MURNAGHAN, Circuit Judge:
Several newspapers, feeling themselves aggrieved by action of the district judge in United States of America v. Virgil Griffin, in the United States District Court for the Middle District of North Carolina, Criminal No. 83-5301 G and eight connected cases, 83-5302 G through 09 G, have petitioned, in Fourth Circuit Case No. 84-1039, for issuance of a writ of mandamus to annul an order dated January 4, 1984, filed with the district court clerk on January 5, 1984 (hereafter the January 5, 1984 order). The order, by reference to an accompanying memorandum, provided for the conduct of individual voir dire “in camera with defendants present but other potential jurors not present.” The in camera portion of the order manifestly closed the individual voir dire proceedings to the public, including members of the news media.
Alternatively, the petitioners have attempted to appeal in the nine criminal cases, Criminal Nos. 83-5301 G through 09 G. In light of the fact that the appeal has been taken only to guard against a jurisdictional challenge to proceeding by way of [1321]*1321mandamus, we do not decide whether an appeal has procedurally been perfected, and give it no further attention inasmuch as we are satisfied that, as a route for reviewing the questions presented, the petition for writ of mandamus was proper. Central South Carolina Chapter of Professional Journalists v. United States District Court for the District of South Carolina, 551 F.2d 559 (4th Cir.1977); 556 F.2d 706 (4th Cir.1977), cert. denied, 434 U.S. 1022, 98 S.Ct. 749, 54 L.Ed.2d 771 (1978).
The portion of the district judge’s order of January 5, 1984 closing the voir dire to other potential jurors, has not been questioned by any party. That portion required that voir dire held for the purpose of ascertaining whether individual jurors should be dismissed for cause should proceed on the basis of presence, insofar as the potential jurors were concerned, being restricted solely to the potential juror whose status was, at the time, being questioned.1 The reasons underlying that order included concern that the presence of other potential jurors might inhibit the one whose qualification to serve as a juror was being ascertained from being as frank and forthcoming as would otherwise be the case if he or she alone, among potential jurors, were present.
The consolidated criminal case involves charges under 18 U.S.C. § 245 that the defendants, in violation of federally protected activities, had violated or conspired to violate rights of individuals shot down in a wild fracas on November 3, 1979 involving Ku Klux Klan and National Socialist Party of America adherents on the one hand and participants in an anti-Klan parade and other persons on the other. In 1980 a number of the defendants here had been acquitted in state court of murder charges emanating from the most regrettable event. Publicity in North Carolina has been widespread, both with respect to the state charges and trial in North Carolina and in the case of the currently pending federal charges.
The portion of the order, entered January 5,1984, calling for in camera conduct of the voir dire is the focal point of the petitioners’ complaint. However, closures to a) potential jurors and b) the public generally, including the news media, are not unrelated. The second manifestly will significantly aid in preventing a breakdown in accomplishment of the objectives of the first. The district judge, in order to minimize any adverse consequences of the in camera requirement, further spelled out that the voir dire proceedings should be recorded, transcribed, and made fully available to the public, the news media included, upon transcription.2
In the afternoon on January 9, 1984, the petitioners moved for relief from the in camera exclusionary order,3 whereupon the district judge scheduled a hearing to take place on January 11,1984 to afford them an opportunity to state their objections. The [1322]*1322district judge entered on January 12, 1984 an order denying the requested relief.
By the time the petitioners’ January 9, 1984 motion came to the attention of the district judge, two days of in camera voir dire had taken place; a third occurred on January 12, 1984. The district judge, having commenced with in camera voir dire, had identified, by January 12, 1984, 48 prospective jurors not disqualified for cause.4 In the same time frame 37 had been excused for cause.
Both the prosecutors and counsel for all the defendants have favored the entry of the order the propriety of which is contested in the instant proceeding. Both have expressed themselves as satisfied that progress to date in the conduct of the voir dire has been unexceptional and fair to them. At least, any objections they might wish to make are unassociated with the order excluding the public, the news media included, from attendance at voir dire.
On Thursday, January 12,1984, a panel of the Fourth Circuit Court of Appeals, appreciating the serious and imperative nature of the questions presented, entered a stay to allow resolution of the question before completion of the voir dire ended any chance for petitioners to attend even a portion thereof.
On Monday, January 16, 1984, having received written presentations from the petitioners, from the United States as prosecutor in the nine criminal cases and from the criminal defendants, we heard oral argument. On the same day, following argument, we entered an order denying the petition for a writ of mandamus, and, if appeal properly lay from the district court’s closure order, affirming that order. We promised to supply, as quickly as may be, an opinion setting forth the reasons for the decisions we took. What follows is our attempt to redeem that promise.
We start with the premise that claims for First Amendment protection receive especially sympathetic attention and that curtailment of rights of free access to information is the narrow exception to a sweeping general rule. At the same time, absolutism and rigid dogmatic application of concepts, regardless of consequence, are antagonistic to our judicial system. After balancing all the applicable considerations, of principle and of pertinent fact, we are persuaded that, with an unavoidable minimum of interference with freedom of access to information, the district judge has avoided a very real risk of substantial diminution in the rights of the defendants to a trial by an impartial jury guaranteed by the Sixth Amendment to the Federal Constitution. Certainly, on such a sensitive matter, his determination that a risk to that precious right exists is significant and is neither clearly erroneous nor an abuse of discretion.
It must be borne in mind, at the outset, that the rights of the news media, as freely acknowledged at oral argument, are co-extensive with and do not exceed those rights of members of the public in general. Not infrequently, the purveyors of news stand in as surrogates for the other entities and persons comprising the public, but the scope of their rights is defined by the scope of the rights adhering to those for whom they purport to act. Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 573, 100 S.Ct. 2814, 2825, 65 L.Ed.2d 973 (1980); Nixon v. Warner Communications, Inc., 435 U.S. 589, 609, 98 S.Ct. 1306, 1318, 55 L.Ed.2d 570 (1978).
Then it too must be acknowledged that, the district judge having identified and established the existence of a tension between First Amendment concerns of the petitioners and the Sixth Amendment fair trial guarantees of the defendants, he had to investigate whether a less disruptive alternative existed to achieve the protection of Sixth Amendment rights here endangered. The petitioners have, however, nar[1323]*1323rowed the scope of that inquiry for they have insisted on but one solution, putting forth no alternative for consideration.5 They seek to insist upon identifying free and immediate access to the voir dire as the absolute minimum consonant with their perceptions of what their First Amendment rights are. That would, however, fail to alleviate several serious problems. Free access, of course, leads inevitably to widespread dissemination. Despite the best and most well intentioned cooperation of court, parties and prospective jurors, and the strongest exhortations from the district judge to abstain from perusal of newspapers, magazines and the like, or viewing or hearing television or radio broadcasts, when dealing in numbers in the neighborhood of 750 prospective jurors, the likelihood of complete success is small. Just as obviously, segregation of 750 prospective jurors at the outset of the voir dire process, while theoretically, perhaps, possible, would impose an intolerable strain. Not only would location of physical facilities be severely taxing, but, furthermore, demands to monitor and supervise them would create debilitating, indeed exhausting, burdens in terms of operation of the court system as a whole.6
It should be remembered that the closure order was subordinate to, and only in implementation of, the order excluding all other potential jurors from the voir dire examination of any one of them. Such practice is not unusual, nor viewed with suspicion. It is recommended by the Judicial Conference of the United States. See Revised Free Press-Fair Trial Guidelines of the Judicial Conference of the United States, 87 F.R.D. 525, 532-33 (1980). It is endorsed by the American Bar Association. ABA Standards for Criminal Justice, Fair Trial and Free Press, Standard 8-3.5 “Selecting the Jury.” Concern does not appear to have surfaced for perceived First Amendment rights of the excluded potential jurors, who also may lay claim to membership rights of the public. The order under attack has the same ends, and indeed, exists only to insure the same result, as another order generally accepted as appropriate, and, indeed, in some cases, altogether necessary.
Hence, the insistence on all-or-nothing relief professed by the petitioners does not commend itself to us as a reason to choose “all”. Its sole basis is the perceived diminution in the worth of news, if not stale, at least of diminished freshness. That, it is thought, may, in some not quantified way, reduce the beneficial effect for the body politic of steady, immediate promulgation of the public business carried on in court.7
[1324]*1324Yet, as customarily in the law, such a general fear must be measured by the extent of the risk. In light of the success at the district court level in already identifying in three days, through voir dire, almost 70% of the potential jurors presumptively not subject to striking for cause, it is to be expected that in another week, or less, the order of which complaint has been made, and its delaying effect on the promulgation of news, should have expired.
It should be noted, that the petitioners evinced no interest in one suggested solution designed to lessen any loss in freshness or timbre of the testimony at voir dire. They were asked at the oral argument whether, to avoid the possibility of any loss through transcription and through having to rely on cold mechanical reproduction, lacking in intonation, for the words of court, potential jurors, and possibly of counsel for the prosecution and counsel for the defendants, they had sought or would seek an alternative order which would permit newsmen to attend voir dire sessions. The suggested order would require the newspapermen to be segregated, and not allowed to communicate with their employers or others until the moment for release of tapes of the proceedings was at hand. Their response was in the negative. Their lack of interest was palpable.
On balance, we are satisfied that the substantive contents of the statements made at voir dire will be in all essential respects as newsworthy if disseminated to the public on Monday, January 23, 1984 as on Monday, January 16, 1984. The likelihood of candor and consequent enhanced truth will be increased. The threat of theoretical diminution in the impact of a delayed description of the performance of actors, in a fleeting scene in the court drama preliminary to the main stage setting, fades in importance when contrasted with the highly likely slackening in the assurance of the best possible jury if unrestricted access by the petitioners to the voir dire before it is completed and the jury and alternates seated is permitted. Voir dire, as contrasted to the trial, is of limited scope, finite and brief. It must be kept in the forefront of our minds that two weeks of delay, but none whatsoever in denial of full access, are what are involved.
We commend Judge Thomas A. Flannery for a sensitive and sensible solution to an almost intractable problem, given the proper preference of First Amendment rights in almost every situation. “Almost” is the key word, however, and Judge Flannery has properly identified one of the few acceptable exceptions to a virtually unanimous rule. Indeed, in the final analysis, Judge Flannery in his Memorandum Order of January 12, 1984 denying the substantive relief which had been requested by petitioners succinctly but thoroughly outlined the applicable legal authorities, choosing, in any case of doubt, the route most favorable to the petitioners and made the requisite findings to support his January 5, 1984 order. We cannot improve on what he has written, and so simply excerpt the principal portion of his Memorandum Order:
Neither of the opinions rejecting closure orders control the instant case. Unlike the Court in Nebraska Press,8 this Court has not banned the press from reporting any aspect of the case. Indeed, this very point was emphasized in this Court’s Order of January 5, 1984. Order at 10. Similarly, Richmond Newspapers9 is not directly applicable to the case at bar. In Richmond Newspapers, the Court dealt with the trial, as opposed to the voir dire proceedings. Even if voir dire proceedings are considered part of the trial, see United States v. Brooklier, 685 F.2d 1162, 1167 (9th Cir.1982), the facts in the present case meet the criteria for a closure order established by Richmond Newspapers.
[1325]*1325Before turning to the factors of this case, this Court must pause to consider the standard for a closure order. The Supreme Court has not established a practical standard in this area. As noted above, Gan-nett10 requires a somewhat nebulous balancing test, and Richmond Newspapers required an “overriding interest” to exclude the public. The most stringent test was suggested by Justice Blackmun in dissent in Gannett. 443 U.S. at 440-42 [99 S.Ct. at 2936-37] .... Under that test a court must find the following to justify a closure order:
1. A substantial probability that irreparable damage to defendants’ fair-trial right will result from conducting the proceeding in public;
2. A substantial probability that alternatives to closure will not protect adequately his right to fair trial;
3. A substantial probability that closure will be effective in protecting against the perceived harm.
Id. This Court is far from convinced that this stringent test is required to justify a closure order applied to jury selection proceedings.3 It is aware, however, that the safest course is to apply the strictest test adopted by the Courts that have considered the matter. See Brooklier, supra, 685 F.2d at 1167. The test enunciated by Justice Blackmun, therefore, will be applied to the present case.
This Court need not rehash the intense and continuing public interest in this trial. The magnitude of publicity that this case has generated may be unprecedented in this area. A careful and exhaustive voir dire process is crucial to ensure that these defendants will be tried by an impartial jury which will render its verdict on the basis of evidence adduced at trial rather than information from the newspapers or television.4 In order to accomplish this task the Court felt that it was necessary to question individual jurors outside of the presence of other potential jurors. This procedure has a long history of approval by the Courts. See e.g., United States v. Bryant, 471 F.2d 1040 (D.C.Cir.1972); United States v. Cola-bella, 448 F.2d 1299 (2d Cir.1971). This rationale for the procedure was vividly described by the District of Columbia Circuit Court, when it stated that:
It is too much to expect of human nature that a juror would volunteer, in open court, before his fellow jurors that he would be influenced in his verdict by a newspaper story of the trial. Not only so, but had one or more of them said they would be so influenced, and especially if they had explained why, the damage to the defendant would have been spread to the listening other jurors.
Coppedge v. United States, 272 F.2d 504, 508 (1959). That same damage would occur if the press was allowed to observe and report the questions and responses. Potential jurors reading these reports would then be irreparably tainted. Responses would be less trustworthy if jurors felt that the public would look askance at someone too easily influenced by the press. No amount of questioning could remedy this taint. Furthermore, the problem of prejudicial publicity would be multiplied by every response of a potential juror reported in the press. The nature of this case adds an additional problem to the publication of potential jurors’ names, county of residence, and responses to questions. This case involves various organizations that occasionally evoke strong opinions in people. It may be that some people believe that these organizations are prone to violence. If some potential jurors believed these propositions they might be reluctant to express any strong opinions about these organizations, knowing that their responses would be reported by the media. These potential jurors might fear that honest responses would subject them to retribution. Once again, no amount of questioning could remedy responses motivated by fear.
This Court is convinced, therefore, that the first prong of this test is satisfied. The second prong of this test requires the Court to evaluate the effectiveness of alternatives [1326]*1326to the closure order that are less restrictive. At this point it should be noted that the Court has attempted to conduct these proceedings as openly as possible, and still protect the impartiality and integrity of the jury. The group voir dire as well as the statement of the case is open to the press and public. Furthermore, the press and public will be allowed to observe and report the actual jury selection, after the individual voir dire has concluded, as well as the entire trial. In addition, the restriction is limited in duration. The voir dire process is being recorded, and thus, a record of the proceedings is being maintained. This record will eventually be transcribed and at that time will be available to the public and press. Cf. Gannett, supra.
At oral argument, movants did not suggest any alternative measures for the Court to adopt short of its closure order. The number of potential jurors in this case renders most alternatives impractical, if not impossible.5 Sequestration of this extraordinary number of potential jurors is simply not feasible. Furthermore, a limitation on what the press would be allowed to report might remedy the intimidation factor, but it would not prevent the exponential multiplication of prejudicial publicity. This Court believes, therefore, that no alternative to closure would adequately protect the defendants’ right to a fair trial.
Finally, there is no argument that the closure order will fail to protect against the perceived harms. The elimination of publicity concerning the individual voir dire will assure that potential jurors will be insulated from the questions and other jurors’ answers for as long as possible. Furthermore, the potential jurors will be more candid in their responses if they do not have to worry about what the public’s opinion of those responses might be.
Finally we make explicit what was implicit in our order of January 16,1984. The stay, entered by us on January 12, 1984, was vacated, once the order of January 16, 1984, appeared.
Supplemental Opinion
The occasion is not one for surprises, so we state immediately at the outset that we remain unaltered in our views of the law expressed in the opinion released by the Clerk on January 19,1984. In other words, the decision in the Press-Enterprise Go. case,1 and the developments, following the seating of the jurors and alternates, respecting publication, or rather delay in the making available, of the voir dire proceedings, do not shake our composite conviction as to the correctness of what we held in that opinion.
Nevertheless, the extraordinary nature of the way in which the case evolved makes it [1327]*1327appropriate to outline the developments incidental to its origins. In that way we hope to dispel misunderstanding as to how and as to why, especially in aspects of timing, things unfolded as they did.
First it should be borne in mind that the case was one both of extraordinary interest and (not necessarily and same thing) of large importance. The district judge, having decided, for reasons appearing sufficient to him, that void dire as to whether potential jurors were susceptible to striking for cause, should be in camera, had, by January 12, 1984, proceeded along those lines for 3 days, carrying out over half the task. On January 12, 1984, our stay pending consideration of a mandamus application or hearing of an appeal was entered, bringing to an abrupt pause the conduct of a trial whose progress and completion were of great interest and significance to those directly concerned: a) the accused and b) the public (including 1) relatives and associates of the victims and 2) the news media as eyes and ears of us all).
We moved with dispatch to hear on January 16, 1984, the mandamus application or, alternatively, the appeal, and, if we can be said to have been in haste, it was for that reason and that reason only. We were engaged in no race to publication against the Supreme Court, or, indeed, any other court. It was truly an occasion to give the lie to the animadversions on the law’s delay made famous in Hamlet’s sixth soliloquy.
On Monday, January 16, 1984, we met in Charlotte, North Carolina, to hear argument. We were relieved to find ourselves ably assisted by well-prepared counsel appearing for all parties (the prosecution, the defense, and the media). Having canvassed the relevant considerations as thoroughly as could be, and finding the situation to be one presenting readily identifiable issues as to which decision, not time-consuming study and tortured reflection, were called for, we, on that same day, January 16,1984, reached and announced our decision to uphold Judge Flannery’s order, dismissing the attack upon it by way of mandamus, and, in the unlikely event that the attempted appeal was not premature, affirming his decision.
Thereafter, by mid-afternoon on January 17, 1984, a proposed opinion had been prepared, discussed and adopted by the three members of the panel. Having clarified with the Chief Judge of the Fourth Circuit Court of Appeals the propriety of doing so, Murnaghan, J., as the author, forthwith dispatched by mail to the Clerk in Richmond, Virginia, the opinion for filing. On that same afternoon of January 17, 1984, he notified a member of the Clerk’s staff, and instructed him immediately to inform all counsel in the case that an opinion had been prepared and mailed to the Clerk and would be available to each of them upon receipt in the Clerk’s office. Such notification was made, we are informed, on that same day, January 17, 1984.
Thereupon arose the potentially confusing chain of events. We were, of course, when we heard argument in Charlotte, North Carolina on January 16, 1984, and when we shortly thereafter took decision, aware that the Press-Enterprise Co. case had been granted certiorari and been argued, and was awaiting decision. We had no inkling, however, that decision would come so soon. In fact, the opinions in it were handed down on January 18, 1984. The fact that the publication date of our opinion was Januarey 19, 19842 has led to ummerited suggestions, by counsel who knew better, that we were somewhat remiss in not at least commenting on and attempting to distinguish Press-Enterprise Co. Our opinion, however, was completed and launched by mail from the judge’s chambers to the Clerk’s office for filing before Press-Enterprise Co. at the Supreme Court level had seen the light of day. Counsel for the media had been so informed on January 17, 1984.
We did not, and today do not, regard Press-Enterprise Co. as particularly close or [1328]*1328relevant to the Greensboro News Co. case. The certiorari note in United States Law Week clearly indicated that Press-Enterprise Co. concerned a general rule allowing the closing, for reasons of privacy rights protection, of voir dire in any, if not every, ease.3 We, on the other hand, were concerned with a carefully tailored solution for the particular problems of the particular case, and we were permitting no broader an incursion into free press and free speech rights than absolutely necessary to protect competing and overwhelming Sixth Amendment rights arising in that one case.4 The facts justified the limited, narrowly tailored closure in Greensboro News Co. “essential to preserve higher values.”5
Also, Judge Flannery, in the district court, had called for preservation, through transcription, and prompt release, of the voir dire proceedings, while the California court in Press-Enterprise Co. would have denied access for all time.
Indeed, perhaps the principal significance of Press-Enterprise Co. lies in its unambiguous statement that the voir dire is part of the trial itself, so that public access may not be denied on the grounds that it is but a preliminary and, as such, somehow less important stage in the proceedings. We, following Judge Flannery’s lead, had proceeded on the assumption most favorable to the media, namely, that the voir dire took on all the attributes of the trial itself. Furthermore, the voir dire in Press-Enterprise Co. consumed six weeks, while in Greensboro News Co. the process required less than two weeks.
Accordingly, we were unjustly charged with failure to give due deference to prior Supreme Court authority. Though publication of our opinion took place one day later than that of Press-Enterprise Co., it is pel-lucidly clear that our promulgation through dispatch to the Fourth Circuit Clerk took place prior to publication of the Press-Enterprise Co. opinions. As a trivial point, we might mention that, on January 17, 1984, a member of the Fourth Circuit Clerk’s staff notified the office of the Clerk of the Supreme Court of the fact that there had come into existence our opinion in the Greensboro News Co. case. He indicated that provision of a copy to the Supreme Court would be expedited. He was informed merely that the opinion in Press-Enterprise Co. had not yet issued. The office of the Clerk of the Supreme Court supplied no indication that publication was imminent.
Finally, the parties constituting the media base their efforts to secure reconsidera[1329]*1329tion or rehearing on supposed delays and other defects in the making of transcriptions available, now- that voir dire has been completed and the jurors and alternates seated. Without in any way indicating whether they have valid grounds, we point out that what has occurred, or failed to occur, subsequent to January 16, 1984, the date of our order affirming the action taken by Judge Flannery on January 5,1984, is not relevant to the correctness of our opinion filed January 19, 1984. To raise the questions the news media parties seek to present, they will have to institute a new proceeding by way of mandamus or appeal, or otherwise.
Accordingly, the petition for rehearing and the motion for reconsideration are denied. The suggestion for a poll on rehearing en banc has failed to receive the affirmative vote of the members of the Court in regular active service. The motion for leave to file additional affidavits and documents in support of the petition for rehearing is granted.