Knutson, Chief Justice.
This is an appeal from an order of the District Court of Ramsey County discharging a writ of habeas corpus.
The facts may be briefly stated. On May 22, 1970, just after midnight, a call was received by the St. Paul Police Headquarters requesting that assistance be sent to a pregnant woman who was about to deliver, who lived at 859 Hague Avenue. Within minutes two officers, Officer Glen Kothe and Officer James Sackett, were dispatched to that address. When they arrived, they went to the front door and knocked but received no response. Officer Kothe then went around to the rear door. While at the rear of the house he heard an explosion, saw a flash, and heard Sackett cry out. Kothe ran to the front of the house and found that Sackett had been shot. He died shortly afterwards. It turned out that he had been shot with a high-caliber rifle.
All emergency calls received by the St. Paul Police Department are recorded. The tape which contained the call that lured Officer Sackett to his death was forwarded to the Michigan State Police Crime Laboratory for examination. There, Detective Sergeant Ernest Nash, who was in charge of the voice identification unit, made voiceprints (spectrograms) of the tape. Subsequently, tape recordings of 13 voices were also sent to Sergeant Nash and he made voiceprints of these. One such recording was of the voice of petitioner herein, Miss Trimble.
After comparing the tape originally sent with that known to be of Miss Trimble’s voice, Sergeant Nash concluded that the voice of the unknown caller and that of Miss Trimble were one and the same.
The recording of Miss Trimble’s voice was made without her knowledge while she was at the Ramsey County Welfare Department office, where she had been requested to come to check up on “eligibility factors” of AFDC payments which she was receiving. Prior to this time, arrangements had been made with the St. Paul Police Department to have Miss Trimble converse with a member of the department, Officer Carolen Bailey, while at the welfare office in the belief that she was discussing her welfare payments. The sole purpose of the call from Officer Bailey was to record Miss Trimble’s conversation and have voiceprints made therefrom.
Prior to recording Miss Trimble’s voice, approval for calling and surreptitiously recording her voice and that of another person was obtained from District Court Judge Stephen Maxwell, upon application by Officer Earl Miels of the St. Paul Police Department Homicide Division.
Based on the opinion of Sergeant Nash that Miss Trimble had made the call on May 22, 1970, which lured Officer Sackett to his death, Captain Ernest Williams obtained a warrant for her arrest and also a warrant to search the premises where she lived. The information given to the judge to show probable cause for the arrest warrant was that Sergeant Nash after comparing tapes by spectrograms had identified Miss Trimble as the person who had made the call.
Miss Trimble was arrested on October 30, 1970, pursuant to the warrant so issued. She appeared specially in municipal court and objected to the jurisdiction of the court on the grounds that the warrant was improperly granted and the arrest illegal. On November 12, 1970, she was indicted by a grand jury and charged with murder in the first degree. She then applied to the
district court for a writ of habeas corpus which, after hearing, was discharged. This appeal was taken therefrom.
At the outset, it should be noted that we deal here with the sufficiency of the proof to justify issuance of an arrest and search warrant, not with the sufficiency of proof to sustain a conviction.
In Henry v. United States, 361 U. S. 98, 102, 80 S. Ct. 168, 171, 4 L. ed. 2d 134, 138 (1959), the court stated that “probable cause” for arrest did not require evidence sufficient to establish guilt. This court followed that decision in State v. Fish, 280 Minn. 163, 169, 159 N. W. 2d 786, 790 (1968), saying:
“* * * [P] robable cause for arrest exists where there is a reasonable ground for suspicion supported by circumstances sufficiently strong in themselves to warrant a cautious man in believing the accused to be guilty. Probable cause is concerned with probabilities and is something more than mere suspicion and something less than evidence which would sustain a conviction.”
See, also, State v. Sorenson, 270 Minn. 186, 198, 134 N. W. 2d 115, 124 (1965), where we said:
“Since the police deal with probabilities, not legal proof, in deciding whether to arrest without a warrant, this means, according to Jackson v. United States, 112 App. D. C. 260, 263, 302 F. (2d) 194, 197:
“ * * The police are entitled to rely upon hearsay which would be inadmissible in judicial proceedings, and upon various other factors which would not be admissible in evidence against an accused on trial. Rumor, gossip, common conversation, general reputation, hearsay, all these may well go into the police evaluation of the total situation which confronts them.’
“The arresting officer need not have in hand evidence which would suffice to convict. Wong Sun v. United States, 371 U. S. 471, 83 S. Ct. 407, 9 L. ed. (2d) 441.”
At the habeas corpus hearing three experts testified. Dr. Oscar Tosi, who testified for the state, is a professor at Michigan State University. He has an impressive background of education and experience in audiology and speech sciences and electronics. He belongs to many acoustical societies. Detective Sergeant Nash of the Michigan State Police, who also testified for the state, is in charge of the voice identification unit of the Michigan State Police Crime Laboratory. He has studied under Kersta, who will be mentioned hereinafter, at voiceprint laboratories in New Jersey, and has been trained in the use of the voiceprint-sound spectrograph. He does his work in this field with that machine and was, at the time he testified, a student at Michigan State University majoring in audiology and speech sciences. He also has studied under Dr. Tosi.
Dr. Peter Nielsen Ladefoged testified for the defense. He is professor of phonetics at the University of California in Los Angeles, where he has been employed for 8 years. Prior to that time he was a phonetician in the University of Edinburgh, Scotland, in the Department of Phonetics there, where he had been employed for many years. The qualifications of all of these experts are hardly open to question.
The spectrograph was invented during World War II, but its use in identification of people began about 10 years ago. At that time, Mr. Lawrence G. Kersta and a team of researchers at Bell Laboratories began experimenting with voice identification by spectrograph analysis. Their experiments showed a remarkably high, 99.65-percent success in identifying speakers by comparison of voiceprints.
These claims of almost absolute identification by use of voice-prints were criticized by many scholars in the field of acoustics as being unscientific. It is clear also that Kersta has a financial interest in the development of the process since by 1966 Kersta had left Bell Laboratories and founded his own firm, Voiceprint Laboratories, which began manufacture of spectrograph equipment for commercial distribution. For criticism of Kersta’s
work, see P. Ladefoged and R. Vanderslice,
The “Voieeprint” Mystique,
Working Papers in Phonetics 7 (U.C.L.A. 1967).
Because of the criticism being leveled at Kersta and his method of identification as opposed to his claims of success, the Acoustical Society of America’s Technical Committee on Speech Communication made a study of the legal implications of speaker identification by the Kersta method. That report ended:
“We conclude that the available results are inadequate to establish the reliability of voice identification by spectrograms. We believe this conclusion is shared by most scientists who are knowledgeable about speech; hence, many of them are deeply concerned about the use of spectrographic evidence in the courts. Procedures exist, as we have suggested, by which the reliability of voice identification methods can be evaluated. We believe that such validation is urgently required.” Bolt et al.,
Speaker Identification of Speech Spectrograms: A Scientist’s View of its Reliability for Legal Purposes,
47 Journal of the Acoustical Society of America 597, 603 (1970).
In the article, the group of scientists called for additional research into speaker identification by spectrographic analysis, especially in more controlled experiments. The Kersta experiments had always been done with closed groups, i. e. the spectrogram of the unknown voice was always that of one of the group of known voices being used. This was highly criticized in the article, where it was pointed out that only by using an “open” group — i. e. where the unknown voice may or may not be contained among the known voices — could it be said that a speaker could be identified. Under the “closed” Kersta tests, all the examiner did was find the best match among a number of given voices. If an “open” test were used, it would be more closely analogous to a criminal investigation where the examiner would be required to test a sample voice against several voices, any one or none of which might match.
Identification of a person by recognition of his voice from
hearing it at different times is generally upheld as admissible. Annotation, 70 A. L. R. 2d 995; 29 Am. Jur. 2d, Evidence,. §§ 368 and 832; 31A C. J. S., Evidence, § 188.
We have followed this general rule for many years. William Deering & Co. v. Shumpik, 67 Minn. 348, 69 N. W. 1088 (1897), where it was held that evidence of identity of a person conversing through a telephone was sufficient to establish identity. In Katzmarek v. Weber Brokerage Co. 214 Minn. 580, 583, 8 N. W. 2d 822, 824 (1943), we said:
“* * * is wen settled that conversations had over the telephone are admissible in evidence where the witness testified that he recognized the voice over the phone. * * * However, the identity of the person called can be established with reasonable certainty by means of the surrounding facts and circumstances.”
In City of St. Paul v. Caulfield, 254 Minn. 142, 144, 94 N. W. 2d 263, 264 (1959), we said:
“It is the rule in this state that the foundation for admission of testimony as to the identity of the voice of a telephone caller is sufficient when it appears that the witness to whom the telephone call is made testifies that he is reasonably certain as to the voice of such caller and can identify it.”
While most of the cases dealing with the subject relate to conversations over the telephone or to one listening to a person speak and comparing the voice with that heard prior thereto, the rule has been applied to evidence of conversations received by other mechanical means. The general rule is stated in 31A C. J. S., Evidence, § 188, p. 509:
“* * * The rules governing the admissibility of testimony of telephone conversations have been applied to evidence of conversations received by other mechanical means. Thus, testimony of a conversation from another room, received through an electrical apparatus transmitting the voices through ear phones, is
competent, provided the identity of the speaker is established.
“* * * A recording of a conversation is admissible to the same extent that testimony as to such conversation would be admissible * *
Thus, in Calumet Broadcasting Corp. v. F. C. C. 82 App. D. C. 59, 62, 160 F. 2d 285, 288 (1947), the court said:
“Appellant urges that the admission of certain phonographic records in evidence invalidated the proceedings. These records were identified as being mechanical recordings of the conversation at the Thompson conference, above discussed. It was further shown that the recording was done at the request of Mr. Thompson but without the knowledge of the other participants. Amazing though this revelation may be, it could not prevent the use of the recordings if and when they later came into the possession of Government authorities. No participation of those authorities in the making of the recordings was shown. A full foundation for their use to impeach the testimony of the president of the applicant was laid. We do not find their admission erroneous.”
In N. V. Simons’ Metaalhandel v. Hyman-Michaels Co. 7 App. Div. 2d 840, 181 N. Y. S. 2d 267, 268 (1959), the court said:
«* * * our courts neither prohibit the recording of a telephone conversation by one of the participants, nor the use of the recording as evidence * *
In Kilpatrick v. Kilpatrick, 123 Conn. 218, 224, 193 A. 765, 768 (1937), the Connecticut court held that conversations intercepted by a microphone placed in the defendant’s room were admissible, saying:
“* * * The intervention of this mechanical means of audition no more rendered the words spoken inadmissible than would that of a telephone under similar circumstances.”
State v. Sawyer, 266 Wis. 494, 63 N. W. 2d 749, certiorari denied, 348 U. S. 855, 75 S. Ct. 80, 99 L. ed. 674 (1954), again involved a conversation over a telephone.
There are other cases that could be cited, but we think these are sufficient to show that identification by voice, whether the words are spoken in a lineup, over a telephone, or recorded by mechanical means, is generally admissible for identification purposes.
Compelling a person to speak certain words in a lineup for identification is not a violation of any constitutional right. State v. King, 44 N. J. 346, 209 A. 2d 110, 9 A. L. K. 3d 847 (1965). For the same case on a petition for a Federal habeas corpus, see King v. Pinto, 376 F. 2d 593, 595 (3 Cir. 1967), where the court said:
“Appellant’s remaining complaint deals with use of his rainhat and coat and voice in the identification lineup. These were not testimonial. They did not rise to the height of self incrimination. Of the numerous decisions supporting this doctrine, Schmerber v. State of California, 384 U. S. 757, 86 S. Ct. 1826, 16 L. ed. 2d 908 (1966) is perhaps of most present significance.”
While United States v. Wade, 388 U. S. 218, 222, 87 S. Ct. 1926, 1930, 18 L. ed. 2d 1149, 1155 (1967), dealt mainly with the right to counsel at a lineup, the court in a sharply divided decision said with respect to compelling a person to speak certain words at a lineup for purposes of identification:
“* * * [(Compelling Wade to speak within hearing distance of the witnesses, even to utter words purportedly uttered by the robber, was not compulsion to utter statements of a ‘testimonial’ nature; he was required to use his voice as an identifying physical characteristic, not to speak his guilt. We held in
Schmerber, supra,
[384 U. S.] at 761, that the distinction to be drawn under the Fifth Amendment privilege against self-incrimination is one between an ¿ccused’s ‘communications’ in whatever form, vocal or physical, and ‘compulsion which makes a suspect or accused the source of “real or physical evidence,” ’ * * *. We recognized that ‘both federal and state courts have usually held that * * * [the privilege] offers no protection against compulsión to submit to fingerprinting, photography, or measurements, to write
or speak for identification, to appear in court, to stand, to.assume a stance, to walk, or to make a particular gesture.’ * * * None of these activities becomes testimonial within the scope of the privilege because required of the accused in a pretrial lineup.”
It would seem to follow that if identification can be made by comparing a voice over a telephone or by requiring an accused to speak certain words in the presence of an accuser in a lineup or by means of other mechanical recording, the two tapes involved in this case, one of the voice of a known person and the other of an unknown, should be admissible for the purpose of comparison aurally, the same as if the words were spoken in some other manner, assuming that a foundation is laid showing that there has been no alteration of the tape and that the tape is mechanically perfected to the point where voices can be identified from it.
So far as we can determine, only three courts have passed on admissibility of spectrograms such as we have here for the purpose of identification. In State v. Cary, 49 N. J. 343, 230 A. 2d 384 (1967), the New Jersey court remanded the case to the trial court for further pretrial proceedings. In so doing, the court said (49 N. J. 352, 230 A. 2d 389):
“An actual demonstration which shows the accuracy of the voiceprint process would be desirable. The testimony of experts in addition to Mr. Kersta who believe that voiceprint identification is scientifically sound would be helpful. Although we do not wish to dictate in advance the details of the hearing, we do feel that something more than the bare opinion of one man, however qualified, is required. Certainly, the prosecutor must satisfy the trial judge that identification by voiceprint technique and equipment has a sufficient scientific basis to produce uniform and reasonably reliable results and will contribute materially to the ascertainment of truth.”
It is quite noticeable that the New Jersey Supreme Court was
reluctant to rule that spectrograms were inadmissible. At the pretrial proceedings held upon remand, State v. Cary, 99 N. J. Super. 323, 239 A. 2d 680 (1968), Dr. Tosi and Dr. Ladefoged were witnesses. At that time, Dr. Tosi was of the opinion that there had not been sufficient experiments with the use of spectrograms to establish that they are reliable. The court finally came to the conclusion (99 N. J. Super. 334, 239 A. 2d 685) that identification opinion based on use of the tape recordings, would not, “as of this time,” be admissible as evidence in the case.
In People v. King, 266 Cal. App. 2d 437, 72 Cal. Rptr. 478 (1968), the California Court of Appeals (Second District) exhaustively reviewed the experiments in the use of spectrograms up to that time. It, too, came to the conclusion that the spectrograms were not then admissible, but quoted from Frye v. United States, 54 App. D. C. 46, 47, 293 F. 1013, 1014, 34 A. L. R. 145, 146 (1923), the following:
“* * * Just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.”
United States v. Wright, 17 U.S.C.M.A. 183, 37 C.M.R. 447 (1967), involved the review of a general court-martial. Here, again, an exhaustive discussion of the development and use of voice identification by spectrograph was undertaken. Contrary to the New Jersey and California courts, the court, with one judge dissenting, held that voiceprints such as we have were admissible. The court said (37 C.M.R. 453) :
“Mr. Kersta’s testimony established that his system of voice identification had, experimentally and in practical application,
demonstrated a high degree of accuracy and, further, that he was personally qualified to testify as an expert on comparisons of sound patterns made by human voices. True, two defense expert witnesses expressed reservations as to the complete reliability of Mr. Kersta’s system and procedures. The specifics of their reservations need not detain me. Courts have consistently recognized the admissibility of the testimony of experts in areas where there is neither infallibility of result nor unanimity of opinion as to the existence
vel non
of a particular condition or fact. * * * Here, the tape recording of one of the obscene calls and the recording of the test call made by the accused were both before the court-martial. Each was played in open court. Since voice identification by ear is fully acceptable in the courts, the court members could thus determine for themselves the margin of error, if any, in Mr. Kersta’s expert opinion.”
Since the decisions in the three casés mentioned above, Dr. Tosi has conducted extensive experiments over about 4 years under the auspices of the Michigan State Police Department. The experiments, funded by the United States Government under a grant of $300,000, have been carried on for the purpose of studying speech spectrograms for voice identification. Dr. Tosi has compared over 34,000 samples of voiceprints and was of the opinion at the time of the hearing on the case now before us that the voiceprint spectrograms were reliable. His testimony in that respect was:
“Q. Dr. Tosi, could you express an opinion as to the reliability of voiceprint examination by the Kersta voiceprint method by a trained examiner, by the Kersta voiceprint spectrogram method by a trained examiner, who is given all of the time that he needs to make an examination, who listens to the spoken material as well as analyzes the voiceprint, who is allowed to grade and produce an opinion as to the reliability of what he is saying, who is given adequate and sufficient training in the manufacture of voiceprints and reading of voiceprints, who has
a high degree of job responsibility and experience? Could you express an opinion as to the degree of reliability of such an examiner’s opinion?
“A. Well, yes, certainly. Providing that all these conditions that you have expressed, especially that the examiner is responsible and he is allowed to say, Well, I don’t know, I cannot produce in this case an identification,’ and only in those cases in which he is absolutely sure of his statement, I think that then the method is very highly extremely reliable.
“Q. Extremely reliable?
. “A. Extremely reliable.”
Sergeant Nash of the Michigan State Police, who has been trained in the use of voiceprint identification, listened to the tape recordings and also prepared spectrograms from them. His testimony :
“Q. Did you listen with your ears to both of the tapes submitted ?
“A. Oh, yes.
“Q. So that — and then did you ultimately come to an opinion in this particular case as to, from your analysis of the voiceprint submitted from and received by you on June 1st, and the known submitted to you on about September 29th, were you able to form an opinion as to the identity of the unknown caller as it would relate to the caller on the tape that you received June 1st of 1970?
“A. Yes, I did.
“Q. What is your opinion?
“A. In my opinion, the voice of Connie Trimble and the voice that made the call that I received and the tape on June 1, 1970, are one and the same and could be no other.
“Q. And could be no other person?
“A. Yes.
“Q. What is your degree of certainty in your expression of this opinion?
“A. Beyond any doubt.”
While Dr. Ladefoged, who testified for the defense, was still of the opinion that spectrographic voiceprint identification was too uncertain to be admitted as evidence in a trial, he did agree that use of the spectrograms to corroborate an opinion formed from listening to the tapes would make the opinion more reliable. His testimony in that regard was as follows:
“Q. Of course, the human ear is going to vary greatly, the human auditory system, from one human to another in that respect, is it not?
“A. And the skill in listening was affected, yes.
“Q. And then, of course, if we threw in the factor of memory, that also would tend to affect it, too, would it not?
“A. Yes, indeed.
“Q. Well, then, would you agree with me with this proposition, Doctor: That the human ear listening to voices, coupled with a skilled spectrogram operator analyzing spectrograms, spectrographs, could be more accurate than the human ear alone?
“A. Again, there is evidence to show you’re correct, yes.
“Q. And particularly if the human ear could hear the two sounds at the same time or at a very close contemporary time, matter of seconds or minutes,—
“A.
Uh-huh.
“Q. —and analyze also, also independently, Voiceprint spectrograms, could he not achieve a higher degree of reliability in his opinion in that manner ?
“A. He would certainly be helped by having the spectrograms and achieve a higher reliability, yes.
“Q. To his ear?
“Q. So that, in a sense, couldn’t we say that those two systems might complement each other in assisting a person to render an opinion?
“A.
Yes. Definitely.”
All of the articles mentioned above were written and State v. Cary, 49 N. J. 343, 230 A. 2d 384; Id. 99 N. J. Super. 323, 239 A. 2d 680; People v. King,
supra;
and United States v. Wright,
supra,
were decided prior to the experiments conducted by Dr. Tosi. Dr. Ladefoged agreed with Dr. Tosi’s experiments “[a]s far as [he] has gone.” He also said that Dr. Tosi’s results were accepted by the scientific community with certain limitations. He contended that all of the experiments were made with male voices and that it is more difficult to compare female voices by spectrograph due to the fact that they normally have a higher pitch. Dr. Tosi disagreed with this. He claimed that he can identify female voices as well as male.
It is common knowledge that the opinion of an expert on an identification subject is seldom so infallible that others in his field do not disagree with him. But disagreement alone does not make the opinion inadmissible. Where experts disagree, it is for the factfinder, whether that be jury or court, to determine which is more credible and therefore more acceptable. In the field of medicine, it is not unusual to have doctors disagree as to the cause or effect of an illness or accident. The opinion of an expert is admissible, if at all, for the purpose of aiding the jury or the factfinder in a field where he has no particular knowledge or training. The weight and credibility to be given to the opinion of an expert lies with the factfinder. It is no different in this field than in any other.
In recent years there have been giant strides made in the perfection and improvement of many mechanical and electrical devices that are used not only in the field of medicine (see, An
notation, 66 A. L. R. 2d 536), but in other identification areas. The courts have been faced with the question of whether the results of the use of such devices are admissible or not in a great many cases. A glance at 7 Dunnell, Dig. (3 ed.) §§ 3332 to 3334, ought to convince anyone that the employment of expert testimony in many of the fields involved in litigation has greatly expanded in recent years.
With respect to the surreptitious taping of the appellant’s voice in the county jail, the court in People v. King, 266 Cal. App. 2d 437, 464, 72 Cal. Rptr. 478, 495, said:
“There is no violation of the Fifth Amendment in compelling an accused to repeat words for voice identification. It is not testimonial compulsion. (Schmerber v. California, 384 U. S. 757 [16 L. ed. 2d 908, 86 S. Ct. 1826].) It is an identification of physical characteristics.
“The use of the concealed tape in the county jail does not change this rule. While we do not condone this practice, the prosecution’s conduct was much less objectionable in this case than in Miller v. State of California, June 17, 1968, United States Supreme Court [392 U. S. 616, 88 S. Ct. 22, 20 L. ed. 2d 1332], where the United States Supreme Court dismissed a writ as improvidently granted where an informer was sent by the prosecution to live in appellant’s cell to obtain incriminating statements.”
In view of the fact that identification by aural voice comparison, either respecting telephone conversations or words spoken at a lineup, or recorded by other mechanical means is admissible, and the admission that voice comparisons by spectrograms corroborate identification by means of ear, we are convinced that spectrograms ought to be admissible at least for the purpose of corroborating opinions as to identification by means of ear alone. They ought also to be admissible for the purpose of impeachment. The weight and credibility of such evidence lie with the finder of facts, but that does not involve the question of admis
sibility. In this case, the information submitted to the magistrate for the purpose of determining probable cause was sufficient to justify the issuance of a warrant for arrest and a search warrant.
While we deal here with the sufficiency of proof to establish probable cause for the issuance of an arrest and search warrant, we are convinced that in the trial of the case spectrograms ought to be admissible for the purpose of corroborating voice identification by aural means if a sufficient foundation is laid to satisfy the trial judge that the expert whose opinion is sought is qualified to assist the factfinder in coming to the right conclusion. The qualification of an expert is normally left to the discretion of the trial judge and we think that ought to be the rule here as well as in other fields of scientific study.
Nor do we think that the manner in which a recording of Miss Trimble’s voice was obtained prevents its use for the purpose of identification in establishing probable cause. No information regarding Miss Trimble’s welfare status was elicited or used. All that was obtained was a sample of her voice. While the method used might be criticized by some, we think it did not violate any constitutional or statutory right. There was no interception of her conversation as such but only of the physical characteristics of her voice. Osborn v. United States, 385 U. S. 323, 87 S. Ct. 429, 17 L. ed. 2d 394 (1966); State v. King, 44 N. J. 346, 209 A. 2d 110, 9 A. L. R. 3d 847; People v. King, 266 Cal. App. 2d 437, 464, 72 Cal. Rptr. 478, 495.
The decision of the trial court denying a writ of habeas corpus is affirmed.
Mr. Justice Nelson and Mr. Justice Hachey took no part in the consideration or decision of this case.