COX, J.

A person who is convicted of a crime is not entitled to a new trial solely because a key prosecution witness recants important trial testimony. If the recantation is not credible, then it is not material, and an essential factor that would support a new trial is missing. Here, Kea Ieng, who was convicted of first degree murder, fails to show that the trial court abused its discretion by deciding, under the totality of the circumstances, that the testimony of the recanting witness was not credible. Accordingly, we affirm.

In January 1991, Kea Ieng shot and killed Him Chhom just outside the Ieng family's apartment. The defense's theory at trial was that Ieng shot Chhom in self-defense. The evidentiary basis for this theory was primarily the trial testimony of Ieng. He testified that Chay Chap called the Ieng family before the shooting to warn Ieng that Chhom wanted to kill him. Ieng's sister, Borin, and his mother also testified that Chap had called to warn them about Chhom's threat. But a detective who interviewed Borin after the shooting indicated that she did not tell him anything about such a phone call.

Ieng also testified that just before the shooting, Borin told him that Chhom was at the apartment door. Ieng said that he took his shotgun with him to answer the door because he was afraid of Chhom's threats. When Ieng opened the door, Chhom tried to get in. According to Ieng, Chhom tried to grab the barrel of the shotgun, and Ieng fired once, hitting Chhom in the chest. Believing that Chhom was reaching for a weapon, Ieng fired a second time, again hitting Chhom in the chest. Chhom died from these wounds.

Chay Chap testified for the State. He testified that he dated Borin Ieng for several months before the shooting. Kea Ieng did not approve of the relationship. Chap testified that Ieng had threatened Chap and his friends, Him Chhom and Sarouen Bou, before the shooting. But Chap denied calling the Ieng family prior to the shooting to warn of any threat by Chhom against Kea Ieng.

Chap's friend, Sarouen Bou, also testified for the State. He said that Ieng had threatened him and questioned him about Chap's relationship with Borin. Bou saw Chhom on the day of the shooting and said he was acting normally. Bou denied that he, Chap, or Chhom were involved in a gang.

Forensic evidence showed that at the time of the shooting, Chhom had a gun in his right hip pocket. The safety on the gun was on, and there were no rounds in the chamber. Blood spatter evidence indicated that Chhom was shot while standing in the stairwell leading to the Ieng apartment. The blood evidence was inconsistent with the testimony of Ieng, who had testified that he shot Chhom while the victim was standing in front of the door to the apartment. Chhom's wounds were also consistent with the shooter being above him.

The jury found Ieng guilty of first degree murder. Ieng appealed. We affirmed his conviction in an unpublished decision.

Some three years after testifying at Ieng's trial, Chap recanted. Ieng then moved for relief from the judgment and sentence. At the hearing on the motion, Chap testified differently than he did at trial in several respects. After being informed by his lawyer of the consequences of being charged with perjury, he claimed that he lied at trial because he was afraid of Chhom's "group." The trial court found that Chap could not identify any member of the group and none were present at trial. The substance of Chap's recanted testimony was that he did in fact call to warn the Ieng family that Chhom had threatened to kill Kea Ieng. He also testified that he did not know if he ever personally heard Kea Ieng threaten Chhom. Chap testified that he had again been dating Borin Ieng in July 1994, when he recanted. But he was no longer dating her at the time of the hearing on his new testimony because of "domestic problems."

The trial court denied Ieng's motion. Ieng appeals.

[1-4] We review the trial court's decision of whether to vacate the judgment and grant a new trial for abuse of discretion. We will leave undisturbed findings of fact that are supported by substantial evidence in the record. If they are so supported, we then decide whether they support the conclusions of law and judgment. We review issues of law de novo. [5] Under CrR 7.8, the trial court may grant relief from a judgment based on "newly discovered evidence which by due diligence could not have been discovered in time to move for a new trial under rule 7.6 . . . " Recantation of trial testimony is generally treated as "newly discovered evidence."

In State v. Williams, the Supreme Court set forth five factors a defendant must establish in order to obtain a new trial based on newly discovered evidence:

a defendant must prove that the evidence: (1) will probably change the result of the trial; (2) was discovered after the trial; (3) could not have been discovered before trial by the exercise of due diligence; (4) is material; and (5) is not merely cumulative or impeaching. A new trial may be denied if any one of these factors is absent.

Citing its earlier decision in Rolax, the Supreme Court stated in Macon that Rolax stands for the proposition that the trial court's determination of the recanting witness's reliability, whether or not the original testimony of that witness was the sole basis for conviction, is controlling. The court then concluded by expressly overruling Powell. Thus, whether or not independent corroborating evidence exists to support the original testimony of the recanting witness is not a controlling factor.

Ieng first contends that the trial court abused its discretion by making its own subjective determination of Chap's credibility rather than determining whether the recantation would be persuasive to a reasonable juror. We disagree.

Ieng relies heavily on State v. Smith, a Division II case, to support his argument. There, the court stated that when a trial court determines whether recantation testimony is material, "[t]he question is not whether the trial court believes the recanting witness but whether the recantation has such indicia of reliability or credibility as to be persuasive to a reasonable juror if presented at a new trial." That court relied, in turn, on its earlier case, State v. York. There, the court held that it was inappropriate for a trial court to sit as a "thirteenth juror" to determine the credibility of a recanting witness.

In contrast to Smith, we held in Eder that "[t]he requirement that recantation testimony be material demands that the trial court in all cases decide whether, more probably than not, it is credible." Eder, like York and Smith, looked to Rolax to resolve the question of the trial court's proper role in assessing credibility. But Eder reached different conclusions about Rolax than the York and Smith courts.

In Eder, we reasoned that the purpose of the remand in Rolax was for the trial court to assess the credibility of the recantation, not to determine whether other independent corroborating evidence existed to support the conviction. Our rationale was that if the sole task of the trial court on remand was to determine whether independent evidence existed, the Supreme Court would not have instructed the trial court to determine on remand whether the recantation was perjured or not. Thus, it is the trial court, not the jury, that must assess credibility in all cases.

While the Supreme Court has not directly addressed the question presented here, its Macon decision controls the outcome of this case. In Macon, the court stated that Rolax stands for the proposition that a trial court does not abuse its discretion by determining the reliability of recantation testimony even when that is the sole basis for conviction. It pointed out that before considering a defendant's motion for a new trial based on a recantation, the trial court must determine whether the recantation is reliable. The court's interpretation of the earlier Rolax case and its express overruling of Powell support the conclusion that the Eder court's reading of Rolax is correct. We adhere to Eder and hold that the trial court is to make its own determination of the credibility of a recanting witness, whether or not there is corroborating evidence and without regard to whether a jury might find the witness credible.

Ieng next contends that a new trial is warranted because there was no evidence to corroborate Chap's original testimony that he did not call the Ieng family to warn of Chhom's threats. The trial court found that Bou's testimony corroborated Chap's story at trial. But the Supreme Court has rejected any requirement that corroborating evidence exist before a trial court has discretion to deny a motion for a new trial based on a recantation. Thus, the trial court's finding on this issue, though interesting, is not relevant. Ieng's argument fails.

Ieng next argues that the recantation testimony "nullified" Chap's trial testimony. He stresses that the State relied on its assertion at trial that Chap had not warned him of Chhom's threats to prove lack of self-defense and to undercut the testimony of the Ieng family that there was such a warning. The essence of this argument is that because Chap's trial testimony was important to the State's case, Ieng is entitled to a new trial. As we have noted above, the Macon analysis requires that the trial court make an initial assessment of credibility regardless of the importance of the recanted trial testimony.

Ieng further contends that because the credibility decision was "close," he is entitled to a new trial. For the most part, his argument suggests there is not substantial evidence to support the trial court's decision. We disagree.

Chap's assertion at the recantation hearing that Chhom had friends who had threatened him was inconsistent with Bou's trial testimony that the three only associated with each other and not with others. Chap had continued to date Borin, Ieng's sister, and wanted to resume dating her. Therefore, he had a motive to fabricate testimony at the recantation hearing favorable to the Ieng family. The trial court did not abuse its discretion by determining that Chap's recantation was not credible and denying Ieng's motion for a new trial.

Ieng argues that Macon and Eder are distinguishable from this case because they are child sexual abuse cases. While the facts of those cases are not identical to the facts here, Macon and Eder set forth general principles applicable to all cases involving motions for new trials and to vacate judgments based on newly discovered evidence. It is true that in Eder the court noted that it is likely in a child sexual abuse case that the child witness' family may pressure the child into recanting her or his testimony. But the trial court may find motives for recantation in other types of cases, such as this one. In fact, the trial court here noted that Chap's relationship with Borin at the time of recantation provided such a motive.

Ieng next argues that Chap's severe language problems at trial and inaudible responses throughout the trial transcript in contrast to the recantation testimony bolsters the latter testimony. Our previous unpublished decision considered and rejected the argument that the inaudible words and the lack of an interpreter for Chap at trial prejudiced Ieng. We determined that Chap's testimony was clear enough to permit review of the issues on appeal. Ieng fails to convincingly explain why we should question the trial court's credibility determination at the recantation hearing on the basis of these arguments that we rejected in our prior decision. We decline to do so.

Ieng finally argues that the trial court erred because it was unable to assess Chap's credibility based on his demeanor. As noted above, it is the job of the trial court to assess the credibility of a recanting witness. Demeanor is not necessarily controlling. It is the totality of the circumstances, not just demeanor, that controls.

We also reject Ieng's due process argument that the original conviction was based on Chap's perjured trial testimony. The trial court properly found that the recantation was not credible. Thus, there is no showing that the trial testimony was perjured.

Because the trial court properly found that Chap's recantation was not credible, we need not reach the other arguments that Ieng and his lawyer in this case advance.

We affirm the judgment and sentence.

BECKER and ELLINGTON, JJ., concur.

Review denied at 134 Wn.2d 1014 (1998).