HKSAR v. YEUNG KA HO AND ANOTHER

Press Summary (English)

Press Summary (Chinese)

FACC No. 11 of 2012

IN THE COURT OF FINAL APPEAL OF THE

HONG KONG SPECIAL ADMINISTRATIVE REGION

FINAL APPEALNO. 11 OF 2012 (CRIMINAL)

(ON APPEAL FROM CACC NO. 207 OF 2011)

_______________________

_______________________

_______________________

J U D G M E N T

_______________________

Mr Justice Chan PJ :

1.  I agree with the judgment of Sir Thomas Gault NPJ.  For the reasons given by him, I too would dismiss the appeal.

Mr Justice Ribeiro PJ :

2.  I agree with the judgment of Sir Thomas Gault NPJ.

Mr Justice Tang PJ :

3.  I have had the advantage of reading in draft the judgment of Sir Thomas Gault NPJ.  For the reasons given by him I agree that the appeals should be dismissed.  In deference to counsel’s submissions, I will add a few words.

4.  We are concerned with voice identification of an audio recording, by Mr Lau (PW1), who claimed to be familiar with the voice of the appellants.  In connection with such evidence, the learned judge reminded himself that he shouldapply:

“13. … Turnbull warning with relevant and necessary modifications.”

It is clear from the reasons for verdict that throughout the learned judge was mindful of those warnings. 

5.  Mr Chan submitted that PW1’s voice identification evidence was the result of or influenced by the content of the recorded conversation.  If so, I agree, that might undermine the reliability of the voice identification.  However, that was not the evidence of PW1.  Nor did the reasons for verdict give any hint that the learned judge took that view. 

6.  Quite the contrary, the learned judge said:

“102. Mr. Lau heard the voices of the two Defendants for many times within a long period of time. Evidence of Mr. Lau on the identification of their voices is absolutely credible and reliable.”

7.  Nor did it appear that this finding was influenced by the circumstantial evidence in the case[1]. It was after this clear statement (and, an earlier statement to similar effect in para 100), that the learned trial judge went on to say:

“103. Besides, there was also other peripheral circumstantial evidence to support the evidence of voice identification of Mr. Lau.”

8.  The learned judge then took note of the circumstantial evidence which strongly supported Mr Lau’s evidence that the appellants were the speakers.  

9.  I believe in deciding what weight to put on the voice recognition evidence of Mr Lau, the learned judge was entitled, indeed, bound to have regard to all relevant and probative evidence in the case.  This is not a case where the judge was uncertain about the reliability of the voice identification evidence.  There was no undermining of the importance of maintaining the integrity of the modified Turnbullwarnings.

Mr Justice Mortimer NPJ :

10.  Having had the advantage of reading Sir Thomas Gault NPJ’s judgment in draft, I agree for the reasons he gives, that the appeals must be dismissed.

Sir Thomas Gault NPJ :

11.  This is an appeal pursuant to leave granted by this Court on 12 November 2012.

12.  The issues for consideration are the same in respect of both appellants, and involve evidence of identification said to reside, in part, in a sound tape recording.

13.  The appellants were tried in the District Court before HH Judge D Yau on 21 April 2011.  Both were convicted on two charges; the first of doing acts tending and intended to pervert the course of public justice, and the second of misconduct in public office.

14.  The appellants were police officers.  In a criminal trial in 2007 in the District Court (“the earlier proceeding”), the 1^st appellant was a prosecution witness having been deployed as an undercover officer investigating allegations of conspiracy to manage a vice establishment and money laundering.  The 2^nd appellant was an exhibits officer in the same trial.  It is claimed that while the 1^st appellant was giving evidence, but during the lunch break, the appellants had a discussion in a witness room that was recorded and disclosed the offending for which they were subsequently charged and convicted.

15.  When the appellants were tried in the District Court, the judge’s reasons for verdict stated at the outset:

“3. The Prosecution and Defence did not dispute on the admissibility of the relevant voice recording pen and the contents therein. In paragraph 4 of the Agreed Facts, it was agreed between all parties that the relevant voice recording pens had been produced by counsel for the 4^th and 6^th Defendants as exhibits in DCCC 1280/2005. It was also agreed between all parties that those voice recording pens were not handled improperly after they had been produced to the court.”

16.  In this Court, Mr Charles J Chan represented both appellants.  Mr Gerard McCoy SC and with him, Mr Jonathan Man Tak Ho (“Mr Man”), represented the respondent.

17.  Mr Chan contended that the admissions of fact did not preclude an argument that the tape recording is inadmissible as a matter of law.  Further, he said the admission, in any event, went no further than accepting the provenance of the taped evidence after the tape was produced to the court in the earlier proceeding.  However, the tape, having been received in evidence, the case proceeded essentially as requiring determination of whether the evidence before the court proved that the participants in the recorded discussion were identified as the appellants and that they engaged in the charged criminal conduct.

The evidence

18.  There is no direct evidence in the present case establishing who made the recording, when it was made or the circumstances in which it was made and kept prior to its production to the court in the earlier proceeding.  The prosecution contends that sufficient facts can be inferred from the content of the recordings when taken with other available evidence.

19.  The tape records a conversation between two persons from which, the prosecution says, relevant evidence can be inferred.  There was no challenge to the finding that whoever it was talking on the recording, there was a discussion concerning the giving of false evidence.

20.  In his reasons for verdict, after referring to the admissions and recording the charges, the District Court Judge said that the only real issue in the case was the voice identification evidence of the prosecution witnesses in relation to the interlocutors in the audio recordings.  At the outset of his consideration of the evidence, the judge said:

“13. When considering the evidence on the voice identification, the court applied Turnbull[2] warning with relevant and necessary modifications, and reminded itself that even an honest witness might make a mistake in voice identification.”

21.  Referring to the prosecution evidence, the judge described the witness PW1 (“Mr Lau”) as the most important prosecution witness and the only person who could positively identify the voices of the two appellants in the recorded conversations.  He said that Mr Lau is a prosecuting counsel and had been counsel representing two of the defendants in DCCC 1280/2005.  It was in the course of the trial in that case that the present appellants are said to have offended.

22.  Mr Lau’s evidence in the present case was that towards the end of the prosecution case in DCCC 1280/2005, counsel for other parties produced to the court 45 voice recording pens.  One contained the recording with which we are concerned.  Mr Lau testified that he could identify the voices on the tape as those of the present appellants.  The reliability of that evidence is in contention.

23.  Mr Lau’s evidence was that, in the course of the earlier proceeding, in the morning of 1 June 2007, he had complained to the judge in the court that one of the police officers in the prosecution team whom he subsequently identified during an identity parade as the 2^nd appellant, had peeped at a document that Mr Lau was reading.  He knew that police officer from a previous case in the Kowloon City Magistracy.  Mr Lau said that when he listened to the tape, he recognized the voice of the same police officer he knew as “Billy”.

24.  The District Court Judge in his reasons for verdict said:

“28. The court must handle this part of the evidence of Mr. Lau very carefully. The only useful part of such evidence and the only purpose that it can be admitted as evidence is the identification of voices of the people who spoke in (the audio recordings contained in) the voice recording pen.”

25.  The judge summarized the voice identification evidence as follows:

“31. First of all, Mr. Lau testified that the 1^st Defendant had testified in court in the capacity as prosecution witness in a trial case before Magistrate Mr. Peter Law at Kowloon City Magistracy. Such a case at Kowloon City Magistracy was one of the cases of keeping vice establishments deriving from DCCC 1280/2005. In such a case at Kowloon City Magistracy, the 1^st Defendant as an undercover police officer pretended to be a customer and visited the vice establishment concerned.

32. Moreover, in the case (DCCC) 1280/2005, the 1^st Defendant testified at the witness box for a long period of time. Mr. Lau was responsible for cross-examining the 1^st Defendant so he had sufficient time and chances to familiarize himself with the voice of the 1^st Defendant. Although the 1^st Defendant had not yet started testifying in court during the first suspended trial, the 1^st Defendant had assisted in handling documents. According to Mr. Lau, the 1^st Defendant was relatively familiar with documents, and sometimes after Mr. Lau had asked the 2^nd Defendant i.e. the exhibits officer something about documents, the 2^nd Defendant would seek assistance from the 1^st Defendant.

33. Mr. Lau testified that even before commencement of the re-trial of (DCCC) 1280/2005, since Mr. Lau had had contacts with the 1^st Defendant in the trial case at Kowloon City Magistracy, he was already able to identify the voice of the 1^st Defendant. The trial case at Kowloon City Magistracy lasted for about 12 to 15 days. The 1^st Defendant testified for about 1 to 2 days.

34. All in all, in the presence of Mr. Lau, the 1^st Defendant testified for about 30 days or 120 hours. The last time Mr. Lau heard the voice of the 1^st Defendant was when the 1^st Defendant testified in the voir dire of DCCC 1280/2005 i.e. in about late 2007 or early 2008.

Regarding the 2^nd Defendant

35. Regarding voice identification of the 2^nd Defendant, first of all, Mr. Lau testified that the 2^nd Defendant never acted as a witness in DCCC 1280/2005 or in other cases handled by Mr. Lau.

36. Mr. Lau said that he talked with the 2^nd Defendant for many times because the 2^nd Defendant had also assisted in the trial of the case at Kowloon City Magistracy. Besides, the 2^nd Defendant also assisted in the process of the trial on many occasions during the suspended trial of DCCC 1280/2005 and the subsequent 130-day retrial.

37.   Mr. Lau said that the 2^nd Defendant was responsible for passing documents to the defence.  Through these contacts and occasional chit-chats with the 2^nd Defendant, Mr. Lau became familiar with the voice of the 2^nd Defendant.  Mr. Lau said that the 2^nd Defendant was rather talkative so during court adjournments Mr. Lau talked with the 2^nd Defendant for many times.  Although contents of those conversations were not something important and also not personal, such kind of chit-chats occurred for about 50 to 60 times, each lasting for about 1 to 2 minutes.”

26.  In cross-examination, Mr Lau said that when asked by the police about the matter in April 2010, some 3 years and 2 months later, his initial recall was vague though he remembered having complained to the judge about the person he remembered having peeped at documents he was reading.  His evidence was that, after having located and reviewed notes of the earlier proceeding he had retained, he recalled that the person who peeped at his papers was “Billy”.

27.  Mr Lau did concede that he had not seen the two appellants alone together in the witness room during court adjournments. 

28.  The judge heard some expert testimony to the effect that there was no indication that the tape recording had been edited or interfered with and that there was no technical indication of when it had been recorded.

29.  PW4 (“Mr Ng”) was the prosecuting counsel in the initial suspended trial and re-trial in DCCC 1280/2005.  He confirmed that the 1^st appellant was the second prosecution witness.  He said that one of the police officers who assisted him was called “Billy”.  PW4 had also prosecuted in the trial at Kowloon City Magistracy where the same undercover police witnesses gave evidence.

30.  In the reasons for verdict, the District Court Judge noted PW4’s evidence was that he had listened to the relevant audio recordings and was unable to identify the voices recorded.  The judge then stated:

“57. … After hearing the evidence of Mr. Ng on this aspect, I deliberately remind myself that I must exercise extra care when dealing with the evidence of PW1 Mr. Lau in relation to voice identification. The reason is that Mr. Ng was the prosecutor in that case. He should have more chances and in closer proximity than any other people to have contacts with the Defendants. Theoretically he should be in a better position to identify the voices of the Defendants.”

31.  The judge went on to note that PW4 could not recall who was the subject of the complaint on 1 June.

32.  In his summary of the evidence of PW4, the judge set out the following findings:

“59. …. Mr. Ng testified that as exhibits officer, the 2^nd Defendant had been assisting Mr. Ng to pass exhibits to the judge for most of the time. Mr. Ng said that he was sure he had more opportunity than PW1 Mr. Lau to talk to the 2^nd Defendant.

60. Another important point in the evidence of Mr. Ng is that when the Defence produced a series of voice recording pens to the court in that case, Mr. Ng and PW1 Mr. Lau were assigned to take part in the preparation of the transcript. According to the evidence of Mr. Ng, during the course of preparing the transcript, Mr. Lau made mistakes for many times and mixed up which voice in the audio recordings belonged to whom. Mr. Ng further said that Mr. Lau had to guess the identity of the persons in the recordings and that Mr. Lau himself had made mistakes for many times and even had to be corrected by his own defence team.

61. However, the Defence did not cross-examine Mr. Ng in relation to the voice identification of the relevant sessions of audio recording in the present case to clarify whether Mr. Lau had made any mistakes concerning such sessions. On the contrary, Mr. Lau gave specific evidence in relation to the relevant session(s) of audio recording in the present case and confirmed accuracy of his evidence on voice identification.

62. Upon invitation by the court and being agreed by the Prosecution and Defence, Mr. Ng listened to the relevant parts of the audio recordings in court again.  What Mr. Ng could do was only to speculate that the voice which the Prosecution alleged to belong to the 2^nd Defendant could be the voice of the 2^nd Defendant.  Mr. Ng was completely unable to give evidence in relation to whether the voice of the 1^st Defendant was that of the 1^st Defendant.”

33.  The 5^th prosecution witness was the officer-in-charge of case DCCC 1280/2005 over the relevant period.  She gave evidence directed to aspects of the proceedings and the roles of participants.  In relation to our present task, her evidence does not take matters much further.  One point from her evidence noted by the judge in his reasons is:

“72. Another important part of the evidence of WDIP Lee is that according to Madam Lee, in the case DCCC 1280/2005 exhibits were all kept in a large box and such large box was in the custody of her subordinates. During lunch breaks or court adjournments, they would normally take the box with them when they left the courtroom. Generally, they would put it in a witness room near the courtroom. Madam Lee said in general the 2^nd Defendant was responsible for watching over the said exhibits box.”

34.  PW5 was unable to identify the voices on the tapes.

35.  Four further police officers gave evidence of having listened to the relevant tape in September 2009 and had been unable to identify the voices recorded.  The judge did not find their evidence added significantly.

The District Court Judge’s Findings

36.  Having set out the evidence, including that of a character witness for the defence (the appellants elected not to give evidence), the judge found all of the witnesses honest and reliable.  He found the contents of the transcript of the audio recordings prepared by PW3 are accurate.  His findings on the voice identification evidence of PW1 were:

“100. I find that evidence of PW1 Mr. Lau on voice identification of the audio recordings is true and accurate. I give full weight to his evidence. I find that regarding the 1^st Defendant, Mr. Lau had ample opportunities to fully familiarize himself with the voice of the 1^st Defendant at the trial at Kowloon City Magistracy and at the trial of DCCC 1280/2005.

101. Regarding the 2^nd Defendant, although Mr. Lau heard the voice of the 2^nd Defendant on relatively fewer occasions, he still had sufficient opportunities to hear the voice of the 2^nd Defendant in the process of handling documents throughout the whole trial of DCCC 1280/2005. As the officer in charge of exhibits in the case, the 2^nd Defendant definitely had many opportunities to talk with Mr. Lau. Furthermore, although the trial period was relatively shorter, the 2^nd Defendant was also the exhibits officer in the trial at Kowloon City Magistracy and Mr. Lau was also the defence counsel in that case. This also increased the opportunity of him hearing the voice of the 2^nd Defendant.

102. Mr. Lau heard the voices of the two Defendants for many times within a long period of time.  Evidence of Mr. Lau on the identification of their voices is absolutely credible and reliable.”

37.  The judge then added his findings on the peripheral circumstantial evidence.  He accepted that references to names and nicknames on the tape were to police officers and police activities connected with case DCCC 1280/2005 and another case with which the appellants and those referred to on the tape were closely familiar. The District Court Judge found also that the taped conversation revealed that the two males were recorded as talking about the incident in the courtroom on 1 June 2007 when complaint was made about “Billy” peeping at PW1’s documents.

38.  The judge also mentioned reference on the tape to another previous case:

“110.  In counters 205 and 211 of the audio recordings, Male 2 asked Male 1 if he could remember another previous case involving “city”.  According to the evidence in our present case, both the 1^st and 2^nd Defendants had been involved in the trial of a case concerning vice establishments which was heard at Kowloon City Magistracy, and the trial of such case concerning vice establishments also involved the undercover operation that was involved in the trial of DCCC 1280/2005.  In both cases, the 1^st Defendant was one of the members in the undercover operation, while the 2^nd Defendant was an exhibits officer.  Notebook of the 1^st Defendant confirmed that he had testified in a case called city sauna (literal translation).”

39.  With reference to the time when and the place where the recorded conversation occurred, the District Court Judge’s reasons for verdict state:

“116. Evidence which is able to assist us in trying to find out on what date and at what time the audio recordings were made are from the contents of the conversations. In the recorded conversations, Male 1 and Male 2 mentioned a floppy disc. Transcript of the trial of DCCC 1280/2005 shows that the 1^st Defendant testified before lunch on 1^st June that he would bring along with him a floppy disc when he performed the undercover operation. Obviously the two males were discussing the evidence which had been given by one of them. I find that the only irresistible inference is that the two males were the 1^st and 2^nd Defendants, and that they were discussing the evidence which had been given by the 1^st Defendant at the trial of DCCC 1280/2005 at an earlier time on 1 July 2007.

Place of making audio recordings

117. Without direct evidence of the person who placed the voice recording pen, the Prosecution could only rely on circumstantial evidence to try to convince the court to make an inference of the place of occurrence of the conversations recorded in the audio recordings. The evidence referred to by the Prosecution included the following.

118. The trial of DCCC 1280/2005 was heard by Mr. Jenkins. Mr. Jenkins’ court was at court no. 24 on 8/F of Wanchai Law Courts, Wanchai Tower at that time. According to the evidence of PW2 the expert witness, such conversation took place in a room. It was heard from the audio recording that the 1^st Defendant told someone else over the phone that he needed what kind of food. The 2^nd Defendant brought back some food and lemon tea later. They talked with each other while eating the food. According to the evidence of PW5 WDIP Lee, the 2^nd Defendant was the officer responsible for watching over the exhibits box in the case during court adjournments. She also testified that in general the exhibits box would be put in the witness room during court adjournments. In counter 232 of the transcript of the audio recordings, someone mentioned quarter past. This seems to indicate that the then time was 2:15pm which was also the finishing time for the lunch, the time for them to prepare to return to the courtroom and continue with the trial. Since they had only a very short time left at that time, they were definitely at a place very near the courtroom.

119. I find that based on the above evidence, I am able to make an irresistible inference that the conversation in the audio recordings took place in the witness room on 8/F of Wanchai Law Courts, Wanchai Tower during lunch time on 1 June 2007.”

40.  The judge took into consideration that others were not able to identify the voices on the tape and set out why he nevertheless accepted the evidence of PW1.  When doing so, he mentioned again the “Turnbull” warning.  He stated his conclusion as:

“123. I find that the 1^st and 2^nd Defendants had conversations in the witness room at 8/F of Wanchai Law Courts, Wanchai Tower, while the 1^st Defendant was still giving evidence for the Prosecution in court in the case DCCC 1280/2005 on 1 June 2007. I find that the contents of their conversations are as stated in Exhibit P23 the transcript. I find that their conversations relate to the evidence which had been given by the 1^st Defendant at the trial of DCCC 1280/2005. I find that they also discussed how the 1^st Defendant should further testify when the trial was resumed, that the 1^st Defendant should not mix up the true and false parts of his evidence, and that the 1^st Defendant should memorise the false parts so as to avoid confusion in future.”

The Case in the Court of Appeal

41.  In their appeal to the Court of Appeal, the appellants advanced four grounds all of which were rejected.  Of those the first; that the contents of the audio recording did not disclose the charged offending, and the fourth, that the District Court Judge had a conflict of interest, were rejected.  They were not pursued in this Court and need not be mentioned further.

42.  The remaining grounds were dealt with together in the Court of Appeal judgment.  The court considered four matters; they were: the making and custody of the tape recording, the possibility that it had been tampered with, whether the voice identification evidence of Mr Lau should have been admitted by the judge, and the weight that should be accorded in Mr Lau’s evidence.

43.  The judgment of the Court of Appeal cited the leading authorities on evidence by audio recordings.  By reference to HKSAR v Lee Chi Fai, the court accepted that the content of the tape was obviously relevant to the charges.  The court then said:

“27. … Secondly, its prima facie authenticity must be considered. In general, the prosecution would have evidence concerning the provenance and history of the audio recording. However, this is not ‘indispensible’. As Roberts CJ said in Choi Kit Kau v The Queen, this is the ‘best’ method of proving. However, the authority case shows that this is not the only way to prove.”

44.  The court accepted that the authorities establish that authenticity can be proved circumstantially.  The judgment reviewed the expert evidence and accepted the finding of the District Court Judge that the expert evidence established the tape as prima facie authentic.  The court then addressed the identification evidence of Mr Lau and concluded that, although he was not an expert, the judge was right to accept his evidence because of the opportunities he had to become familiar with the voices of the appellants he said were recorded on the tape.  The court said:

“37. … The Applicants emphasized that there were other witnesses in the case who were unable to identify the voices depicted in (Exhibit) P37. However, the judge had the advantage of seeing and hearing testimony of every witness. He could observe the response, wisdom, alertness and manner etc. characteristics of every witness, and then he had the right to accept the evidence of one of the witnesses.”

The Case in this Court

45.  In this Court, Mr Chan, for the appellants, presented extensive written argument on the difficulties inherent in dealing with evidence in the form of tape recordings. He argued that only with clear evidence of the making and preservation of any recording and expert evidence of its accuracy and reliability with available voice recorded material for comparison, should such evidence be admitted.  He contended that warnings adapted from the guidance in Turnbull are not sufficient to avoid the risks of relying on evidence of tape recordings.

46.  When he dealt more specifically with the circumstances of the present case, Mr Chan was critical of the inadequacies in the identification evidence.  In that regard, he went considerably beyond the evidence and cross-examination of PW1 before the trial judge.  That cross-examination focused almost entirely on opportunities PW1 had to become familiar with the voices of the appellants. Counsel’s criticisms of the inadequacy of the evidence explaining how PW1 was able to recognize the voices when others could not, seems to have rested on the one answer that Mr Lau acknowledged having never received any special training on voice identification.

47.  The real issue, however, is whether it was open to the judge to conclude that the evidence that was given proved that the voices recorded on the tape were those of the two appellants. 

Decision

48.  The issues surrounding evidence in the form of recordings are broadly similar though the means of recording may be different.  Audio tapes, video tapes, films, photographs, even DNA analysis, whatever the technology used, raise issues of authenticity and accuracy that must be addressed with reference to the particular circumstances.  When a recording is relied on for identification, not only the accuracy of the record, but also the reliability of the witness identifying what is recorded may arise.  These matters are facts which must be proved.

49.  Visual identifications are well recognized as requiring care and the courts recognize this as the decision in Turnbull clearly shows.  The same care and equivalent warnings are applicable to other forms of identification evidence, including voice identifications.

50.  In HKSAR v Lai Wai Cheong[3], the Court of Appeal accepted that the Turnbull guidelines are applicable to evidence of identification, whether of appearance or voice.  A tape recording can be admitted and relied upon as evidence of the contents where a proper basis for its reliability is established.  In the early case of R v Maqsud Ali[4], Marshall J, for the Court of Criminal Appeal, said:

“… We can see no difference in principle between a tape recording and a photograph. In saying this we must not be taken as saying that such recordings are admissible whatever the circumstances, but it does appear to this court wrong to deny to the law of evidence advantages to be gained by new techniques and new devices, provided the accuracy of the recording can be proved and the voices recorded properly identified; provided also that the evidence is relevant and otherwise admissible, we are satisfied that a tape recording is admissible in evidence. Such evidence should always be regarded with some caution and assessed in the light of all the circumstances of each case. There can be no question of laying down any exhaustive set of rules by which the admissibility of such evidence should be judged.”

51.  In Choi Kit Kau v The Queen[5], Roberts CJ said:

“[T]he best method of proving that a tape recording produced in court is authentic is to show that it has been, since the time it was recorded, continuously in the custody of persons who assert that it was not tampered with. This is a course which should be followed if the authenticity of a tape is challenged and is desirable even if it is not.”

52.  Where the “best” method is unavailable, the authenticity of a tape may still be capable of proof.  In HKSAR v Lee Chi Fai and Others[6], the Court of Appeal adopted the analysis in R v Murphy and Another[7]:

“ Authenticity, in our view, like most facts may be proved circumstantially. In the case of a video film, the direct way is to call the cameraman who took it and the court will normally expect him to be called. But if he is not available, he need not be called; other evidence will suffice if it is logically probative that the video was authentic. That evidence may be adduced in other ways and from other sources.”

53.  In R v Chen[8], the Victoria Court of Criminal Appeal received evidence of tape recordings where the makers were dead or unavailable.  The court said:

“The test is whether there is sufficient material before the court to allow the tribunal of fact acting reasonably to conclude that the recorded sounds reproduce those originally made by the persons identified by the evidence. In other words, there must be evidence, which the tribunal of fact is entitled to accept, that the recording is of a conversation which occurred and which would be admissible if proved by oral testimony. In our opinion, admissibility does not depend on the party tendering the tapes having removed absolutely any chance that they are inaccurate.”

54.  There is much of assistance in the lengthy judgment of the Court of Appeal (Criminal Division) of Northern Ireland in R v Murphy and Another.  The case involved video-taped evidence forming part of the prosecution case though the makers of the films did not give evidence. 

55.  The Court of Appeal in England in R v Flynn and St John[9] ruled that evidence of recognition of a voice by a lay person was admissible, but on the facts of that case evidence of police officers of voice recognition should have been ruled inadmissible.  The opportunities for familiarization with the voices of the accused were very much less than in this case and there was expert evidence casting doubt on the reliability of the tape recordings.  The court’s conclusions were summarized at p 279 as:

“ There are other considerations which apply to both appellants. We have already noted the two most important factors, namely the limited opportunity for the officers to acquire familiarity with the appellants’ voices and secondly, the poor quality of the covert recording. To these we add the following. First, the police officers’ purported recognition of the appellants’ voices is in marked contrast to the evidence of the two experts who are unable to recognise their voices; in Dr Holmes’ case she was unable to identify individual voices. Secondly, nothing is known of the ability of any of the police officers to recognise voices. There is no evidence that any of them had any training in auditory analysis. Thirdly, the identification of the voices was carried out by listening to the covert tape on a standard laptop computer, as opposed to the sophisticated equipment used by the experts. Fourthly, in our judgment, it is significant that DC Gittings’ transcript contains words attributed to Flynn which Dr Holmes could not distinguish as words, let alone recognise as attributable to Flynn. Fifthly, the expert evidence shows that lay listeners with considerable familiarity of a voice and listening to a clear recording, can still make mistakes.”

56.  To the judgment in that case the court added this postscript:

“ As appears from the above we have been dealing in these appeals with issues arising out of voice recognition evidence. Nothing in this judgment should be taken as casting doubt on the admissibility of evidence given by properly qualified experts in this field. On the material before use we think it neither possible nor desirable to go as far as the Northern Ireland Court of Criminal Appeal in O’Doherty which ruled that auditory analysis evidence given by experts in this field was inadmissible unless supported by expert evidence of acoustic analysis. So far as lay listener evidence is concerned, in our opinion, the key to admissibility is the degree of familiarity of the witness with the suspect’s voice. Even then the dangers of a mis-identification remain; the more so where the recording of the voice to be identified is poor.”

57.  In Jones & Harris[10], the Court of Criminal Appeal of Victoria rejected submissions for two appellants that evidence of voice identifications should not be received.  In the case of the appellant Jones tape-recorded telephone conversations together with circumstantial evidence consistent with factual statements made during the telephone calls was received as probative of identity.  Similarly, in the case of Harris, voice identification evidence from witnesses familiar with his voice and supported by circumstantial evidence was held to have been rightly admitted.

58.  I see no reason why, in certain circumstances, the identity of participants and, the timing and circumstances of a recorded conversation, cannot be proved by reference to the content of the recording.  It must be possible as where the maker is unknown or unavailable and the content when placed in a context of known facts establishes time and place.  Similarly, the identity of a speaker may emerge from what is said.  Statements revealing the facts or knowledge that only an offender could know frequently can lead to identification.  For example, statements made by disguised offenders in the course of a robbery can be relied upon to identify them.  Statements made anonymously in writing or in the course of a telephone call demanding ransom after a kipnapping may support identification.

59.  Of course other possible contingencies must be excluded – such as voice imitation.  Those matters go to the weight to be accorded the evidence of the conversation.

60.  The authorities make clear the care with which tape recordings must be considered.  But with appropriate safeguards, and careful directions, there is no reason to construct rigid exclusionary rules.  As with all evidence, reliability is to be assessed in all the circumstances.

61.  In many cases, it would be question-begging to seek to prove the authenticity of a tape recording from the contents of the recording itself.  Mr Chan, for the appellants, submitted that the voice identification in this case is flawed because PW1 was not shown to have recognized the speakers solely by their voices and without reliance on the context with which the speakers exhibited familiarity.  He argued that:

“In substance, the whole exercise would not stand to test the ability of PW1’s ability in the voice identification. It is apparent that both PW1 and essentially the learned trial Judge had been affected by this potential contextual bias, using the content to reinforce their ‘belief’ that the voices belonged to the Appellants.

…

The Appellants respectfully submit that to assess the evidence of voice identification of PW1, it has to be done in separation of the content of Exh. P37.  Otherwise, one would fall into the same flaw or weakness in suggesting that PW1 was capable of identifying something apparently representing the details of a content that he was familiar with and borrow force from this to his ability of voice recognition because it is a ‘content recognition’ which might subconsciously affect the listener and also it is a logical fallacy amplified infinitely because one cannot examine how much the recognition is attributable to voice and content respectively.”

62.  I do not accept that the “context”, indeed all the surrounding circumstances, may not be taken into consideration in evaluating evidence of identification.  The argument to the contrary is unreal and inconsistent with the authorities already mentioned.

63.  Of greater significance is whether PW1’s evidence of his recognition of the identity of the voices on the tape may have been not solely his own, but was contributed to by others who did not give evidence of recognition or identification of the voices.  PW1’s evidence was that when the existence of the tapes was exposed at the end of 2007, he was one of the people involved in preparing transcripts.  There were 45 tapes produced at the trial of DCCC 1280/2005 though only one contained the conversation with which we are directly concerned.

64.  At the trial of this case, when PW1 gave evidence that he could identify the voices on the tape, he was asked when he first disclosed this.  His evidence was:

“A: Actually, I disclosed (this) at the trial of DCCC 1280 because actually in the process of preparing these transcripts, (the identities of) Male 1 and Male 2 had been mutually agreed by the prosecution and defence in that trial.

Q: The prosecution and defence did not agree on who was Male 1 and who was Male 2; otherwise it would have been written at an early stage --- it would have been specified at the beginning that Male 1 was whose voice and Male 2 was whose voice. Do you agree? This was not agreed by both sides. Is it?

A: I dare not say (whether it is the case or not) for individual transcript. For example, if you are talking about whether the transcript for 1 June had been mutually agreed at that time, we --- I have to go back and ‘check’ my notes, but regarding the transcripts being produced to the court at that time, most of the Male 1, Male 2 or Female 1, or Male 1, Male 2, Male 3, all were mutually agreed by the prosecution and defence before the transcripts were produced to the court; otherwise the cross-examination could not be continued.

Q: I put it to you that, by saying that the prosecution and defence mutually agreed, you mean agreeing the contents of the transcripts but not the identities of the voices appeared in the transcripts. This has never been agreed. Do you agree? Since …

A:  I have to think about this first.  I do not have much recollection about this.”

65.  Later in his evidence, PW1 explained why in October 2010 he was prompted to consider the details of the “peeking” incident of 1 June 2007.  He said that in connection with other litigation it had become necessary for him to review the relevant documents and his contemporaneous notes.  There was also a professionally prepared transcript then available.  He said this brought back memory of the incident.  It was not put to him that this recollection and identification in 2010 and subsequently were aided by having been involved, with others, in 2007 in preparation of the first transcripts (which in any event did not include the identities of the recorded speakers).  Of course this would not exclude the possibility of there having been discussion or speculation at the time, but there is no evidence of that.  At the trial of the present proceeding, the tape was played and PW1 identified the voices as those of the appellants.  Doubtless by then he would have listened to the tapes a number of times.  Standing alone his evidence that he could identify the voices after the lengthy lapse of time would warrant concern.  But the circumstantial evidence gives real credibility to his recollections, as they clearly did for the District Court Judge and the Court of Appeal.  And the peeking incident gave him a particular reason to remember.

66.  I have not been persuaded that the reasons for verdict contain material errors of law.  Nor has the judgment of the Court of Appeal been shown to address incorrectly the arguments presented to that court.  The case has involved an assessment of the reliability of oral evidence of identification bolstered by circumstantial evidence directed to the authenticity of the tape recording.  The courts below were convinced by the evidence tying the recorded conversations to the appellants and to the occasion on which they occurred.  Both lower courts reviewed the absence of evidence from others whose exposure to the voices of the appellants was no less than PW1 but nevertheless accepted his evidence.

67.  The dangers of misidentification of voices especially by lay listeners were acknowledged.  PW1’s familiarity with the voices of the appellants was fully investigated and his evidence was firm.

68.  Accordingly, I am satisfied that the verdicts were open on the evidence and there has been demonstrated no “substantial and grave injustice” (So Yiu Fung v HKSAR[11]) as would be required under s 32(2) of the Court of Final Appeal Ordinance to warrant disturbing the convictions.

69.  I would dismiss the appeals.

Mr Justice Chan PJ :

70.  The Court unanimously dismisses the appeal.