BEASLEY, Judge.

Marion Franklin Campbell was convicted of burglary (OCGA § 16-7-1), criminal attempt (OCGA § 16-4-1), and robbery (OCGA § 16-8-40). He enumerates as error the admission of an audio tape of a pretrial voice lineup, arguing that he did not consent to this use of his voice and that the lineup was impermissibly suggestive.

At about 5:00 a.m. on April 2, 1995, an intruder unlawfully entered a woman's apartment, threatened her into submission, and unsuccessfully tried to rape her. She pleaded with him not to hurt her and offered all her cash for him to leave. He dragged her around her dark apartment until they found her purse, whereupon she gave him $200. Tearing out her phones, he placed her in the bathroom and left. The episode lasted about fifteen minutes, during which time the intruder spoke often to the victim.

Immediately thereafter, the victim described the intruder to police as a white male about six feet tall, thin build, wearing a dark jacket, jeans, and work boots, and having a strong odor. She described his voice as slow and Southern. Because officers had encountered and talked with such a person in the vicinity the same night and had obtained his identity and recorded his tag number, they located a photo of him and showed it that same day to the victim with five photos of other white males. She quickly identified Campbell.

Four days later the police brought Campbell to the station in for questioning and advised him of his Miranda rights. He signed a form consenting to the police recording him by audio and video machines while he made a statement. The victim had been trained and was experienced in recognizing voices, so the police dubbed a few sentences from the audio recording onto a new tape and recorded another five voices from white men with slow Southern accents who repeated the same few sentences. She readily identified Campbell's voice as the voice of her attacker.

Campbell unsuccessfully moved to exclude the pretrial voice lineup on the grounds that it was impermissibly suggestive. At trial, the victim unequivocally pointed to Campbell as the perpetrator, and the pretrial voice lineup was also introduced. He moved for a new trial on the grounds that the lineup was impermissibly suggestive and, for the first time, that he had not consented to the use of his voice in the lineup. A new trial was denied.

1. Campbell does not identify any state constitutional grounds for the errors alleged on appeal, but cites only cases based on the federal constitution. See Redd v. State, 154 Ga. App. 373, 374 (2) ( 268 S.E.2d 423) (1980) (federal due process guarantee); Beasley v. State, 204 Ga. App. 214, 216 (1) ( 419 S.E.2d 92) (1992) (Fourth Amendment rights). We will therefore restrict our inquiry to the federal constitution. See Stephenson v. State, 206 Ga. App. 273 ( 424 S.E.2d 816) (1992) (court will not sua sponte consider state constitutional standards). Any independent rights Campbell has under the state constitution are waived.

2. Campbell contends the voice lineup was impermissibly suggestive. Under the due process guarantee of the federal constitution, an impermissibly suggestive pretrial lineup can deprive the defendant of a fair trial and invalidate an in-court identification. Simmons v. United States, 390 U.S. 377 ( 88 SC 967, 19 L.Ed.2d 1247) (1968). See Heyward v. State, 236 Ga. 526, 527 ( 224 S.E.2d 383) (1976). Simmons "set forth a two-part test to determine whether the procedure followed violated due process. The first inquiry is whether the procedure used was impermissibly suggestive. Only if it was, need the court consider the second question: whether there was a very substantial likelihood of irreparable misidentification." (Citations and punctuation omitted.) Tate v. State, 153 Ga. App. 508, 509 (1) ( 265 S.E.2d 818) (1980).

(a) Because the purpose of objecting to pretrial identification is to exclude in-court identification, a defendant's failure to object to the in-court identification can be fatal to the claim of prejudice. Where "in-court identification of appellant was not challenged, any error in admitting the out-of-court identification was harmless. [Cits.]" Evans v. State, 177 Ga. App. 820, 821 (1) (b) ( 341 S.E.2d 483) (1986).

Campbell did not object to the victim's unequivocal visual identification of him in the courtroom. Thus, the admission of the pretrial voice lineup identification appears harmless. But because the voice lineup identification was based on sound rather than sight, it reinforced the validity of the in-court visual identification; we will therefore consider Campbell's arguments.

(b) Campbell argues that the tape impermissibly suggested his voice because (i) in the introduction the police inform the victim that the voice of the suspect in custody is on the tape; (ii) the background noise surrounding the recording of his voice is noticeably different from the others; (iii) he is cut off while continuing to speak whereas the others stop after they have finished their statements; and (iv) the other five speakers speak in dull monotones, which suggests they were reading, whereas Campbell speaks with intonation and expression.

Telling the victim that the suspect's voice is on the tape did not preclude her from determining that the voice of the man who came to her apartment was not on the tape. The supervising officer specifically testified and the court found that the victim understood she was only to determine whether the voice of her intruder was on the tape and if so, which one. This finding is not clearly erroneous. See Phillips v. State, 204 Ga. App. 698, 700-701 (1) ( 420 S.E.2d 316) (1992). Nevertheless, providing such information should not be done because it could influence the person being tested. At the least, it raises that question, and it affects the weight of the identification.

The background noise tape is not remarkably different amongst the voices. Each differs slightly from the others. Similarly, although Campbell's voice is cut off at the end, another voice is cut off or stopped at the beginning and two others in the middle. There is no discernible pattern pointing to Campbell. The same is true of the intonations. Some are more animated than others, and Campbell's is not remarkably different. And there is no evidence the police told the victim that the suspect was the only one who did not read his statement.

In Williams v. State, 163 Ga. App. 866 ( 295 S.E.2d 361) (1982), the appellant pointed to varying volume, manner of speech, and background noises to argue that his particular segment was singular and thus irreparably suggestive. We rejected this argument and explained "[t]hough there are certain differences in the volume and background noises, this indicates only that the voices were recorded at different times and different locations. Each of the identifying witnesses testified that he or she recognized the voice because of the peculiarities of the voice and not because the voice was singularly isolated by its setting. The jury was able to hear the tape played and judge for itself the likelihood of misidentification. This objection goes to the weight and credibility of the evidence [cits.], not to its admissibility." Id. at 869 (3). The same obtains here.

All things considered, the evidence supports the court's finding that the tape was not impermissibly suggestive. We need not even reach the second question.

(c) Even if the tape were suggestive, Campbell does not show it was so "impermissibly suggestive as to give rise to a very substantial likelihood of irreparable misidentification." Simmons, supra, 390 U.S. at 384. "The factors to be considered include the opportunity of the witness to view or hear the criminal at the time of the crime, the witness' degree of attention, the accuracy of the witness' prior description of the criminal, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation." (Citations and punctuation omitted). Jefferson v. State, 206 Ga. App. 544, 546 (2) ( 425 S.E.2d 915) (1992). See Neil v. Biggers, 409 U.S. 188, 199-200 ( 93 SC 375, 34 L.Ed.2d 401) (1972).

Just as the victim in Jefferson, the victim in this case, who in this case was able to recognize voices, "was clearly focused on her assailant's voice, had ample opportunity to hear him speak during the crime, and described his voice in detail. Her identification of his voice was unequivocal and made [within days] of the crime. Under all these circumstances, there was no substantial likelihood of irreparable misidentification. [Cits.]" Jefferson, supra, 206 Ga. App. at 546.

Even if there were some taint in the pretrial identification, "a witness' in-court identification may be admitted if it has an independent origin." (Citations and punctuation omitted). Phillips, supra, 204 Ga. App. at 702 (2). Before the voice lineup, the victim had unequivocally identified Campbell in a photo lineup. She visually identified him again in the courtroom. Any error in admitting the voice lineup was harmless.

Furthermore, the victim's identification of defendant fit police identification of Campbell as observed just before and just after the crimes were committed. Shortly past 1:00 a.m., an officer came upon Campbell in his parked car in a shopping center about half a mile from the victim's apartment. He was alone, asleep, haggard, dirty, and smelling of alcohol. The officer obtained his identification and auto tag number. Close to 5:00 a.m., the officer saw the car parked a few hundred yards from the victim's apartment. A white male was in it, but the officer did not see his face. When the officer drove by again shortly after 5:00 a.m., no one was in the car. At 5:25 a.m. at an under-speed limit six-minute driving distance from the victim's apartment, the officer saw Campbell headed towards the highway. Eighteen minutes later the officer received the call regarding the victim's ordeal.

3. Campbell contends he never consented to his voice being used in the voice lineup and therefore the lineup was inadmissible. He claims he consented only to the use of the content of his speech, not the manner in which he spoke.

(a) First, he does not identify the specific constitutional right violated by failing to obtain his consent. He cites Beasley, supra, 204 Ga. App. at 216 (1), which held that under the federal Fourth Amendment a urine sample was inadmissible for failure to obtain proper consent. But in United States v. Dionisio, 410 U.S. 1, 14 ( 93 SC 764, 35 L.Ed.2d 67) (1973), "the Supreme Court ruled that voice exemplars were not within the scope of the Fourth Amendment protections. `The physical characteristics of a person's voice, its tone and manner, as opposed to the content of a specific conversation, are constantly exposed to the public. Like a man's facial characteristics . . ., his voice is repeatedly produced for others to hear. No person can have a reasonable expectation that others will not know the sound of his voice, any more than he can reasonably expect that his face will be a mystery to the world.' [Cit.]" Reeves v. State, 139 Ga. App. 214, 215-216 (2) ( 228 S.E.2d 201) (1976).

Nor can a voice exemplar be a violation of the Sixth Amendment right against self-incrimination. "`Requiring a suspect to give a voice exemplar for identification purposes does not violate his privilege against self-incrimination accorded by the United States . . . Constitution. [Cit.]' [Cit.]" Jenkins v. State, 167 Ga. App. 840, 841 (1) ( 308 S.E.2d 14) (1983). Since the police can require a suspect to give a voice exemplar without violating his rights, Campbell's alleged lack of consent is irrelevant.

(b) Second, even if consent were required, Campbell's grounds for objecting to the voice lineup both in his pretrial motion and at trial did not include the issue of lack of consent. It arose only in his motion for new trial, which was too late. "In order to preserve an objection upon a specific ground for appeal, the objection must be made at trial upon that specific ground. Norman v. State, 197 Ga. App. 333, 334 [(2)] ( 398 S.E.2d 395) (1990)." Smith v. State, 222 Ga. App. 366, 368 (3) (a) ( 474 S.E.2d 272) (1996). Objecting on specific grounds waives the grounds not asserted. McGee v. State, 205 Ga. App. 722, 727-728 (10) ( 423 S.E.2d 666) (1992) (court would not hear new grounds as to why audio tape should not be admitted).

(c) Aside from waiver, the argument fails on its merits. When Campbell consented to the audio recording of his statement, Campbell acknowledged in writing that he received the warning: "ANYTHING YOU SAY CAN AND WILL BE USED AGAINST YOU IN A COURT OF LAW." He claims this only gave the State permission to use the content of his statement, for impeachment purposes, and not the manner in which he spoke (his accent and voice inflections).

In Gadson v. State, 223 Ga. App. 342 ( 477 S.E.2d 598) (1996), defendant was arrested for assaulting a particular victim. His written consent allowing the State to draw his blood stated "`[a]nything found by them can and may be used as evidence in any court of law.'" Id. at 345 (4). When the blood results were used in prosecutions involving other victims, he urged this exceeded the scope of his consent. We held this language was sufficiently broad to allow the State to use the blood results in other prosecutions, even without having explained such a possibility to defendant.

Similarly, Campbell's general consent included the use of the audio recording of his statement in a voice lineup. Campbell's citations to Beasley, supra, State v. Gerace, 210 Ga. App. 874 ( 437 S.E.2d 862) (1993), and Deckard v. State, 210 Ga. App. 421 ( 436 S.E.2d 536) (1993), are unavailing, for in each the police either told defendant the evidence was sought only for a specific purpose or affirmatively misled defendant as to the consequences of his not consenting.

Judgment affirmed. McMurray, P.J., and Smith, J., concur.

BEASLEY, Judge.

DECIDED AUGUST 26, 1997.