Defendant contends that the in-court identification by Rita was tainted by an improper pretrial photographic and one-man lineup identifications.
The practice of showing suspects singly to persons for the purpose of identification and not as part of a lineup is usually questioned. Stovall v. Denno (1967), 388 U. S. 293, 302, 87 Sup. Ct. 1967, 18 L. Ed. 2d 1199; and Foster v. California (1969), 394 U. S. 440, 89 Sup. Ct. 1127, 22 L. Ed. 2d 402. The reason is that such identifications are deemed to be unfair because the atmosphere is highly suggestive and conducive enough to cause an irreparable mistaken identification. This is also true to some extent with photographic identifications. The improper employment of such photographs by the police may cause the identifying witness to err and lead to a misidentification. Further, regardless of how the misidentification comes about, the witness thereafter is apt to retain in his memory the image of the photograph rather than of the person actually seen. This thereby reduces the trustworthiness of a subsequent lineup or courtroom identification. Despite these hazards the need for both types of identification is at times necessary. Photographic identification is necessary in order to aid and effectuate criminal law enforcement “. . . from the standpoint both of apprehending offenders and of sparing innocent suspects the ignominy of arrest by allowing eyewitnesses to exonerate them through scrutiny of photographs. . . .” Simmons v. United States (1968), 390 U. S. 377, 384, 88 Sup. Ct. 967, 19 L. Ed. 2d 1247. Showing a suspect singly is necessary in situations where the usual lineup procedures are inadequate or simply cannot be employed, Stovall v. Denno, supra; for example, when a witness is seriously *441injured and about to die in a hospital. In such cases time is crucial and the only feasible means of identification is to bring the accused to the witness. Therefore, a claimed violation of due process of law because of a one-man lineup or showing photographs depends on “. . . the totality of [the] surrounding circumstances. . . .” As stated in Foster, Stovall and Simmons, each case must be considered on its own facts and it is only a denial of due process of law when the conduct of the identification procedures are so impermissibly and unnecessarily suggestive and conducive as to give rise to a very substantial likelihood of irreparable misidentification. This court is in accord with this doctrine. State v. Brown (1971), 50 Wis. 2d 565, 185 N. W. 2d 323; Zdiarstek v. State (1972), 53 Wis. 2d 420, 192 N. W. 2d 833; Wright v. State (1970), 46 Wis. 2d 75, 175 N. W. 2d 646; and State v. Biastock (1969), 42 Wis. 2d 525, 167 N. W. 2d 231.
The defendant argues that the photograph used in the August 12th identification violates his fourth amendment rights under Davis v. Mississippi (1969), 394 U. S. 721, 89 Sup. Ct. 1394, 22 L. Ed. 2d 676. In Davis, the United States Supreme Court held that the fourth amendment applies to involuntary detentions occurring at the investigatory stage as well as the accusatory stage. The court said that illegally detaining one for the “sole purpose” of obtaining fingerprints violates the individual’s fourth amendment rights, and that evidence of the fingerprints so taken was inadmissible in a state court by virtue of the fourteenth amendment. By analogy defendant argues that the August 4th photographs fall within the meaning of Davis. There is nothing in the record to show that the August 4th photographs were ever used at the August 12th identification. Such photographs were never introduced into evidence, nor does Rita’s testimony imply that such photographs were used. *442Other photographs of the defendant were also taken when he was legally arrested. The point is that the photographs used could have been procured from many sources. The defendant maintains though that it is the state’s burden to prove that the August 4th photographs were not used under the doctrines of State v. Brown, supra. This burden did not exist. The instant case was tried before the Brown decision, which was prospective in application and not retroactive. Even if Brown was applicable, the case is still of no help because before the burden of proof is cast upon the state the defendant must prove the illegality of pretrial identifications. The state then only has to prove the in-court identification is of independent origin and untainted by the photographic identification. The defendant maintains that an inference of an improper photographic identification exists in the record. The defendant did testify that such photographs were taken, but without some showing that the August 4th and August 12th events are somehow connected the inference fails and only conjecture and speculation are left. Further, the lack of evidence in the record can raise no improper inferences because the defendant had ample opportunity and did intensively cross-examine and recross-examine the state’s witnesses. Quinn v. State (1971), 50 Wis. 2d 96, 183 N. W. 2d 61.
As for the photographic identification itself, it does not appear to have been conducted in any impermissibly suggestive manner. Rita was shown photographs throughout the year, and never picked out or misidentified anyone. Finally, after going through numerous pictures that year, she picked one picture out of twelve for the first time. The assailant’s photograph was the defendant. The police did not in any manner suggest or indicate that defendant’s picture was in that selection.
The one-man lineup conducted on August 13th was improperly carried out and impermissibly suggestive. The defendant was the only one there for the purposes *443of identification when confronted with his accuser. Further, a police officer suggested to Rita before she identified the defendant that the defendant admitted to committing such a crime. Although this confrontation is improper, the question remains whether under “the totality of the circumstances” this was sufficient enough to taint the in-court identification — thereby preventing it to independently stand and causing defendant to be convicted without due process of law.
The defendant was identified in court on a basis sufficiently distinguishable and independent of the improper one-to-one confrontation. The in-court identification clearly purged itself of any primary or prior taint. State v. Brown, supra. Rita observed the defendant throughout his attack. The incident was a direct eye-to-eye confrontation between the defendant and herself. She not only testified that she observed the defendant but even saw and described the weapon he used. The identification was based on her own personal knowledge and observation at the time of the offense. She had more than a sufficient opportunity to see his face. She even talked with him and scuffled with him when he attacked her. The lack of taint on the in-court identification is further borne out by the fact that she never made any erroneous identifications. She correctly identified him from the August 12th photographs and the in-person identification on the 13th. She stated throughout the trial and the pretrial identification procedures that she would not identify anyone unless she was absolutely positive, since she considered an identification to be “a very serious accusation.” The pretrial errors in identification, if any, were harmless beyond reasonable doubt. State v. Brown, supra.1 It is evident then that the court*444room identification can stand independently of the August 13th identification and even without the corroboration of the August 12th photographic identification.
By the Court. — Judgment affirmed.