In August 1985 the petitioner Wayne E. Dumond was convicted of kidnapping and rape and sentenced to consecutive terms of twenty years and life in the Arkansas Department of Correction. We affirmed. Dumond v. State, 290 Ark. 595, 721 S.W.2d 663 (1986). Petitioner Dumond now seeks an evidentiary hearing in circuit court pursuant to Criminal Procedure Rule 37.1, alleging that he did not receive effective assistance of counsel.
In assessing the effectiveness of counsel, we adhere to the standard set out by the United States Supreme Court in Strickland v. Washington, 466 U.S. 668 (1984). In Strickland, the court held:
A convicted defendant’s claim that counsel’s assistance was so defective as to require reversal of a conviction . . . has two components. First, the defendant must show that counsel’s performance was deficient. This requires showing that counsel made errors so serious that counsel was not functioning as the “counsel” guaranteed the defendant by the Sixth Amendment. Second, the defendant must show that the deficient performance prejudiced the defense. This requires showing that counsel’s errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable. Unless a defendant makes both showings, it cannot be said that the conviction . . . resulted from a breakdown in the adversary process that renders the result unreliable. [Strickland, 466 U.S. at 687.]
There is a strong presumption that counsel’s conduct at trial falls within a wide range of reasonable professional assistance, and the petitioner has the heavy burden of overcoming that presumption. A court hearing an ineffectiveness claim must consider the totality of the evidence before the jury and must judge the reasonableness of counsel’s conduct “on the facts of the particular case, viewed as of the time of counsel’s conduct.” Strickland, supra, at 690. An error by counsel, even if professionally unreasonable, does not warrant setting aside the judgment of a criminal proceeding if the error had no effect on the judgment. A reasonable probability that, but for counsel’s conduct, the result *384of the proceeding would have been different is a probability sufficient to undermine confidence in the outcome. Strickland, supra; Pruett v. State, 287 Ark. 124, 697 S.W.2d 872 (1985).
Petitioner was represented by two attorneys at trial — Larry Horton and William McArthur. Petitioner states that Horton retained McArthur less than a week before trial to assist him and that McArthur had little to do with preparation for trial and was not fully informed as to all the issues in the case. The record, however, reflects that at a pre-trial hearing on April 12, 1984, more than four months before trial, Horton advised the court that he had associated McArthur as co-counsel and requested that the court enter his name of record. To what extent McArthur participated prior to trial is not evident, but, as co-counsel of record for four months before trial and in light of his extensive participation in the trial itself, McArthur as well as Horton must be considered responsible for the quality of the defense afforded petitioner.
Petitioner’s initial allegation of counsel error concerns scientific testing of evidence. At trial a serologist at the Arkansas State Crime Laboratory testified for the state that he had examined semen stains on the victim’s clothing and body fluids obtained from the victim in a medical examination of her shortly after the attack. The serologist testified that the semen contained no spermatozoa and was from a man with type A blood. He further said that the blood type could be ascertained because the man was a secretor; that is, one of approximately seventy percent of the population who secrets his ABO blood grouping into all his body fluids. (The victim was determined to be a type O secretor.) Relying on estimates that forty percent of the population has type A blood and seventy percent are secretors, the serologist concluded that only twenty-eight percent of the population could have produced the semen. When the incidence of males having undergone a vasectomy was figured in, the percentage of male persons who could have produced the semen was narrowed to .6 percent. As petitioner had type A blood, was a secretor and had undergone a vasectomy, the evidence by the state serologist suggesting that he could have been the rapist was strong. Petitioner’s assertion of ineffective assistance of counsel arises out of his attorneys’ failure to counter the state’s serological evidence with serological evidence favorable to the defense. First, *385petitioner asserts that there was a test not conducted by the state crime lab which would have excluded him as the rapist. According to petitioner, the test, immunoglobulin allotyping, has been available since the early 1970’s and can be used to genetically type body fluids that are several years old. As support for the allegation that counsel was ineffective for not arranging for the allotyping test to be conducted before trial, petitioner states that in mid-1987 he obtained the original semen samples and submitted them to an independent serological expert who performed the allotyping and concluded that the semen was genetically inconsistent with petitioner and could not have come from him.
We need not be concerned with whether the serological expert’s report (appended to the petition) is as conclusive as petitioner claims since the report clearly constitutes new evidence and as such will not be considered. Petitioner as well as the state, which has joined in the request for an evidentiary hearing on the allotyping evidence, place great significance on the outcome of the test despite the fact that evidence obtained after a conviction is not a basis in itself for postconviction relief. A claim of new evidence is a direct rather than a collateral attack on the judgment and not within the purview of our postconviction rule. Chisum v. State, 274 Ark. 332, 625 S.W.2d 448 (1981). The issue therefore is not whether the new evidence is reliable but rather whether counsel’s conduct in not seeing to it that allotyping was done before trial was reasonable. As noted, in deciding an ineffectiveness claim the reasonableness of counsel’s challenged conduct must be judged on the facts of the particular case when viewed as of the time of counsel’s conduct. It must be determined whether in light of all the circumstances the identified acts or omissions were outside the wide range of professionally competent assistance. While counsel’s function is to make the adversarial process work in a particular case, it must be recognized that counsel is strongly presumed to have rendered adequate assistance and to have made all significant decisions in the exercise of reasonable professional judgment. Counsel has a duty to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary. A decision not to investigate must be directly assessed for reasonableness under all the circumstances, applying a heavy measure of deference to counsel’s judgments. Strickland, supra.
*386We cannot say that the failure of counsel to secure allotyping on the semen samples in this case constituted an unreasonable professional decision given the relative obscurity of the test and its general lack of availability. Petitioner himself acknowledges that allotyping, which was not available at the State Crime Laboratory, was also unavailable at Roche Labs in North Carolina where the experts he had consulted were employed. In fact, petitioner says very little about the specific availability or recognition of the tests by serological experts at the time of trial, although he makes the blanket statement that highly discriminatory genetic tests were available at the time of trial. Petitioner does cite one treatise to support his statement that PGM testing is highly discriminatory and has been known since 1964, but he fails to cite any recognized treatise on the availability and generally recognized definitiveness of allotyping. There are numerous scientific tests which could be conducted on physical evidence in a criminal trial and failure of counsel to seek a particular test will not amount to a denial of the counsel guaranteed by the sixth amendment unless it can be concluded that the test was one which any competent attorney under the same circumstances would have sought. Petitioner has not met his burden of demonstrating that allotyping was reasonably available and widely recognized at the time of his trial and generally accepted in the relevant scientific community and therefore should reasonably have been made a part of his defense. See Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).
Petitioner notes that McArthur asked the state’s expert whether PGM testing had been done and mentioned the failure of the state to do the test in his closing argument, indicating that he recognized its significance. As with allotyping, neither the state nor Roche Lab did the test but he contends that it was readily available at other labs.
Petitioner states that counsel consulted Dr. Jim Geyer and Dr. Ron Barwich of Roche Labs who had performed the same ABO testing done by the state’s expert. Geyer was unavailable to testify because of a reaction to a bee sting, and petitioner contends that counsel was remiss in not calling Dr. Barwich who was present at trial to contradict the state’s expert who testified there was a 99.4% probability that the semen was petitioner’s. He further alleges that Geyer and Barwich could both have testified *387that 25 % of the male population has the same blood group as the petitioner. When discussing another allegation, petitioner states that Horton told Geyer after trial that he did not call Barwich because he planned to put Walter Stevenson on the stand and then expose Stevenson as the assailant.
The state’s serologist narrowed the percentage of males in the population who matched the evidence to .6 percent, but he did not testify that there was a 99.4 percent certainty that the petitioner himself was the rapist. Even if his testimony could be so construed, on cross-examination by McArthur he conceded that he could not say that his figures were accurate based on the testing done and that he could not say with certainty that spermatozoa were not present when the attacker ejaculated. The state’s expert further conceded that with the testing he had done he could only narrow the group of individuals from whom the semen could have come to twenty-eight percent of the male population.
When assessing an attorney’s decision not to call a particular witness or witnesses, it must be taken into account that such a decision is largely a matter of professional judgment which advocates could endlessly debate, and the fact that there was a witness or witnesses who could have offered testimony beneficial to the defense is not in itself proof of counsel’s ineffectiveness. Tackett v. State, 284 Ark. 211, 680 S.W.2d 696 (1984). In light of the cross-examination by petitioner’s attorney which resulted in the serologist’s admitting that the semen could have come from twenty-eight percent of the male population, petitioner has not shown that counsel’s decision not to call Geyer or Barwich was prejudicial to the defense or more than a simple matter of professional judgment.
In a footnote petitioner argues that ABO typing alone is no longer sufficient evidence in rape and paternity cases since other, more definitive testing, is available. He urges that the use of ABO typing when more definitive tests are available amounts to a violation of due process of law which should be confronted by this court, but the only way the issue could be within the purview of Rule 37 is if it presents a question so fundamental as to render the judgment in petitioner’s case absolutely void. Griswold v. State, 290 Ark. 79, 716 S.W.2d 767 (1986); see also Madewell v. State, 290 Ark. 580, 720 S.W.2d 913 (1986). There is no requirement *388that there be any scientific evidence to sustain a conviction for rape, see Cope v. State, 292 Ark. 391, 730 S.W.2d 242 (1987), but even if serological testing forms a significant part of the state’s evidence, we are not persuaded by the arguments in this petition that a conviction is void for failure to perform any particular serological test.
The next portion of the petition concerns a publication contract between Larry Horton and petitioner and his wife. Petitioner alleges that Horton obtained the publication and literary rights to their story after petitioner was castrated in a widely-publicized attack. Petitioner asserts that the desire to have a inore interesting, dramatic story to tell led Horton to abandon all rational defense tactics. As the first of several examples of Horton’s disastrous strategy, petitioner states that Horton’s plan whereby he would put Walter Stevenson on the stand and try to prove he was the rapist “blew up” shortly before Stevenson was called to testify when it was discovered that Stevenson had the wrong blood type. Petitioner cites several references to Stevenson in cross-examination by the defense and the fact that he was mentioned in the list of witnesses for the defense in an attempt to show that counsel built the defense around Stevenson as the true perpetrator of the crime. Petitioner contends that when counsel did not call Stevenson, he was called instead by the state to rule Stevenson out as the attacker. In its closing argument the state noted that the defense had not called Stevenson. He alleges that the defense was left to lamely offer in closing argument that the evidence also pointed to Stevenson who was likely an innocent man.
After reviewing the record of the trial, we do not agree that the jury would have reasonably concluded that the defense was depending entirely on incriminating Stevenson to exonerate the petitioner. For one thing, Horton did not claim in his opening statement that he would produce the real rapist. While the state did note in its closing argument that Walter Stevenson had been listed as a witness for the defense and suggested as the attacker, McArthur responded in closing argument for the defense that the defense was not contending that Stevenson was guilty; the defense was merely pointing out that a strong case could be made against Stevenson who was probably innocent. The strategy was unsuccessful and may have been improvident, but petitioner has *389not shown that the tactical decision to suggest that Stevenson could have committed the crime amounted to ineffective assistance of counsel. We have consistently held that mere improvident strategy is not grounds to grant a Rule 37 petition. Leasure v. State, 254 Ark. 961, 497 S.W.2d 1 (1973).
Petitioner also cites counsels’ failure to call James Taylor as a witness for the defense as proof that counsel abandoned a rational defense when they pinned their hopes on exposing Stevenson as the attacker. According to petitioner, Taylor, who was subpoenaed by the defense but released without testifying, would have corroborated his claim that he was sick on the day of the crime and would have told the jury that he saw petitioner a couple of days later and petitioner said that the illness had left him weak. Since petitioner and several defense witnesses testified about the illness, Taylor’s testimony would have been cumulative. Omitting him as a witness therefore did not deprive the defense of any vital evidence.
Petitioner also faults counsel for not offering into evidence a taped telephone conversation between him and his former employer Richard Kellum. He says that the heated conversation was taped before his arrest and before he knew he was a suspect and contained a statement wherein he explained that he had tried to reach Kellum several times on the afternoon of the crime to tell him that he was too sick to return to work but Kellum’s answering machine was not turned on. Petitioner does not state under what circumstances the tape came to be made. To constitute ineffective assistance of counsel, the failure to offer the tape into evidence would have to have been an error so serious that there is a reasonable probability that, but for the error, the result of the proceeding would have been different. Strickland v. Washington, supra. The mere fact that the tape might have provided some corroboration for petitioner’s alibi does not create a reasonable probability that the result of his trial would have been different if the tape had been introduced into evidence.
Also with regard to Kellum, petitioner alleges that counsel should have cross-examined him for bias because Horton overheard the prosecutor “browbeating” him outside the courtroom trying to get him to shade his testimony to be more favorable to the state. He further states that he had a bitter unemployment *390compensation case against Kellum, a fact which would have shown Kellum to be biased.
Petitioner does not explain what the prosecutor said to Kellum which led him to conclude that Kellum was being browbeaten into shading his testimony, and he does not contend that any action by the prosecutor actually caused Kellum to be untruthful. There are a myriad of questions which an attorney could ask in cross-examining a state’s witness. It is not difficult in hind-sight to think of questions which could have been asked. Petitioner has not offered factual support from which it can be concluded that counsel was ineffective in not cross-examining Kellum about his dealings with the prosecutor or the compensation claim. It cannot be said that the failure to examine Kellum on the two points prejudiced the defense.
Another contention is that defense counsel failed to challenge the lineup or search and seizure and to cross-examine the victim about the lineup and photo arrays. He offers nothing, however, to show that any act or omission by counsel was actually motivated by Horton’s desire to make the publication contract more valuable.
Petitioner concludes his arguments about the publication contract with the allegation that the entering into the contract was unprofessional conduct which created a highly prejudicial conflict of interest. He asks that we adopt a per se rule that a conflict of interest exists where there is an agreement between defense counsel and an accused bestowing to the attorney publication or literary rights concerning the case. As petitioner concedes, courts usually hold that there is no per se rule regarding such contracts. Wojtowicz v. United States, 550 F.2d 786 (2nd Cir.), cert. denied, 431 U.S. 972 (1977); Ray v. Rose, 535 F.2d 966 (6th Cir.), cert. denied, 429 U.S. 1026 (1976). While recognizing the potential for a conflict of interest to develop in such cases and noting the language of Arkansas Rules of Professional Conduct, Rule 1.8(d), we will not presume prejudice. A petitioner who collaterally attacks his conviction under Rule 37 is required to show that he suffered some actual prejudice arising from a specific error by counsel. “Prejudice is presumed only if the defendant demonstrates that counsel ‘actively represented conflicting interests’ and that ‘an actual *391conflict of interest adversely affected his lawyer’s performance.’ ” Strickland v. Washington, supra, citing Cuyler v. Sullivan, 446 U.S. 335 (1980). As with any claim of ineffective assistance of counsel, the petitioner has the burden of providing factual support to demonstrate that counsel actively represented conflicting interests and that the conflict of interest adversely affected counsel’s performance. See Neff v. State, 287 Ark. 88, 696 S.W.2d 736 (1985); Smith v. State, 264 Ark. 329, 571 S.W.2d 591 (1978).
The petitioner also alleges that he was arrested without probable cause. He contends that either the warrant for his arrest was one that had been improperly presigned by the municipal judge without a determination of probable cause, see Stewart v. State, 289 Ark. 272, 711 S.W.2d 787 (1986), or that the judge issued it based on a statement by the victim which did not provide probable cause. His argument is that the illegal arrest and subsequent lineup tainted the victim’s in-court identification of him as her attacker and that his attorney was ineffective for not challenging the arrest. He claims that the in-court identification should have been suppressed. We find that even if there was no probable cause and the arrest was illegal, that illegality did not taint the in-court identification so that the petitioner was denied a fair trial.
The petitioner states that he was arrested on November 1, 1984, pursuant to a warrant based on affidavits of the victim and Officer Dye. He was then placed in a lineup. According to the record, a warrant was issued on November 28, 1984, accompanied by an affidavit signed by James Cole which reads:
I am employed by the Forrest City Police Dept. On September 11,1984, Wayne Dumond did go to the home of [victim] and did force her to leave the home with him at gun point and did take her to an unpopulated area in St. Francis County, Arkansas, and did force her to have sexual intercourse and oral sexual activities with him by the use of a deadly weapon. This information is based upon my investigation and the investigator of other officers of the Forrest City Police Department.
Officer Dooley testified at trial that the victim came to him after seeing the petitioner driving a truck and described him as *392her attacker. Based on the victim’s description of the man and the truck, Officer Dooley said that he arrested the petitioner. This testimony summarizes the facts surrounding the petitioner’s arrest as stated in Dooley’s affidavit for a search warrant for the petitioner’s truck and house. Although the facts of the arrest as recited by the petitioner might not provide probable cause, nor would the affidavit by James Cole which appears in the record, Officer Dooley’s testimony of his knowledge prior to the arrest indicates that he had probable cause for the arrest.
However, even if there were not probable cause for the arrest so that the arrest was illegal, the petitioner’s attorney was not ineffective for not challenging the arrest since we find that the in-court identification was not the product of the arrest. An illegal arrest, alone, is not grounds for Rule 37 relief. Gunn v. State, 291 Ark. 548, 726 S.W.2d 278 (1987). The issue is whether the arrest tainted the in-court identification of the petitioner by the victim so that the identification should have been excluded.
In Wong Sun v. United States, 371 U.S. 471 (1963), the United States Supreme Court stated that the test for determining whether evidence from an illegal arrest should be excluded is:
[W]hether, granting establishment of the primary illegality, the evidence to which instant objection is made has been come at by exploitation of the illegality or instead by means sufficiently distinguishable to be purged of the primary taint.
The question then is whether the lineup and then the in-court identification was an exploitation of the arrest or whether the in-court identification came from a source that is independent of the initial illegality. Silverthorne Lumber Co. v. United States, 251 U.S. 385 (1920). We are satisfied that the in-court identification stemmed from the victim’s extended contact with her assailant during the attack rather than from the lineup. The victim testified that she was with the petitioner at approximately 2:45 p.m. for forty to forty-five minutes. He was not disguised and she stated that she saw him clearly. She had to persuade him not to kill her while standing face to face with him. The victim was absolutely positive about her identification of him in the courtroom. All of these factors are indicative of the reliability of the victim’s identification. Under these circumstances we find that the in-*393court identification was not the product of the arrest or resulting lineup. Therefore, the petitioner’s attorney could not be said to be ineffective for failing to challenge the arrest.
Neither do we accept the argument that the petitioner’s attorney was ineffective for failing to challenge the lineup as suggestive. Because we find that the victim’s in-court identification of the petitioner was based on her observation of him at the scene of the crime, no prejudice to the petitioner as a result of the lineup has been demonstrated.
The petitioner also claims that his attorney should have challenged the search that the police made of his pickup truck. The search yielded a pistol. The petitioner submits arguments that the warrant might have been presigned, that the search warrant might have been issued after the search, that the information in the affidavit was stale, and that the affidavit did not provide probable cause. All of these arguments do not suffice to demonstrate that had the search been successfully challenged and the evidence suppressed, there would be a reasonable probability that the outcome of the trial would have been different. See Strickland v. Washington, supra. The only item recovered was a pistol similar to that which the victim said that her attacker used. It is very unlikely that if the evidence had been suppressed the outcome of the trial would have been different since the victim positively identified the petitioner and in light of the other strong evidence that the state introduced against him.
Petitioner next states that after he was convicted investigators assisting him learned that the victim had picked a picture of a man named Ricky White from a police photospread as being her assailant. He further states that the Forrest City Police Chief admitted that the misidentification was withheld. He further states that Horton was aware, presumably before trial, that the victim had allegedly identified White, but the Forrest City Police Department “stonewalled” and denied it. He says that McArthur felt helpless on the issue since he got into the case so late. In a similar allegation, he alleges that Henry Leary, a St. Francis Deputy Sheriff, told “former defense counsel” (again, presumably after trial) that he was present when the victim viewed the lineup containing the petitioner and that she was unable to *394identify him until police Chief Joe Goff took her into another room and brought her back for a second viewing. Petitioner charges that the state withheld information about her inability to immediately recognize him in the lineup but does not claim that the information was available to counsel before trial. In fact, petitioner does not claim that either Horton or McArthur rendered ineffective assistance of counsel in regard to either piece of information which he alleges was withheld. The information was uncovered after trial and clearly constitutes new evidence. As with the genetic typing evidence discussed previously, it amounts to a direct attack on the conviction and is not therefore cognizable under our postconviction rule. Chisum v. State, supra.
Petitioner’s final allegation is related to an argument he raised on appeal. During the testimony of Bill Dooley, the police officer who had lifted fingerprints from the victim’s car in which she and her abductor had ridden, Dooley referred to a report from the State Crime Lab. McArthur immediately asked to approach the bench, stating that the defense had not been made aware of the report’s existence until that moment. The prosecutor responded that there may have been no written report. When McArthur insisted that the defense was entitled to the report if it existed, the prosecutor requested and was granted a recess to go over the matter. No record was made of what occurred during the recess, and petitioner now alleges that a discussion was held in the judge’s chambers among the court and the lawyers in which defense counsel moved for a mistrial on the ground that the fingerprint report had not been made available to the defense on discovery. Petitioner states that Horton did not learn that the in camera discussion containing the motion for mistrial was not made a part of the record until the transcript was prepared for appeal. (Horton did not mention in the appellant’s brief that any motion had been made in the conference.) He argues that Horton, who alone represented him on appeal, was responsible for moving to supplement the record if a part of the trial was omitted from the transcript. He contends further that the motion for mistrial made in camera was of great significance since the failure to request a mistrial in a timely manner, i.e. at the first opportunity, lead to this court’s declining to consider the issue on appeal.
The sixth amendment guarantee of effective assistance of counsel extends to effective appellate as well as trial *395representation. Evitts v. Lucey, 469 U.S. 387 (1985). The court in Evitts did not set out specific criteria for assessing the effectiveness of an attorney on appeal, and we have not addressed the question beyond the general holding that the petitioner is responsible for making a clear showing in the petition that counsel failed to raise some possibly meritorious issue. Howard v. State, 291 Ark. 633, 727 S.W.2d 830 (1987). Some courts have held that the standard for ineffective assistance of counsel established in Strickland v. Washington, supra, can be used as a basis for establishing a standard for effective assistance of appellate counsel. Jenkins v. Coombe, 821 F.2d 158 (2nd Cir. 1987); Gray v. Greer, 800 F.2d 644 (7th Cir. 1986); Bowen v. Foltz, 763 F.2d 191 (6th Cir. 1985);Schwander v. Blackburn, 750 F.2d 494 (5th Cir. 1985); Mitchell v. Scully, 746 F.2d 951 (2nd Cir. 1984). We agree that the Strickland guidelines are applicable to a review of a claim of ineffective assistance of counsel on appeal.
The Strickland standard requires two findings. First, that counsel’s performance was deficient, and second, that the deficiency resulted in prejudice to the defense. Prejudice will be presumed to have resulted where counsel failed to raise on appeal an issue of such significance that counsel’s performance can be said to have fallen below an objective standard of reasonableness and there was a reasonable probability of a different result but for counsel’s error. Counsel’s conduct must have undermined the proper functioning of the adversarial process. Since there is room in appellate, as well as trial, advocacy for professional judgment, it may be necessary to examine the record to see what issues counsel could have raised as compared with those issues which were raised. The case before us presents a situation which is unusual in that counsel is not taken to task for not raising the issue of the failure of the state to disclose the fingerprint report but rather for not supplementing the record on appeal to show that he made a timely motion for mistrial when he first learned that the fingerprint report had not been disclosed.
We regret the necessity of again pointing to “off-the-record” conferences held in chambers or at the bench pertaining to substantive matters before the trial court. All conferences should be on the record unless they involve matters unrelated to the trial in progress, in which case a note to that effect should be made. Ward v. State, 293 Ark. 88, 733 S.W.2d *396728 (1987); Fountain v. State, 269 Ark. 454, 601 S.W.2d 862 (1980). There can be no doubt that the discussion about the fingerprint report concerned a substantive matter, and the state had a duty in accordance with Ark. R. Crim. P. 17.1 (a)(iv) to provide for the discovery of all reports or statements of experts made in connection with the case, including results of scientific tests or comparisons. Petitioner had made the appropriate request under A.R.Cr.P. Rule 17.1 and sought the basis of the results of any test. The prosecutor denied knowledge of the fingerprint report, but that information is imputed to him. Lewis v. State, 286 Ark. 372, 691 S.W.2d 864 (1985). As we said on appeal, while the information testified to by the expert was neutral and nonprejudicial, petitioner was entitled to challenge the state’s conclusion by having his own test performed. Although the state’s witness said that he could not determine from his examination of the fingerprints whether petitioner had been in the victim’s car, it was important for the effective preparation of the defense for the defense to study the state’s fingerprint report to prepare for the state’s evidence as well as to decide whether further investigation of the fingerprint evidence was necessary. We held when this case was affirmed on appeal that the defense had waived its objection about the fingerprint report when it made no further objections to the evidence until the state had rested. If an objection and request for mistrial were made during the off-the-record conference, counsel on appeal should have moved to supplement the record so that this court could have been aware of the timely objection and motion for mistrial in order to decide whether the trial court’s reasons for overruling the objection and denying the motion were proper. We will not presume that the accused suffered no prejudice in an off-the-record conference where the petitioner has made a factual allegation which, if true, may establish that he was prejudiced in his right to effective assistance of counsel on appeal. Since a complete record is necessary to determine whether there was a motion for mistrial and, if so, the exact context in which it was made and the reasons it was denied, we remand this cause to the trial court to supplement the record with the off-the-record conference. When the complete record is before us, we will decide whether the petitioner was denied effective assistance of counsel on appeal.
*397In conclusion, the petition for postconviction relief is without merit in all respects except for the allegation that petitioner was denied the effective assistance of counsel on appeal by virtue of counsel’s failure to secure a complete record and raise the denial of a timely motion for mistrial as grounds for reversal.
Remanded.