Ceasar Sanchez appeals from a judgment of the United States District Court for the Eastern District of New York entered on February 2, 1979, convicting him after a jury trial before Henry Bramwell, Judge, of distribution of heroin on August 23,1977, in violation of 21 U.S.C. § 841(a)(1) and 18 U.S.C. § 2. /The only issue on appeal is whether the4n-court identification of SancKiz 'By~añ üridereover~^eñt~shouid D'aveT T)ieir^cludedZfcom-..e-v-idence-»because-nf impermissibly suggestive prior photographic identifications^ Although the photographic Identificatioirprocedure used by the Drug Enforcement Administration (DEA) Task Force must be condemned as impermissibly suggestive, the district court did not err in finding that füñder the totality of the circumstances tühis improprietylfid not leacTto áT^übstantiaLlikeíihood ofñnadgHtÍfica±ÍOTi|r Accordingly, we affirm the conviction.
Assuming that the admission of the agent’s in-court identification testimony was proper, the evidence of appellant’s guilt was clearly sufficient. The testimony at trial, viewed in the light most favorable to the Government, established that at approximately 5 p. m. on August 23, 1977, Frank Marrero, a detective with the New York City Police Department assigned to the DEA Task Force, went to the El Obrero Market on Central Avenue in Brooklyn in an undercover capacity to purchase heroin. He was accompanied by one David Vega, a confidential informant. The informant introduced Marrero to one Luis Sanchez (hereinafter referred to as “Luis”), appellant’s uncle. Luis told Marrero and the informant to wait in the store while he went to get the “stuff.” Luis then left the store, walked across the street, and entered 142 Suydam Street; he returned soon thereafter and told Marrero he was “ready to do business.” Marrero and the informant were instructed to go around the corner and enter the first building on Suydam Street, which was next to the El Obrero Market, with a door from the Market entering on to a hallway of the building.
When Marrero and the informant got to the building, Luis was already inside and let them into the hallway. Luis then opened the door leading from the Market, and a person later identified at trial by Marrero as appellant, Ceasar Sanchez, entered the hallway. Luis introduced appellant to Marrero as “Ceasar” and instructed Sanchez to give Marrero the heroin. Luis and Marrero discussed a possible future sale of heroin and then Luis left. Appellant then gave Marrero two bags of heroin. In response to Marrero’s question appellant stated that the heroin would take a “three cut.” Sanchez and Marrero had a brief discussion of possible future sales, and then Marrero gave appellant $2,700, the price quoted by Sanchez for the two bags. The entire transaction took approximately five to ten minutes.
“Ceasar” then re-entered the Market through the back door. Marrero and the informant exited on to Suydam Street and got into their car in front of the Market. As he entered the car, Marrero saw appellant leave the Market, pass in front of the detective’s car, and enter 142 Suydam Street.
*383The testimony of Detective Marrero was corroborated in certain respects by Detective Arthur Drucker. Drucker, who was conducting surveillance near the Market as part of the back-up team for Marrero, observed the detective and the informant enter the Market around 5 p. m. He then saw Luis leave the Market and walk up Suydam Street out of Drucker’s sight, returning within a few minutes. Drucker watched Marrero and the informant leave the Market, walk up Suydam Street until they were out of view, and return approximately ten minutes later to their car. He also testified that he observed appellant leave the market and walk up Suydam Street.1
Appellant, testifying in his own behalf, admitted that he lived at 142 Suydam Street in August of 1977 and that Luis Sanchez was his uncle. He denied having ever seen Marrero or having delivered heroin.
Immediately before the trial a suppression hearing was held when it was revealed by the Government that Marrero had on two prior occasions selected appellant’s photograph out of photographic spreads and that the Government did not plan to introduce evidence of these identifications at trial but would rely upon Marrero’s in-court identification.2 At the hearing Marrero testified that he had two years experience as a detective and eight years experience on the New York City Police Department; that he had spoken to Sanchez for five to ten minutes in the hallway during the heroin sale, that the hallway was illuminated by sunlight coming through clear glass windows in the door, and that he had observed Sanchez’ face. Appellant was standing one to two feet away from the detective during the transaction. Marrero also testified that when appellant passed in front of the car after the sale, the detective had another opportunity to observe Sanchez’ face for 10 to 15 seconds. (¡Marrero had not seen Sanchez in person between the date of the sale, August 23, 1977/ and the first day of trial, December 5, Í978.
The first photographic identification, according to Marrero, was made one or two days after the sale. Marrero stated he selected appellant’s picture from a photo spread of approximately ten pictures. The detective testified that the second photographic identification occurred in January of 1978; this time Marrero selected defendant’s picture from a spread of approximately eight photographs. Detective Drucker testified at the suppression hearing that he was present at the first photographic identification. He could not remember how many pictures were in that spread.
Detective Steve Caracappa testified that he had composed the August, 1977, spread, consisting of approximately six pictures. Caracappa thought the identification occurred five to seven days after the sale, at the DEA Building. After Marrero selected appellant’s photograph, all the other pictures were returned to various files without any special marking of the photographs used or initialing of the one selected by Marrero. Caracappa also composed the January, 1978, spread, which he stated also contained approximately six photographs. As before, after Marrero selected appellant’s photo the other pictures were returned to various files without taking any steps to insure that the photos could be retrieved if needed again.
Both Caracappa and Marrero testified that it is the normal procedure of the PEA, not to make^Sñ^UHoSlSISriS^ the photographic spreads shown to undercover agents.3 No report of'any kind--was'ever made"' of the first photographic identification: a report was made of the second *384photographic identification, but it did not contain any information as to which pictures were used in the spread and made no mention of the prior photo identification. No identifying marks were ever made on the pictures of appellant actually selected by Marrero. Caracappa was unable to remember whether the same mug shot of appellant was used each time, and Marrero stated that he believed different pictures of the appellant had been used. Appellant was not arrested until January, 1978, after the second photographic identification and over four months after the heroin sale which was the subject of the indictment; the in-court identification was made over fifteen months after the sale.
At the end of the suppression hearing Judge Bramwell denied appellant’s motion to suppress the in-court identification by Marrero. The court ruled that “the impermissible suggestiveness, if any, in the photo arrays, does not give rise to the substantial likelihood of misidentification. Officer Marrero is an experienced and trained undercover police officer who spoke to the defendant for five to ten minutes in an illuminated hallway. Moreover,' Agent Marrero demonstrated certainty in his identification.” 4
DISCUSSION
A conviction ([“Eased on eyewitness identification at trial foTlowing a pretrial identification by photograph will~~be~~set aside on’ that ground only if the photo^áphícídentificatiori procedure 'was 'so' impermissibly süggestive as to give rise tcTá t>^ly^lMafitigniMIihb5d”orTrreparaEIo. misidentification." Simmons v. United States, 390 U.S. 377, 384, 88 S.Ct. 967, 971, 19 L.Ed.2d 1247 (1968). See also Manson v. Brathwaite, 432 U.S. 98, 109-14, 97 S.Ct. 2243, 53 L.Ed.2d 140 (1977); Neil v. Biggers, 409 U.S. 188, 198-99, 93 S.Ct. 375, 34 L.Ed.2d 401 (1972). “[R]eliability is the linchpin in determining the admissibility of identification testimony,” Manson v. Brathwaite, supra, 432 U.S. at 114, 97 S.Ct. at 2253, and the court must look to the “totality of the circumstances” to determine whether the in-court identification is reliable even if the pretrial identification procedure was suggestive. Neil v. Biggers, supra, 409 U.S. at 199, 93 S.Ct. 375. See generally United States v. Williams, 596 F.2d 44, 48-49 (2d Cir. 1979).
Since the two pretrial photographic spreads used by Detective Marrero were deliberately not retained by the Government, thus making independent judicial review of their contents impossible, appellant argues that as the initial step in the analysis this court should assume that both spreads were impermissibly suggestive. The Government contends that there is “ample evidence” in the record concerning the contents of the photo spreads to negate such an assumption.5 We disagree. The record of the suppression hearing furnishes glaring proof of why the detectives’ description of the contents of the spreads cannot substitute for independent judicial review. The testimony was filled with contradictions. Marrero thought the first spread contained ten photographs, Drucker said he could not remember, and Caracappa said there were six photos. Caracappa said all of the pictures were mug shots; Marrero could not remember. Drucker did not describe the photographs in that first spread; Caracappa’s and Marrero’s descriptions overlapped only to the extent that both agreed all the pictures were of Hispanic males, some with mustaches, roughly forty years old. Marrero said all the individuals had fair skin and curly hair, but Caracappa made no mention at all of these characteris*385tics. Caracappa thought one of the individuals pictured might have had a beard, but Marrero made no mention of that. The Government’s contention that the descriptions of the composition of the first spread “substantially corroborated” each other is simply factually erroneous unless the Government means to assert that all Hispanic males roughly forty years old look alike — a proposition that is obviously absurd.
The testimony concerning the composition of the second photo spread is even less illuminating. Marrero testified there were eight photographs; he gave no description „o___ o____________r.„„ at all of the individuals pictured. pa stated that there were only six pictures, and his description consisted solely of one statement that all the individuals were Hispanic males in the same age group as defendant. The detectives could not even agree on whether the picture of Sanchez was the same as that used in the first spread. Marrero testified that he believed different mug shots of appellant were used in the two spreads; Caracappa originally testified that the same mug shot was used in each spread, although he thought different prints from the same negative might have been involved, but on cross-examination after learning of Marrero’s testimony he backed down from that assertion and admitted that perhaps different photographs of appellant were used in the two spreads.
In short, the testimony concerning the compositions of the two photo spreads was vague, contradictory, and of absolutely no assistance to a court in making the threshold inquiry of whether the pretrial identification procedure was impermissibly suggestive. Police officers cannot be expected to recall in sufficient detail photographs whose major importance is that they are not photographs of the suspect. The only reliable way for a court to resolve that initial issue is to see the actual photographs, sometMng_whicL.CO.uld_e_asily be done were it not for the Government’s failurejto-retain "any record of the spread used._ Since it is the Government’s action which deprives the court of this essential evidence, it is the Government which must bear the resulting unfavorable inference. Therefore, in the absence of the actual photo spreads usecl, this“rourt~wnr"assmme that the pretrial identification suggestive. procedure was' impermissibly
The “central question” thus becomes “whether under the ‘totality of the circumstances’ the identification was reliable” in spite of the impermissibly suggestive pretrial identification procedure. Neil v. Biggers, supra, 409 U.S. at 199, 93 S.Ct. at 382. See also Manson v. Brathwaite, supra, 432 U.S. at 109-14, 97 S.Ct. 2243. The relevant factors are the witness’ opportunity to view the criminal at the time of the crime, the degree of'attention of the witness at that time, the accuracy of any prior description by the witness of the criminal, the level of certainty of the witness, and the length of time between the crime and the identification. See Manson v. Brathwaite, supra, 432 U.S. at 114-16, 97 S.Ct. 2243; Neil v. Biggers, supra, 409 U.S. at 199-200, 93 S.Ct. 375. Against these factors must be weighed “the coiru^ .ot.th.e.a.ui£ V. gestivA,ldentific,atiQiLftself.” Manson Brathwaite, supra, 432 U.S. at 114, 116, 97 S.Ct. at 2253.
Detective Marrero had an excellent opportunity to view the appellant. They were standing one to two feet apart from each other in a small hallway for five to ten minutes. For a period of that time Marrero and the seller were engaged in direct conversation. The hallway was illuminated by the afternoon summer sunlight coming through the clear glass “portholes” in the door. These circumstances present at least as good an opportunity to view as existed in Manson v. Brathwaite, supra, 432 U.S. at 114, 97 S.Ct. 2243.
The degree of attention of an undercover police officer engaged in the purchase of narcotics for the explicit purpose of gathering evidence for a later prosecution of the seller will obviously be quite high. In addition, Marrero had eight years’ experience on the police force, two of those years as a detective. He was “not a casual or passing *386observer, as is so often the case with eyewitness identification.” Manson v. Brathwaite, supra, 432 U.S. at 115, 97 S.Ct. at 2253.
-There is no evidence in the record of any prior description of the criminal by Marrero.6 However, Marrero did demonstrate certainty in identifying appellant as the seller, as was found by the trial court after seeing and hearing the witnesses during rigorous cross-examination by defense counsel.
Finally, although there was a period of over fifteen months between the crime and the in-court identification, which is a “somewhat negative factor," United States v. William, supra, 596 F.2d at 49, we do not feel that this is so long an intervening period of time as to render the identification unreliable. Cf. id. (in-court identification by lay witness was sufficiently reliable to warrant its admission into evidence despite time lapse of two years and eight months between crime and in-court confrontation).
Following the criteria established in Neil v. Biggers and Manson v. Brathwaite, therefore, we conclude that under the totality of the circumstances, even assuming as we must that the pretrial photographic identification procedures were impermissibly suggestive, there is not a “very substantial likelihood of irreparable misidentification.” Simmons v. United States, supra, 390 U.S. at 384, 88 S.Ct. at 971. Although the jury in this case was unaware of the impermissibly suggestive photographic identification, it obviously chose to believe the testimony of Detective Marrero after thorough and accurate instructions by the trial court regarding the factors to be considered by it in evaluating the identification testimony. The decision whether or not to introduce evidence of the impermissibly suggestive photographic spread was, of course, one of trial strategy to be made by appellant’s counsel after weighing the risks and possible advantages of such a course. The record does not show any request by appellant for a live line-up prior to trial.
Appellant invites us to exercise our supervisory power to reverse his conviction in light of the Government’s acknowledged policy of not keeping any record of the photographs used in photo spreads shown to undercover agents and of not marking the photo selected by the agent for identification. We decline to take that step at this time, but we note that such a prophylactic measure may well be necessary if the Government continues the careless procedures described in this case.
That the photographic identifications may have ostensibly been for purposes of enabling others to arrest the suspect rather than for evidence at trial hardly justifies the cavalier approach of the Government. It should have been painfully obvious to the DEA at the time of the first photo spread that virtually its entire case against appellant turned on the identification testimony of one person, Detective Marrero. As far as the record shows, nothing prevented Marrero from furnishing a detailed written description of appellant immediately after the heroin transaction forming the basis of the charge or from joining in the later arrest. Once the Government chose instead *387to use photo spreads it assumed the burden of retaining them for judicial review to insure that there had been no “corrupting effect,” Manson v. Brathwaite, supra, 432 U.S. at 114, 97 S.Ct. 2243, on that vital identification testimony by the use of improperly suggestive pictures. The danger of impermissibly suggestive pretrial identification procedures is that if a misidentification occurs, “the witness thereafter is apt to retain in his memory the image of the photograph rather than of the person actually seen, reducing the trust-worthiness of subsequent lineup or courtroom identification.” Simmons v. United States, supra, 390 U.S. at 383-84, 88 S.Ct. at 971. The mere fact that a police officer may have better training and experience in identifying criminals does not render such an officer free from the human frailties which have motivated the concern of the courts in this area. Cf. Manson v. Brathwaite, supra, 432 U.S. at 114-16, 97 S.Ct. 2243 (experience of police officer only goes to one factor, the degree of attention, in the “totality of the circumstances” test which applies to all witnesses). See also id. at 130-31, 133-35, 97 S.Ct. 2243 (Marshall, J., dissenting).
It is therefore important, if the rights of the defendant are to be preserved and the accuracy of the truth-finding function of the trial is to be protected, that the photographic spread shown to the undercover agent be recorded for later judicial review. The announced policy of the Government of not retaining the spread or any records from which the spread could be accurately reconstructed is totally without justification. We trust that the Government will abandon this policy rather than force us to exercise our supervisory power to adopt a more harsh prophylactic rule, such as the exclusion of the testimony of the witness who saw the unreviewable photographic spreads.
The judgment of conviction is affirmed.