by the High Court in Alexander v The Queen (1981) 145 CLR 395. After examining the Australian authorities and leading United States legal writings and cases, his Honour went on to say at 475:

"The general rule has been held to be that testimony by a witness that he recognised the accused by his voice is admissible in evidence, provided that the witness has some basis for comparison of the accused's voice with the voice which he identifies as the accused's and this is satisfied if the witness acquires his knowledge of the accused's voice after the event to which the witness testifies as well as before that time."

(Page 37)

And then, after considering further United States authority and the Devlin Report on Evidence of Identification in Criminal Cases (1976) his Honour said:

"It is to be noted therefore that according to both United States authorities and to the observations in the Devlin Report, evidence of the voice of a person present at a crime as being the same as the voice of the accused can only amount to positive identification where the witness is very familiar with the voice before hearing it at the crime, or that the voice heard at the crime was very distinctive, which means that the witness need not have heard the voice before the crime but heard it as the voice of the accused for the first time after the crime and then noted it to have the same very distinctive features as had the voice at the crime."

As previously mentioned, this approach has been followed by the Court of Criminal Appeal in New South Wales in R v Brownlowe (1987) NSWLR 461 and in a series of other cases in that State.

88 The vital point appears to be that whether the evidence amounts to positive identification or not is a question for the jury - R v E J Smith (supra) 444 at 458. From this it seems to follow that where there is evidence of visual identification or visual recognition of the accused, the method by which that visual identification or recognition is made, whether at an identification parade, or from a photo board or as a result of extensive prior familiarity these are all matters which can be examined in determining the probative value of the evidence. This can be done in the course of submissions by counsel and by a direction from the presiding Judge to a jury about the factors which may go to its reliability. Where necessary, the usual warnings must be given in relation to such identification evidence. As the proposed evidence relating to identification of the accused by voice will go directly to identity, it can almost always be expected to have some probative value and, therefore, be prima facie admissible. If, however, the prejudicial effect of the evidence, for any of a number of possible reasons, is so great as to outweigh a probative value which may be low or questionable, then there will be the discretion of the trial Judge to exclude that evidence, notwithstanding any probative value. That discretion will need to be exercised on established principles.

89 I can appreciate that the rules which have evolved in New South Wales by which tests of familiarity or distinctiveness have been developed in order to determine whether or not the purported voice

(Page 38)

identification is admissible at all are directed to the sole question of admissibility. But, with respect, that does not appear to me to be consistent with the general rule that, subject to the discretion to exclude because of prejudicial effect, all relevant evidence is admissible and the weight which can or should be attributed to that evidence will be a matter for the tribunal of fact at the particular trial.

90 Accordingly, if the witness has any familiarity with the voice which is to be identified I consider that his or her evidence is therefore admissible and that this Court should prefer the line of authority which has developed in Victoria: R v Callaghan [2001] 4 VR 79, in Queensland: R v Corke (1989) 41 A Crim R 292, in the Australian Capital Territory: R v Miladinovic (1992) 109 ACTR 11 at 15 - affirmed Miladinovic v The Queen (1993) 47 FCR 190 and in South Australia: R v Bueti (1997) 70 SASR 370 per Doyle CJ at 380 in preference to the approach taken by the courts in New South Wales. This also appears to me to be implicit in the approach taken by this Court in Nguyen v The Queen (2002) 26 WAR 59 by Anderson J with whom Steytler J agreed.

91 In this sphere it is possible that the voice identification or recognition evidence may be given or supplemented by the provision of expert opinion: Nguyen v The Queen (supra) and R v McHardie [1983] 2 NSWLR 733 at 763, even where that expert opinion may rely on scientific analysis by a sonagraph or other technical means. But this does not mean that all voice identification or recognition evidence is a matter of opinion upon which expertise must be established as a condition for its admissibility. Rather, the process of voice recognition or identification is the application of ordinary human experience to information received via the sense of sound and is but another example of normal human perception. As already observed, the reception of the evidence appears to depend upon no more than the witness's familiarity with the voice to be identified or, as was said by Toohey and Gaudron JJ in Bulejcik v The Queen (supra) at 396:

"However, the issue here is not whether the appellant's voice was so distinctive as to be embedded in the listener's memory or whether the jury had prior familiarity. Rather, the question is whether there was enough material of sufficient quality to make the comparison and whether the jury were properly warned of the difficulties involved."

92 In the present case each of the three police witnesses who gave evidence, identifying the killer shown on the videotape of the shooting,

(Page 39)

from the voice on that videotape, as the accused, had familiarity with the voice of the accused. Two of the police officers had become familiar with her voice through face-to-face dealings before the crime had been committed. The third had spent many hours listening to recordings of telephone interceptions which included her voice. As a consequence I am satisfied that each of those officers had sufficient experience of the accused's voice to give evidence about their recognition of the accused as the person whose voice was recorded on the video surveillance tape of the shooting.

93 Just how reliable that evidence was and whether or not the individual officer's experience of the accused's voice was sufficient to make his or her evidence, taken with all the other evidence of identification, enough to satisfy the jury beyond reasonable doubt on this issue of identity was the ultimate question for the jury. In determining that ultimate question the jury was entitled, indeed obliged, to listen to the voice recordings of the accused which were in evidence and to compare them with the evidence on the video tape on the fatal occasion in order to reach a verdict. Indeed, on this approach where evaluation of the strength of the evidence is, as it always should be, for the jury to decide the importance of the jury making its own assessment of all this evidence is all the more obvious. This adds even greater significance to the second ground of appeal which I address later.

94 So far I have proceeded on the basis that the identification by the three police officers of the accused as the person shown on the videotape shooting the deceased was based, in all important respects, upon voice recognition alone. However, as previously set out, two of the police officers indicated that their recognition was, to an extent, also based upon the recognition of the figure shown in the videotape taken in combination with the voice as recorded. Although, it must be said, that the passages in cross-examination of these witnesses previously described tend to suggest that the degree to which visual identification formed part of the recognition by each of the witnesses was, at the most, small. There are cases in New South Wales and Queensland where the requirement for the witness to be very familiar with the voice which is being identified, or for that voice to be very distinctive, need not be satisfied in cases where the voice identification is accompanied by visual identification. These are R v Corke (supra) and R v Brotherton (supra) at 106.

95 In R v Corke (supra) the Court of Criminal Appeal in Queensland (Macrossan CJ, Derrington and McPherson JJ) quashed a conviction and ordered a new trial where the complainant's visual identification of the

(Page 40)

accused occurred when she confronted him alone at a police station rather than in a line up. In addition, the trial Judge had admitted evidence of the complainant's identification of the appellant's voice while at the police station. The court concluded that such a suggestive form of visual identification meant that the conviction was unsafe and should be quashed. In relation to the evidence of voice recognition and the submission that this evidence should also have been excluded because of the principle in New South Wales, for which Smith [1984] 1 NSWLR 462 was cited, that recognition of the voice of the accused can only amount to positive identification in cases where the witness is familiar with the voice before hearing it at the crime or where the voice heard at the crime was very distinctive, their Honours said:

"This citation is good only in the context of that judgment where the discussion is expressly limited to the situation where the identification relates to the voice alone. The necessity for pre-knowledge of the vocal features of the offender or alternatively for the existence of some distinctive characteristic does not apply to the case where the identification of voice is associated with visual identification providing that the former is made in circumstances equivalent to a line up. Even then the same warnings must be given to the jury as in the case of visual identification, and it may be further desirable to warn the jury that in the absence of some distinctive quality, identification of voice is usually less reliable than visual identification and that its support of quality should be judged accordingly."

In R v Brotherton (supra) the Court of Criminal Appeal of New South Wales (Hunt CJ at CL, Wood and Sharpe JJ) was dealing with a case of voice identification where the witness was familiar with the voice before the crime and the question of the application of the rule in R v E J Smith (supra) came under consideration. At 106, Hunt CJ at CL, with whom Wood and Sharpe JJ agreed, said of voice recognition evidence based on a conversation between the witness and the accused some two days before the offence and lasting only for 10 minutes:

"If that had been all, I would have considerable doubt whether the complainant's familiarity with the appellant's voice would have been sufficient for the voice identification by way of recognition to go to the jury. But it was not all. Significantly, it was accompanied by visual identification, in particular the tattoo on the man's thigh when she was forced to fellate him, and which she described in a way which was not markedly

(Page 41)

different from the tattoo which the appellant did in fact have on his thigh. She also heard the appellant speak again the next morning when she confronted him and alleged that he had been her attacker. The voice identification does not have to stand by itself in assessing the strength of the complainant's identification of the appellant as the man who raped her, where one strand of the identification evidence is of poor quality, another strand may support its correctness, and what must be considered is the accumulation of the evidence ... "

On one view R v Corke (supra) and R v Brotherton (supra) may support an approach that the requirement for the witness to be very familiar with the voice to be identified, or the voice to be very distinctive, before voice identification evidence is admissible, does not apply where there is other evidence of visual identification. In this case the limited reliance by the police officers on the appearance of the killer on the videotape as an aid to recognition in combination with the voice identification made might be seen to come within that rule. However, I do not consider that the New South Wales rule should be applied in this State at all or that, in this case, the admissibility of the evidence from the police officers should be upheld simply because of some degree of reliance upon visual recognition as well. Rather, I agree, with respect, with the view that the treatment of voice recognition evidence should be no different to that of other forms of human identification or recognition which led Doyle CJ in R v Bueti (supra) at 379 to say:

"That being so, there is no reason in principle, in my opinion, where the admissibility of evidence of identification by voice recognition should be treated as subject to rules peculiar to that type of evidence. The admissibility of such evidence should depend upon the principles that generally regulate the admissibility of evidence. That is not to deny, of course, that those principles must be applied with reference to the particular nature of the evidence under consideration."

96 This leads to the second aspect of the argument for the appellant that this voice recognition evidence by the police officers should have been excluded. This is the submission that the principle in Smith v The Queen (supra) rendered the police voice identification evidence inadmissible. Smith v The Queen (supra) is not a case dealing with voice identification evidence but, rather, is a case involving the means of identification of an accused as the offender depicted in security camera photographs which recorded a robbery in a bank. Two police officers who had previously

(Page 42)

dealt with the accused gave evidence at the trial that one of the robbers shown on this film was the accused. The court quashed the conviction, allowing an appeal from the Court of Criminal Appeal of New South Wales, and ordered a new trial. All members of the court joined in those orders. Gleeson CJ, Gaudron, Gummow and Hayne JJ did so on the ground that the evidence of identification by the police officers was irrelevant. Kirby J considered that the evidence, while relevant, was opinion evidence and that for that reason it was inadmissible because of s 76 of the Evidence Act 1995 (NSW). The reason why four of the Justices regarded the evidence as irrelevant was because the process of identification involved nothing more than a normal evaluation of the images shown on the bank camera recording of the robbery and that this was a task well within the capacity of any observer and of the members of the jury themselves. These four Justices considered that evidence from police officers of recognition of the accused as one of the members was nothing more than an interpretation of the film of the robbery which the members of the jury could and should undertake themselves. Their Honours said (at 655):

"Because the witness's assertion of identity was founded on material no different from the material available to the jury from its own observation, the witness's assertion that he recognised the appellant is not evidence that could rationally affect the assessment by the jury of the question we have identified. The fact that someone else has reached a conclusion about the identity of the accused and the person in the picture does not provide any logical basis for affecting the jury's assessment of the probability of the existence of that fact when the conclusion is based only on material that is not different in any substantial way from what is available to the jury. The process of reasoning from one fact (the depiction of a man in the security photographs) taken with another fact (the observed appearance of the accused) to the conclusion (that one is the depiction of the other) is neither assisted, nor hindered, by knowing that some other person has, or has not, arrived at that conclusion. Indeed, if the assessment of probability is affected by that knowledge, it is not by any process of reasoning, but by the decision-maker permitting substitution of the view of another, for the decision-maker's own conclusion."

97 In this present case two of the police officers had direct experience of speaking with the accused on occasions before the crime and therefore had a familiarity with her voice which derived from material which was

(Page 43)

not, and which could not be, before the jury. So far as the evidence of Detective Sergeant Noye was concerned, not only had he played over and over the recordings of the intercepted phone calls during the two weeks which he spent transcribing them, but he too had also spoken to the appellant for some 40 minutes in November 1999 before the commission of the crime. He therefore had a greater familiarity with the accused's voice than could be obtained by the jury in listening to the telephone intercepts alone having regard to the time available to the jury to do so. In other words the familiarity with the voice of the accused possessed by each of the three witnesses went beyond the materials which were in evidence. It was based upon a more extensive experience than those materials.

98 In Smith v The Queen (supra), at 656, it is recognised that there may be a variety of circumstances or occasions when a witness may give evidence that he or she recognises a person depicted in a photograph. Most of the examples given involve a situation where the witness's assertion of identity relies on material or features which go beyond the image shown on the particular photograph. Turning to the somewhat different realm of recognition of voices on videotapes, that is also the situation here because, as described, the evidence of recognition of the three police officers of the voice on the video recording of the killing was based on experience and familiarity which extended beyond the materials which were before the jury and upon which the jury could make comparisons.

99 For these reasons I consider that the first ground of appeal must fail.

Second ground of appeal – directions to the jury about the treatment of the Police voice identification evidence

100 The second ground of appeal concerns the directions which the learned trial Judge gave to the members of the jury that they should avoid using their own observations about similarities in the voices of the appellant, as recorded in the evidence, and the voice of the killer recorded on the videotape at the time of the shooting and that they should avoid playing detectives themselves. Unfortunately, this resulted in a misdirection by suggesting that the members of the jury should not themselves evaluate the evidence to make this vital decision but, rather, ought decide whether or not they could accept the police evidence of the identification of the accused made largely from the voice recognition process already described.

(Page 44

101 The need for the members of the jury themselves to make this decision about whether or not all the evidence proved to their satisfaction that the accused was the killer as an essential step in their task of reaching a verdict appears plainly from Smith v The Queen (supra) and Nguyen v The Queen (supra). Indeed, it emerges directly from the application of fundamental principles because, in this case, the determining issue was that of identification and this was the ultimate question which had to be proved by the prosecution to the satisfaction of the jury beyond reasonable doubt.

102 In this respect I agree with the reasons for decision of Miller J and, in particular, with his Honour's observations that:

"The cases stand for the proposition that a jury must be informed that although there is evidence to assist them on the issue, it remains ultimately their decision and a decision which they can take, having regard to their own views on the matter from the material available in the court, irrespective of the opinion or identification evidence which may have been adduced by the prosecution".

103 Earlier in these reasons I explained why I considered an extension of time within in which to appeal should be granted. I would allow the appeal and would quash the conviction of the appellant and order a new trial.