(dissenting). The legally dispositive facts adduced at the pretrial suppression hearing were that, two hours after the alleged robbery and five miles distant from the place of its commission, the codefendants, already under arrest for drug possession, were displayed to the complainant by several police officers, side-by-side, handcuffed, and next to a silver-toned vehicle similar to that in which the perpetrators had, according to the complainant’s earlier report, arrived at and left the robbery scene. The complainant, having been advised in advance of the showup that the police “had a person with the characteristics [he] had given,” did not fail on his arrival at the showup to identify the detained individuals as the persons who had robbed him.
Even if it were reasonable to suppose that “none of the features of this showup rendered it more prejudicial than any other” (majority op at 403), which it is not (contrast e.g. People v Wells, 221 AD2d 281, 281 [1st Dept 1995], lv denied 87 NY2d 978 [1996] [where “the police never indicated to the victim that there was a suspect in custody and made it appear to the victim that defendant was not in custody, and the victim, from a distance, was asked only generally to scan the block to see if she saw anyone whom she recognized”]), that would only beg the question. The majority offers no legally satisfactory answer, why identifications obtained in pursuance of such a patently suggestive and gratuitous display should not have been suppressed as risking convictions premised on irreparable misidentification.*
Showup identifications involving an encounter between a suspect or, as here, ostensibly linked suspects and a witness are inherently suggestive and thus strongly disfavored (People v Riley, 70 NY2d 523, 529 [1987]; People v Ortiz, 90 NY2d 533, 537 *405[1997]; People v Johnson, 81 NY2d 828, 831 [1993]; and see Stovall v Denno, 388 US 293, 302 [1967] [“The practice of showing suspects singly to persons for the purpose of identification, and not as part of a lineup, has been widely condemned”]). We have, accordingly, found station house showups presumptively subject to suppression (see People v Duuvon, 77 NY2d 541, 544 [1991]). And, although we have been somewhat more lenient respecting the admissibility of showup identifications where there is a legitimate investigatory imperative or exigency to justify their use (see id.), we have been clear that this rationale is not so elastic as to permit geographically remote showups hours after the initial confrontation and after the suspect’s arrest. Indeed, although we may be said to have “endorsed” a closely defined category of showups in People v Johnson (81 NY2d at 831) (see majority op at 402), we were careful to qualify that “endorsement” by adding that “the emphasis must be upon the prompt and immediate nature of an identification after the crime has been committed, not, as the People argue, after the defendant has been arrested” (81 NY2d at 831 [emphasis supplied]).
Of course, whatever “endorsement” Johnson may have contained, it was not of the reviewed Appellate Division decision upholding the denial of Johnson’s motion seeking suppression of the complainant’s showup identification. In reversing that decision, necessarily upon the ground that it was not consonant with the legal standards governing the admissibility of showup identifications, we noted that “a showup hours after the crime, with both the complainant and the defendant transported to the crime scene, was improper. In those circumstances,” we said, “an appropriately conducted lineup was required” (id.). Apart from the circumstance that here, the complainant was transported not to, but away from, the crime scene—a circumstance surely not supportive of any conclusion that the showup was within the Duuvon dispensation for show-ups temporally and spatially near the crime—there is nothing to distinguish the present facts from those we found reversible in Johnson. In each case the showup followed the crime by hours and involved an arrested subject who could have, and should have, been placed in a lineup. If the mixed question doctrine was not an impediment to review in Johnson, neither can it be here. The police conduct was equally offensive to due process in both cases. The mixed question doctrine is abused when employed, as it is here, transparently, to shield from judicial scrutiny police conduct that is simply illegal.
*406The majority suggests that, even though the instant identification procedure occurred two hours and five miles distant from the crime, it may be accommodated within the Duuvon exception for temporally and spatially proximate showups (see Duuvon, 77 NY2d at 544) since the perpetrators left the crime scene in a car and because we have never adopted a separation of two hours as a bright line. But this suggestion takes no account of our case law implicitly holding that resort to a showup hours after a crime and subsequent to the defendant’s arrest is not compatible with due process (see Johnson, supra). Moreover, the notion that significant spatial separation between the crime scene and the showup may be treated as de minimis because perpetrators use a car, particularly when paired with the majority’s aversion to temporal limitations, raises the possibility of judicially countenanced showups not merely five miles, but 50 or 100 miles from the place of the crime, long after any legitimate investigatory need for the rapid confirmation or exclusion of an arrestee’s status as a criminal suspect by means of a showup must have abated. This is a development plainly not, even remotely, within the contemplation of Duuvon, which involved a showup conducted three to four minutes after the crime and around the corner from its place of commission. Even People v Brisco (99 NY2d 596 [2003]), until now this Court’s furthest temporal extension of the Duuvon proximity doctrine, approved a post-crime, pre-showup interval of no more than one hour. Today’s decision does not so much extend Duuvon as leave it in the dust, turning what was conceived of as a narrow exception limited by the requirements of due process into a rationale for a proliferating, and from this Court’s perspective largely unreviewable, reliance upon showups.
Indeed, the Court suggests that a showup is permissible whenever arrestees are suspected of non-“garden-variety” drug offenses significantly implicating public safety. The notion, however, that public safety is well served by routine resort to an inherently suggestive identification procedure, is puzzling. The use of a showup when the conduct of a properly constituted lineup is entirely feasible needlessly and substantially increases the risk of irreparable misidentification and, with it, the hazard that an innocent party will be convicted while the culprit remains at large. Of course, a prompt and immediate showup may be justified when, for example, the only alternative to its conduct is to let go a suspect detained without probable cause or when it is the only means of obtaining an identification from *407an ailing or moribund witness. Certainly, there are true exigencies in whose light a lineup would be impracticable and which would render the considerable risks entailed by showups constitutionally tolerable. But, ordinarily, there can be no cognizable exigency premised, as the majority now posits, on the possibility that a party already arrested may not have committed an offense other than the one for which he or she has been taken into custody. If there is a question as to whether an arrestee may be identified as the perpetrator of an offense—either the one for which he has been arrested or some other—due process dictates that the inquiry must be satisfied where at all feasible by means of a procedure that is not inherently suggestive. There will, presumably, be few situations in which an arrestee necessarily bound for the station house to be booked, as the present defendants were, could not be placed in a lineup. The People, in any event, made no showing that conducting an appropriate lineup would have been unduly burdensome (see Riley, 70 NY2d at 530).
Inasmuch as this highly suggestive showup was not a constitutionally permissible surrogate for a fairly constituted lineup identification procedure, I would reverse the appealed order, grant defendants’ omnibus motion to the extent of suppressing the showup identifications, and remand for a new trial, to be preceded by an independent source hearing (see Burts, 78 NY2d at 23).
Even if a reversal were not required in consequence of the illegal showup, I believe that it would be by reason of trial counsel’s inexplicable failure to raise the statutory affirmative defense to first-degree robbery.
While trial counsel vigorously litigated the question of whether the complainant had correctly identified defendants, realistically there was, following the denial of defendants’ suppression motion, little chance of an acquittal premised on misidentification. The discovery of the complainant’s stolen driver’s licence, credit cards and other identification in the vehicle occupied by defendants at the time of their apprehension was not plausibly explicable except by the hypothesis that those items had been placed there by someone involved with the robbery or the robbers; it was extremely powerful confirmation of the accuracy of the complainant’s identifications. On the other hand, there did exist an entirely viable defense to the indictment’s top *408count—one expressly provided for in the first-degree robbery statute under which defendants were charged and which, if successfully interposed, would, as the majority explains (majority op at 398), have significantly reduced defendants’ penal exposure.
As is now conceded, the BB gun allegedly brandished by defendant Howard during the robbery is not a “firearm” within the meaning of Penal Law § 160.15 (4), but rather an “imitation pistol” (see e.g. People v Wilson, 283 AD2d 339, 340 [1st Dept 2001]). Penal Law § 160.15 (4) specifically provides that “it is an affirmative defense [to first-degree robbery predicated upon the display of a firearm] that such pistol, revolver, rifle, shotgun, machine gun or other firearm was not a loaded weapon from which a shot, readily capable of producing death or other serious physical injury, could be discharged” (emphasis supplied).
Even if there had been a viable alternative factual theory to the BB gun display to support the indictment’s first-degree robbery count, it is clear that that theory was far from ironclad. We have held that, to be a sufficient predicate for a first-degree robbery conviction, a firearm display must be reasonably perceived as such by the victim (People v Baskerville, 60 NY2d 374, 381 [1983]). Here, the complainant said that he could not tell whether the object held to his back was a gun or something else. While perhaps the inference that the object appeared to the complainant to be a gun remained available to the jury nonetheless, it was hardly compelling. The salient point is that, even if, as the Appellate Division held in the alternative, there was sufficient evidence to support the first-degree robbery conviction based on the back poke attributed to defendant Stanley (92 AD3d 176, 179 [2012]), there remains a reasonable probability that defendants would not have been convicted of first-degree robbery if defense counsel had only interposed the defense that literally stared out at them from the statute under which their clients were charged in the indictment’s top count.
Moreover, the Appellate Division’s evident assumption that the verdict was in fact alternatively premised on Stanley’s “display” of the object poked in Lopez’s back (see id.)—an assumption now evidently shared by the majority—was completely unwarranted. There is no ground to suppose that the jury’s verdict was so founded. So far as the jury knew, or could have known given the trial court’s charge, which made no mention of the section 160.15 (4) affirmative defense, the factual theory nearly exclusively stressed at trial—that the BB gun was displayed—remained entirely viable. The overwhelming likelihood *409is that, if there was a unanimous verdict at all as to the display element, it was premised on the brandishing of the imitation gun—the factual theory that the affirmative defense would have nixed—and not the object stuck in Lopez’s back. But, to the extent that it may not have been, that raises more problems than it solves for the defense of this representational effort mounted by the majority. This is because the viability of an alternative factual theory upon which the jury could have premised their verdict casts into question whether the general verdict was unanimous. In the absence of any assertion of the affirmative defense, there was evidence to support a verdict on both the BB gun display theory and the back poke theory—or so the majority says. It is thus impossible to avoid the conclusion that the general verdict was duplicitous (see People v Martinez, 83 NY2d 26, 37 [1993]), and the majority conspicuously makes no effort to do so.
Contrary to the Appellate Division’s suggestion, now embraced by the majority, there was no conceivable strategic rationale for counsel’s failure to interpose the defense. At the very least, raising the defense would have succeeded in limiting the factual theory upon which the first-degree conviction was put to the jury, and in so doing avoided the possibility, which now cannot be ruled out, that the jury’s general verdict was not unanimous as to the factual ground for convicting on the top count. Moreover, moving for dismissal of the top count based on the statutory affirmative defense, as made out by the People’s proof, would not in any way have impaired the trial strategy of casting doubt upon the victim’s inculpating identifications. And, while the People defend counsel’s purported election to present an all or nothing defense, or, as the majority puts it, to “go for broke,” going for broke was never an option. It was not as if eliminating the top count or reducing the likelihood of a conviction on that count would have operated to make a not guilty verdict more likely; convicting for second-degree robbery (the second submitted count) was an option the jury would have had in any event. Of course, interposition of the defense would predictably have led to the submission of robbery in the second degree on a display theory, but a conviction of robbery in the second degree would undoubtedly have been a significantly better outcome for defendants than one for robbery in the first degree.
The inclusion of the affirmative defense in Penal Law § 160.15 (4) is all that distinguishes first-degree robbery based upon the display of an apparent firearm from second-degree robbery *410based on the display of an apparent firearm (Penal Law § 160.10 [2] [b]). The legislature then has, for all practical purposes, left it, in the end, to defense counsel to assure that a client is not prosecuted excessively for a robbery involving the display of an imitation gun. The reasonable professional expectation reflected in the statute must be that counsel will interpose the defense when there is a basis to do so, thus giving effect to the statutory scheme’s contemplated gradation in penal treatment. Counsel’s failure to raise the defense was in the present legal and factual context a serious and inexplicable departure from prevailing, and indeed legislatively assumed, standards of professional practice. And, it is at least reasonably probable that had the defense been raised the outcome of the trial would have been significantly less adverse to defendants (see Strickland v Washington, 466 US 668, 694 [1984]). Indeed, raising the affirmative defense to first-degree robbery, which, with respect to the BB gun display, was entirely made out by the People’s proof, was likely the single most obvious and efficacious act of advocacy counsel could have performed for clients whose prospects of outright acquittal by the time of trial were, dispassionately viewed, exceedingly dim. It is true that what is fundamentally at issue is one error, and the proper focus in evaluating a claim of ineffective representation is upon the total representational effort. Nonetheless, we have recognized that a single clear-cut error may be profoundly prejudicial and thus itself a sufficient ground for an ineffective assistance claim (People v Turner, 5 NY3d 476 [2005]). Here, the failure of counsel to interpose the obvious and plainly meritorious affirmative defense more than likely resulted in an insufficiently supported and/or non-unanimous verdict. On this account as well, then, there should be a reversal and a new trial.
Although at trial there was evidence from which the jury could have gathered that the complainant had in fact reliably identified defendants, there was no pretrial finding of independent source to support the admission of the complainant’s identification testimony in the first instance, and that omission, we have held, mandates reversal: “The flaw [where identifications have been received at trial following an impermissibly suggestive identification procedure] cannot be retroactively cured because, simply put, the jury heard impermissible in-court identification evidence and the nature of this kind of defect cannot be sanitized after the irretrievable event has occurred. In such circumstances, nothing short of reversal and a new Wade hearing and new trial will suffice” (People v Burts, 78 NY2d 20, 23 [1991]).