dissenting:
I respectfully dissent. While I have no quarrel with the majority opinion’s exposition of the facts, I believe that the majority misconceives the central issue involved here, which is whether the expert evidence regarding the adjacent statements recorded on the tape affects the conclusiveness which must be given to the trial court’s determinations of credibility. As I do not believe that this “new” evidence undermines the credibility findings of the state court or fundamentally alters the nature of Aiken’s claim, which was fully and fairly litigated before the state court, I would affirm the denial of the writ.
I believe that the issue of exhaustion is disposed of by the Supreme Court’s opinion in Vasquez v. Hillery, 474 U.S. 254, 106 S.Ct. 617, 88 L.Ed.2d 598 (1986). In Vasquez, the respondent had presented his equal protection challenge to the selection of the grand jury at every level in the state courts. 474 U.S. at 256, 106 S.Ct. at 620. In response to a request from the district court, the respondent submitted three items of evidence which formed the basis for the state’s nonexhaustion claim: census data from the year 1900, three affidavits from residents of the County, and a statistical analysis of the cumulative census data from 1900 to 1960.1 474 U.S. at 258-59, 106 S.Ct. at 620-21; see also Hillery v. Pulley, 533 F.Supp. 1189, 1199 (E.D.Cal.1982).
The district court in Hillery carefully addressed this issue and applied a standard similar to that applied here:
The test for “fair” presentation is whether the new material presents the petitioner’s action in “a significantly different posture” from that considered by the state courts.
533 F.Supp. at 1200. The court found that the census evidence and the affidavits were merely cumulative evidence relating to facts which were not in dispute. Regarding the statistical analysis, he said:
The testimony proffered by petitioner is simply an expert analysis of evidence already before the court and is not, in this sense, “new evidence.” In this regard it seems relatively clear that the statistical analysis of census data is the kind of information designed ... to aid the trier of fact in understanding the evidence before the court.
Id. at 1202.
On final hearing, the district court granted the writ. Hillery v. Pulley, 563 F.Supp. 1228 (E.D.Cal.1983). In affirming, both this court and the Supreme Court expressly approved the district court’s opinions. See Vasquez v. Hillery, 474 U.S. at 256 n. 1, 106 S.Ct. at 619 n. 1 (“thorough and well-reasoned”), aff'g Hillery v. Pulley, 733 F.2d 644 (9th Cir.1984) (“excellent and extensive”). The Supreme Court’s reasoning *885in rejecting the lack of exhaustion argument echoes that of the district court. 474 U.S. at 258-59, 106 S.Ct. at 620-21.
In determining whether the “new” evidence presented by the petitioner requires a remand to the state court under the Vasquez v. Hillery test, it is important to understand exactly how that evidence relates to the issue decided in the prior state court proceedings. It is my view that Dr. Shipp’s affidavit, like the expert analysis in Hil-lery, is “simply an expert analysis of evidence already before the court.” This case does not raise the issue of whether Aiken asked for counsel or what the substance of the colloquy was; the parties agree that the requests for counsel could have been heard by the officers and that the officers responded appropriately to statements immediately preceding the request. The only issue is whether Aiken’s request was deliberately ignored by officers who, in fact, heard the request. The issue is thus one of credibility, and unless Dr. Shipp’s affidavit can be said to undermine the trial court’s finding that the officers were truthful in saying they did not hear the request (and therefore did not deliberately ignore it), it does not “fundamentally alter” the claim which Aiken has already presented to the state courts.
When the original tape was played on the original recording device at the trial, neither the judge, the prosecution, nor the defense counsel heard the requests. It was only afterward, using a new machine, that it became apparent that Aiken had, in fact, requested counsel; even then his requests were difficult to discern.2 It was at that point that the supplemental hearing was ordered for the specific purpose of determining whether the officers had heard and deliberately ignored Aiken’s request. At that hearing, both officers testified that they did not hear any requests for counsel. They also testified that Aiken’s manner of speech made him difficult to understand. There was testimony that the interview room was not soundproof and that extraneous noises might have interfered with the officers’ ability to hear what Aiken was saying. The tape was played repeatedly on both the original recorder and a new machine. Aiken was able to cross-examine the witnesses and to bring out the fact that the officers had responded without difficulty to his other statements.
Dr. Shipp’s affidavit establishes that the decibel counts of Aiken’s requests for counsel were higher (louder) than were the immediately preceding statements to which the officers responded appropriately. This, however, does not undermine the credibility determination made by the state court.3 Federal habeas courts are explicitly enjoined from redetermining matters of credibility. Marshall v. Lonberger, 459 U.S. 422, 434-35, 103 S.Ct. 843, 851, 74 L.Ed.2d 646 (1983). The merits of this evidentiary dispute were resolved in a full, fair, and adequate state hearing. If Aiken had been prevented by the court from directing inquiry along the lines of the analysis later performed by Dr. Shipp, the integrity of the proceeding might reasonably be called into question. That was not the case, however, and under the circumstances which did occur, I am unable to say that Dr. Shipp’s evidence so changed the nature of Aiken’s claim that it requires yet another round of litigation before it can be said that Aiken received a full and fair hearing on the issue. Accordingly, I would affirm the denial of the writ.
. Census figures for 1910 through 1960 had been considered in the state court. 533 F.Supp. at 1199 n. 21.
. In fact, a complete transcript of the interrogation was not obtained until after Dr. Shipp analyzed the tape using a variety of physical and electronic sound-enhancement techniques.
. The decibel count is, in a sense, some scientific confirmation that the human ears could have heard the requests. However, it does not at all establish that the human ears did in fact hear and understand that a request was being made. The decibel count is significant, but of far more significance are the conditions of human auditory sensitivity, mental concentration and the receipt and translation of sound.