Miles, Ex Parte Richard Ray Jr.

OPINION HERVEY; J., delivered the opinion of the Court in which KELLER, P.J., and WOMACK, JOHNSON, KEASLER, COCHRAN and ALCALA, JJ., joined. Applicant, Richard Ray Miles, Jr., was convicted of murder and attempted murder and sentenced to forty years’ and twenty years’ confinement, respectively. The convictions were affirmed on appeal. Miles v. State, Nos. 05-95-01280-CR & 05-95-01281-CR, 1997 WL 867103, 1997 Tex.App. LEXIS 3431 (Tex.App.-Dallas July 2, 1997, no pet.) (not designated for publication). Applicant filed initial applications for writ of habeas corpus arguing that the State violated Brady v. Maryland 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963); we denied relief. Ex parte Miles, Nos. WR-64,325-01 & WR-64,325-02 (Tex.Crim.App. April 4, 2007). Applicant now files subsequent applications for writ of habeas corpus claiming that he is actually innocent, that the expert testimony on gunshot-residue analysis is no longer reliable, and that the State violated Brady. The trial court entered findings of fact and conclusions of law recommending that relief be granted on all grounds. Relief is granted. *651 I. FACTS In the early morning hours of May 16, 1994, the victims, Deandre Shay Williams and Robert Ray Johnson, Jr., stopped at a Texaco service station located on Northwest Highway in Dallas. At approximately 2:50 a.m., while Williams and Johnson were sitting in them parked car, a black male wearing dark shorts, a white tank top, and a floppy hat walked up to the car. Reaching into the car from the driver’s side, he shot both men with a nine-millimeter handgun. The shooter then ran from the scene and got into a white Cadillac, which was then driven away. Williams was killed, and Johnson sustained severe injuries. Approximately twenty-five minutes after the shooting, Applicant was arrested on a nearby street. He was subsequently indicted for the murder of Williams and attempted murder of Johnson. A. State’s Evidence On the day of the offense, Johnson and Williams spent the afternoon driving through Glendale Park, showing off Johnson’s car, a Nissan 300 ZX with gold wheels. Johnson and Williams left the park between 8:00 and 9:00 p.m. and went to the New York, New York night club located on Northwest Highway near Bach-man Lake. The Nissan 300 ZX was parked prominently in front of the club. Johnson and Williams left the club together shortly after 2:30 a.m. and drove down Northwest Highway. When they approached a Texaco service station, Williams asked Johnson to stop the car so that he could talk to a woman walking on the sidewalk. Johnson parked at the Texaco. At approximately 2:50 a.m., a male walked up to the car, and reaching into the car from the driver’s side, he shot both men multiple times with a nine-millimeter handgun. Williams was killed, 1 and Johnson sustained severe injuries on the left side of his body and neck. Johnson testified that he recalled seeing someone at the scene who bore a resemblance to Applicant. However, he could not identify Applicant as the shooter because the shooter was “darker.” Johnson also testified regarding a twelve-gauge shotgun that he kept in the back of his car between the passenger seats. He explained that he lived in a hostile neighborhood, which required taking steps to protect himself and his car, and he was told by a police officer that “it was the only thing you could carry legally in your car.” The shotgun had been in the car on the day of the shooting, and Williams had apparently taken the shotgun from the rear of the car and placed it at his feet some time between leaving the club and arriving at the Texaco. Testimony revealed that shortly after the shooting, someone removed the shotgun from the car; Johnson later recovered the gun and sold it to a friend. Johnson denied that the shotgun was pulled out from the seats while he and Williams were at the park earlier in the day. He also denied that he or Williams had any type of confrontation there. However, when he was recalled by the defense, Johnson conceded that, although they did not have problems while at the park, they had been confronted and threatened by a group of black males in a white Cadillac at Redbird Mall earlier that day. Marcus Thurman testified that he was standing in line to buy gas when he heard *652 six or seven shots fired. He turned around and saw a black male running with a gun in his right hand, heading from the direction of the victims’ car and toward the bushes behind the Texaco. Thurman estimated that he was within twenty feet of the suspect, and although the view lasted for only a second or two, Thurman stated that he clearly saw the shooter’s face because of the well-lit parking lot. Thurman also observed the shooter’s body build and noted that he was wearing “a white tank top, some dark shorts and a black floppy hat.” Thurman testified that there was no one else around the vehicle at the time that the shots were fired, and no one else in the immediate area was dressed like the shooter. Thurman moved to the edge of the Texaco building and saw the shooter run down into the bushes. Fifteen to twenty seconds later, a white Cadillac pulled up with its lights off. The shooter came out of the bushes, with his white tank top in his hand, and he got into the right rear passenger side of the vehicle. Thurman testified that he did not' see a gun at that point. The Cadillac drove off down Lemmon Avenue. Thurman immediately got into his own car, and while he used his cell phone to call 9-1-1, he followed about a “[b]lock, block and a half’ behind the Cadillac. He observed the Cadillac make a u-turn in front of the Sewell Cadillac dealership, and the shooter exited the car. The Cadillac continued north on Lemmon Avenue, and the shooter began to walk down the same street toward Northwest Highway. Thurman also made the u-turn and drove north on Lemmon Avenue. As he passed the shooter, he could not see the man’s face clearly, but he was sure that it was the same person. Thurman testified that the shooter’s white tank top was back on and described his shorts as hitting at the knee, although he did not know if they hit above or below the knee. Thurman remained on the phone with 9-1-1 throughout these events. Thurman returned to the Texaco. By that time, officers from the Dallas Police Department (DPD) had arrived and were attempting to protect the crime scene. Thurman approached Sergeant Gary Tolle-son and informed him of what had transpired. At approximately 3:10 a.m., Tolle-son broadcast Thurman’s description over the police radio, stating that the suspect was a black male wearing a white tank top, floppy hat, and dark, baggy shorts and that the suspect vehicle was a white Cadillac. A short time later, a patrol car pulled into the Texaco. Thurman observed a man in the back seat, and he said, “That’s the guy that done the shooting.” Thurman stated that he did not believe that the officers were attempting to show Applicant to him. Police officers then took Applicant out of the car and had him stand up, before putting him back in to the car. Thurman testified that he was subsequently taken to the police station, and after a thirty-minute wait, he was shown a photographic lineup consisting of six photos (State’s Exhibits 20-25), and Thurman identified the person in Exhibit 20 (Applicant’s photograph) as being the shooter. Thurman stated that Applicant’s skin appeared to be lighter than the skin of the person in Exhibit 20. Still, Thurman made an in-court identification of Applicant as the shooter: Q. [STATE]: Can you tell the jury today, is there any doubt in your mind that the person you’ve identified in court today is the same person you saw come by and drive by or, excuse me, run by and get in that car? [[Image here]] Is there any doubt in your mind, sir? *653 A. [THURMAN] No, it’s not. Q. The fact — Well, let me ask you this. Your identification in court today, is that based on viewing those pictures an hour or two later or viewing the person in the car or viewing the suspect as he ran by you with the gun? A. Viewing him as he ran by me. Q. Is that the time you had the clearest opportunity to see the person’s face? A. Yes, it is. Sergeant Tolleson testified that, upon arriving at the Texaco, he attempted to protect the crime scene by controlling the crowd. Within five minutes, Thurman approached him and gave him a description of the shooter (a black male in his mid-twenties, wearing a white tank top, floppy hat, and dark baggy shorts) and the getaway vehicle. Tolleson broadcast that information over the radio immediately. Tolleson stated that he also called for a helicopter “in case there [was] a foot chase or any other need.” Tolleson testified that between three and five minutes had passed between the initial broadcast of the description and the apprehension of Applicant. Officer Ron Goodnow, an off-duty police officer working at Sewell Cadillac, testified that he heard Tolleson’s broadcast on his police radio. Three minutes prior, Good-now had seen a black male wearing a black hat, a white tank, and dark pants or shorts walk by the dealership. Copsequently, he called in that he had seen a possible suspect matching the description walking down Lemmon Avenue in the direction of Northwest Highway, specifically toward Bluffview. Goodnow testified that he had not seen the suspect get out of a Cadillac; rather, he had simply seen the man walk past the dealership. Officer Dale Lundberg and his partner were on patrol that evening, and when they heard of Goodnow’s sighting, they drove to the area where the suspect had been heading. Lundberg testified that they observed Applicant standing on the corner of Lovers Lane and Bluffview Street, located about a block or block-and-a-half from where Goodnow had seen the suspect. Lundberg noted that Applicant had on a floppy, “velvety” hat; a white tank top with some writing; and blue “jams,” which “are like pants that come up in between the ankles and the knees ... they are not quite full length trousers but they are in between pants and trousers.” After verifying the suspect’s description, Lundberg radioed that he had the suspect. At about the same time, the police helicopter arrived and shone a spotlight on Applicant. When Applicant walked across the street, the officers got out of their vehicle and arrested him. Lundberg testified that Applicant did not resist, no weapons were found on him, and he denied any wrongdoing. The officers secured Applicant in the backseat of the squad car and drove him to the Texaco, where they parked in back so that Applicant would not be observed. They removed Applicant from the car, still handcuffed, so Detective Richard Dodge of the DPD’s Physical Evidence Section could perform a gunshot-residue hand washing on him. Subsequently, the officers took Applicant to the police station. Detective Dodge testified that he was responsible for photographing the crime scene and collecting physical evidence. 2 Dodge did Applicant’s handwashing, which is “a process of taking a chemical and applying it to a person’s hands and the object is to see if they have come in contact with a firearm.” Dodge found bullets *654 and nine-millimeter shell casings in the victims’ car and on the driveway, all within two feet of the car. He was also able to lift some prints, but after running the prints through AFIS (the automatic fingerprint identification system), none of the prints were identified. Applicant’s prints did not match. Robert Poole, a firearm and toolmark examiner, testified that all of the bullets recovered at the scene were nine-millimeter-caliber projectiles, and a microscopic examination indicated that they came from the same gun. Detective B.J. Hooker, the lead homicide detective on the case, testified that he arrived at the Texaco around 4:00 a.m. and stayed for fifteen to twenty minutes. 3 He then went to the police station. There, Hooker took a Polaroid photograph of Applicant, which he used to create a photographic lineup of Applicant and five other individuals (State’s Exhibits 20 thru 25, with Applicant’s photo being Exhibit 20). Of the six individuals, Applicant was the only one wearing a white tank top. 4 Hooker showed Thurman the photos within hours of the offense, and Thurman identified Applicant’s photograph as that of the shooter. Thurman also submitted an affidavit, recalling what he had observed. Later that morning, several witnesses were transported to the station where they were also shown the photographic lineup, and affidavits were taken. Hooker stated that five witnesses submitted sworn statements indicating that the gunman was “dark skinned” and wore long shorts and a hat; all except for one (Stacy Williams) described the shooter’s shirt as a “T-shirt,” and only one (Christopher Shawn Beam) mentioned that the shooter’s shirt had writing on it. However, none of those five witnesses were able to pick out a photograph from the array because they had not seen the shooter’s face. Johnson was also unable to identify the shooter from the photo lineup. 5 Hooker further testified that, at the time of his arrest, Applicant was wearing “long blue Dickies pants.” Hooker had retrieved Applicant’s clothing from the property room and had taken pictures of the Dickies pants and the tank top (State’s Exhibits 16-19). The hat had been lost; Hooker stated that he had seen the hat sitting on the desk and believed it had been thrown away. The defense questioned Hooker about the discrepancy between the witnesses’ description of the shorts worn by the shooter and the Dickies worn by Applicant: Q. [DEFENSE]. Would it be reasonable that as a homicide investigator, when you’ve got six witnesses that saw a shooter in shorts and you’re brought a suspect wearing full-length pants, that you begin to think maybe they’ve brought the wrong guy? A. Well, that crossed my mind. Yes, it did. Hooker stated that he also questioned Applicant about his whereabouts on the night of the shooting. Applicant provided him with the names and phone numbers of three individuals who would support his alibi. Hooker testified that those individuals provided information that was potentially beneficial to Applicant. When asked directly if the alibi checked out, Hooker *655 answered, “Somewhat, yes, it did,” but he also stated that it did not tend to exonerate Applicant because it did not disprove the fact that a witness had identified him. Hooker also asked Johnson whether he had had any confrontations prior to the shootings. Johnson told him that he had been confronted by a number of black males in a white Cadillac at Redbird Mall earlier on the day of the shooting. According to Johnson, the Cadillac had attempted to block him in a parking place, and one of the occupants of the Cadillac had a pistol. Hooker stated that he was unable to locate anyone who knew more about the incident at the Redbird Mall. Vicki Hall, a trace evidence analyst with the Southwest Institute of Forensic Sciences (SWIFS), testified that she had examined the gunshot-residue swabs taken from Applicant’s hands. She stated that, relying on FBI standards, she found elevated levels of antimony and barium on only the wipings from Applicant’s right hand palm. Hall explained, “The fact that I found elevated levels of antimony and barium on the palm of the right hand would lead me to believe that the characteristics of gunshot residue are present, meaning that gunshot residue was deposited on the palm of the right hand in some fashion, either firing a weapon or handling a very dirty weapon.” Hall was not concerned that the residue was found only on the right palm because Applicant’s washings were done between thirty and forty minutes after the shooting, and “the more active a person is, the more possibility you have of wiping any residue off.” On cross-examination, Hall emphasized that, because the residue ratios in this case fall within the FBI guidelines, Applicant’s residue combination is “unique” to gunshot residue. 6 However, Hall recognized that the levels found on the swabs from Applicant’s hands were very low. She also referred to documented cases of known shooters having positive results only on the palm and not the back of the hand, of known shooters with negative results, and of known non-shooters with presumptive levels. In addition, Hall agreed that the existence of residue does not necessarily mean that the individual fired the gun; he could have handled it or encountered it from other sources. For example, antimony is found in matches, car batteries, solders, clays, and pottery, and a handcuff transfer could be a possible (but unlikely) source. On re-direct examination, Hall stated that she could not exclude batteries and matches as sources of the residue found on Applicant’s right palm — “All I can do is tell you that the ratio that I found falls within a set-up guideline of the FBI; therefore it’s characteristic of gunshot residue and not of contamination.” She also acknowledged that it would be possible to wipe off the residue on one side of a hand but not the other side. Then on re-cross examination, she again stated that “the antimony and barium could have come from car batteries or matches or dirt; but the fact that they fall within a ratio leads me to believe that they’re not from those sources” — it is more likely that they are not from those sources, but they cannot be completely excluded. B. Defense Evidence Henry Evans testified that he still had his foot on the brake when he heard the shots at the Texaco. From about thirty *656 feet away, he saw a man standing up and shooting into the car. Evans stated that he could not see the suspect’s face, but he had a clear view of his back and observed his height, build, and overall appearance. Evans testified that the shooter was between 6'2" and 6'4" with a slim build and “real dark skin,” and that he wore a tank top, shorts about knee length, white socks rising three inches from his ankle, and tennis shoes. According to Evans, the shooter did not resemble Applicant. Evans asserted that the shooter had darker skin than the person in State’s Exhibit 20, and he also stated that Applicant looked lighter in the courtroom than he did in the Polaroid. Cassandra Knight testified that she had been friends with Johnson since high school. When the shooting occurred, she was parked next to the victims’ car, trying to talk to Johnson through her passenger-side window. In her passenger-side mirror, she had seen someone approaching, and shortly thereafter, he began shooting. Knight stated that she could not identify Applicant as the shooter because she had seen the shooter’s body but not his face. She also claimed that the shooter’s skin was darker than Applicant’s appeared to be at trial, but State’s Exhibit 20 reflected the shooter’s skin color. James Yarborough testified that he lived one block from the intersection of Lovers Lane and Bluffview, and that Applicant had been staying with him. Yarborough had dropped Applicant off in the Oak Cliff area earlier that evening, and he told Applicant to call when he returned to the neighborhood so that he could unlock the front door. Yarborough received a call from Applicant some time after midnight. He estimated it was at 1:00 a.m., but he “really didn’t know actually what time it was really” because he had been asleep when Applicant called him. Yarborough stated that he heard the helicopter overhead within three or four minutes of that phone call. Ernest Clark testified that he saw Applicant at an apartment complex in Oak Cliff, where Clark’s mother lived, on the night of the shooting. Applicant ended up in Clark’s mother’s apartment around 1:30 a.m., watching television. Clark had informed Applicant that he intended to go to his girlfriend’s house on University, about two blocks from Sewell Cadillac, and Applicant asked for a ride home. They left Oak Cliff shortly after 2:00 a.m. They drove down Lemmon Avenue and stopped at a 7-Eleven store for about ten minutes while Applicant went into the store to get cigarettes, went back into the store to get a light, and then partially smoked a cigarette. Subsequently, Clark continued driving down Lemmon Avenue. He made a couple of turns, and around 2:30 or 2:45 a.m., Applicant got out of the car at the corner of Roper and University. 7 Clark testified that a day or so after the offense, a police officer called him, and Clark told him that Applicant was wearing a hat and a sleeveless shirt that night. Applicant testified on his own behalf and denied committing the instant offenses. He asserted that he had never been to the Texaco where the shooting occurred nor had he been to the nightclub visited by the victims. He also maintained that he did not carry a gun and has never shot one. Applicant stated that he is left-handed and that his skin has not changed much since being in jail. Applicant further asserted that he handled matches that day because he smoked cigarettes and that he had also been in and around cars and their hoods. Regarding his alibi, Applicant explained that on the evening prior to the shooting, *657 Yarborough drove him to the apartments in Oak Cliff so that Applicant could visit with his girlfriend, Betty Hogan. Later that evening, Applicant asked Clark if he would give him a ride home because he knew that Clark’s girlfriend lived near the house where he was staying. Applicant and Clark eventually left the apartment complex and drove north. They stopped so Applicant could buy cigarettes at a 7-Eleven on Lemmon Avenue around 2:25 a.m. Then, upon reaching the intersection of University and Roper, Applicant gave Clark five dollars, and he exited the vehicle. Applicant walked down University to Lemmon Avenue. He turned right on Lemmon, walking in front of Sewell Cadillac to the Red Coleman store on Bluff-view. There, he used the pay phone to call Yarborough to tell him that he was just a couple of blocks away and to ask him to unlock the door. That is when Applicant noticed a police car on the side of the road, and the police helicopter’s spotlight shone down on him. Applicant was arrested and taken to the Texaco. Once at the police station, Applicant told the police that he had come from Oak Cliff and gave them the phone numbers of Hogan, Clark’s mother, and Clark’s girlfriend. According to Applicant, the officer who checked out the story said that it had “added up.” C. Closing Arguments and Verdict During closing arguments, the State argued that there was overwhelming evidence supporting Applicant’s guilt. The State focused largely on Thurman’s identification, arguing that “[t]he only impartial witness in this whole thing is Marcus Thurman.” It noted that Thurman followed the shooter, the officers did not see anyone else, Applicant was arrested within a few minutes not far from the location of the shooting, and gunshot residue was found on his palm. The State also attacked Applicant’s alibi by suggesting that it is easy to get an alibi instruction into the jury charge because, according to the State, all one has to do is have a witness or himself say “No, I wasn’t there. It wasn’t me.” The State ended its argument in part by stating “ [i]f nothing else — I mean, there’s no other suspects. No one else is going to get arrested, because he’s sitting right here. And these families need to have some justice.” In contrast, the defense asserted in its closing argument that there was reasonable doubt “raised by all of the testimony and all of the evidence in this case,” including Thurman’s testimony. The defense noted that Thurman is the only witness out of twenty or thirty who identified Applicant as the shooter, and it detailed the discrepancies between the description of the shooter and Applicant’s appearance (e.g., that the shooter was darker than Applicant and that Applicant was found in full length trousers rather than shorts). The defense further argued that Applicant did not concoct his alibi, referring to supporting testimony. The defense also emphasized how low the hand-washing levels were and how the elements can come from other sources such as matches and batteries, items Applicant had touched that day. Finally, the defense suggested that the confrontation that occurred at Redbird Mall provided a motive for the shooting. Subsequently, the jury found Applicant guilty of both the murder of Williams and the attempted murder of Johnson. He was sentenced to forty years’ and twenty years’ confinement, respectively. II. DALLAS COURT OF APPEALS On appeal, the Dallas Court of Appeals upheld Applicant’s convictions. Miles, 1997 WL 367103, 1997 Tex.App. LEXIS 3431. The court of appeals rejected Applicant’s legal-sufficiency and factual-sufficiency arguments. Id. at *3, *4, 1997 Tex. App. LEXIS 3431, at *7, *10. In reviewing the evidence presented at trial (includ *658 ing Thurman’s description of the shooter, Thurman’s three separate identifications of Applicant as the shooter, the gunshot-residue analysis, and Applicant’s alibi), the court of appeals emphasized “the jury’s prerogative to reconcile any conflicts in the testimony” — the jury was free to believe some pieces of evidence and not others. Id. at *4, 1997 Tex.App. LEXIS 3431, at *7. Additionally, the court of appeals overruled Applicant’s argument that the trial court erred in denying the motion to suppress Thurman’s in-court identification of Applicant. Id. at ⅞4-6, 1997 Tex.App. LEXIS 3431, at *10-15. Applicant argued that “Thurman’s in-court identification of appellant was tainted because it was based on an impermissible ‘one man show-up’ and an impermissible photographic array.” Id. at *4, 1997 Tex.App. LEXIS 3431, at *10-11. The court determined, however, that Thurman’s recollections of Applicant’s appearance “sufficiently served as an independent origin for the in-court identification regardless of the propriety of the pretrial identifications.” Id. at *6, 1997 Tex.App. LEXIS 3431, at *14-15. III. INITIAL APPLICATIONS FOR WRIT OF HABEAS CORPUS On October 13, 2005, Applicant filed his initial applications for writ of habeas corpus. Applicant alleged, inter alia, that the State violated Brady when it failed to disclose before trial a police report that contained exculpatory information. The police l’eport concerned the confrontation between Johnson and several black males in a white Cadillac at Redbird Mall on the evening of the instant offense. The trial court entered findings of fact and conclusions of law recommending that relief be denied because the record demonstrated “that, pre-trial, the defense attorney knew about exculpatory evidence, and used this information in opening statement; the prosecution elicited in direct examination that the victim was unable to identify his assailant; and the defense attorney effectively cross examined Detective Hooker about the exculpatory evidence.” We agreed and denied habeas relief without written order. Ex parte Miles, Nos. WR-64,325-01 & WR-64,325-02 (Tex.Crim. App. April 4, 2007). IV. SUBSEQUENT DEVELOPMENTS Applicant’s father made a Freedom of Information Act (FIA) request to the DPD asking for information about their investigation. By a letter dated November 18, 2004, the DPD turned over 25 pages of information related to the cases, but the police reports that are the subject of the pending writs were not disclosed. Subsequently, Applicant personally wrote a letter regarding a second FIA request with the DPD, noting that he had received the information per his father’s request and asking for additional information that he thought was missing. The record does not reflect whether this request was sent to or received by the DPD or whether there was any response to Applicant’s personal FIA request. Centurion Ministries, Inc. filed another FIA request, on behalf of Applicant, which the DPD received on March 27, 2007. Upon reviewing the material subsequently received from the DPD, Applicant’s attorney discovered two police reports for the first time. The first report was a single-page police report reflecting an “Anonymous phone call on shooting” to the DPD’s Crimes Against Persons Division (CAPERS). 8 The call was received on May 8, *659 1995, approximately a year after the offense, but over three months prior to Applicant’s trial. The memorandum indicated that an unknown female “called and stated that she possibly had some information about a shooting that occurred by Bachman [L]ake sometime last year ( [1994]), between May and June.” She stated that her ex-boyfriend, Keith Richard (aka SIX), told her that he shot two black males at a Texaco near the lake using a nine-millimeter pistol that he still possessed, and that the police arrested someone but they had the wrong person. The officer receiving this information apparently did some research and was able to locate the shooting at issue, making notations on the report of such. The second undisclosed report involved an altercation between the victims and a third person, William Garland, that occurred on May 11, 1994, five days before the shootings in the instant case. 9 Garland informed the police that Williams and Johnson came to his place of work, and as he walked to his car, they “verbally abused” him. Then, while Williams searched Garland’s car, Johnson pulled a sawed-off shotgun from behind his car seat and pointed it at Garland, informing him that “he was here to take care of him” because he had beaten up Johnson’s cousin. Williams and Johnson then drove away. Four days later, on May 15, 1994, Garland ran into Williams and Williams’ brother Kimerick at a car wash, and Williams apologized for the previous altercation. Garland also advised the police that Kimerick later “told him that a dude by the name of Deuce had killed his brother.” 10 In an affidavit dated February 23, 2009, Edward Gray, who represented Applicant during his trial, addressed the undisclosed reports. Gray asserted that before trial he “specifically inquired of the prosecuting attorney if he knew of any other evidence which ‘would tend to corroborate the alibi of the defendant or any other evidence which would tend to be possibly exculpatory in nature or which would lead to exculpatory evidence.’ ” The prosecutor responded that he had uncovered no such evidence. Gray also filed a motion for production of exculpatory material, which was granted. However, “[t]hese documents were never produced to me by the State. The first time that I ever saw these documents was when they were provided to me by representatives of Centurion Ministries.” Gray further stated that the two undisclosed reports would have been material to the preparation of his defense of Applicant because they would have allowed him to investigate other possible suspects and to impeach Johnson about the use of the shotgun. James McCloskey, the founder of Centurion Ministries, Inc., also filed an affidavit dated September 11, 2009, related to the undisclosed reports. McCloskey stated that he had interviewed Keith Richard. He asserted, “Richard is much darker in complexion than [Applicant]. Richard is much taller, appearing to be the 6'6" that is described in court records. [Applicant], at 5'9", is average height.” McCloskey explained to Richard that, on behalf of Applicant, he was investigating a shooting at a Texaco station near Bachman Lake, and Richard’s name had surfaced as a result of an open records report. McClos-key stated, *660 Richard’s first words were “I’ll never forget that night [.”] Richard explained that he had been at the New York New York nightclub with his cousin, Richard Parks. When they got home around 3:00 a.m., they saw television coverage of the shooting. Richard recalled that they “said to each other ‘Geez, we were there and left right before the shooting’ but we didn’t hang around and instead ivent home to Oak Cliff.” Richard also told McCloskey that he had been a drug dealer and was very familiar with the streets of Dallas. Additionally, he was a “ ‘great’ basketball player who could play with the best of them,” and he played against the players of Navarro College and Notre Dame. Attached to his affidavit, McCloskey provided a time line reflecting the events surrounding the shooting and Applicant’s activities and a map of the paths followed by the Cadillac and Applicant. These documents indicated that it was reasonable to believe that Applicant was approximately twelve minutes behind the shooter; in other words, when the officers saw Applicant on the corner, the shooter had likely left the area some time before and Applicant was simply walking from where he exited Clark’s car to Yarborough’s house. Faye Springer, a criminalistics consultant, submitted a “Review of Gunshot Residue Analysis by Atomic Absorption,” which was dated August 29, 2009. After reviewing Hall’s report and testimony, Springer explained that Hall analyzed the swabs taken from Applicant’s hands with Atomic Absorption Spectroscopy (AAS). Springer also noted that Hall used the method and criteria established by the FBI, but Springer could not find any reference to these standards in the FBI publications or in her handwritten notes from a 1986 FBI gunshot-residue class. Springer would have reported that the amount of antimony and barium on Applicant’s right palm “would be described as a low level of antimony and if the barium level was equal to or greater than antimony, then [she] would state that these elements could be from gunshot residue. [She] would also describe in the report that [she] could not exclude other environmental sources of these elements.” Springer stated that “[t]he most fundamental problem with Ms[.] Hall’s testimony is that she testifies twice ... that the level of antimony and barium on the swabs from [Applicant’s] hand is unique to gunshot residue.” According to Springer, studies indicate that about ten percent of the non-shooting population have antimony and barium levels like those found on Applicant, so Hall “overstated the significance of her finding in this case.” Further, Springer concluded that “the distribution of the antimony and barium on the hands of [Applicant] is not typical of an individual who had fired a gun.” She asserted that the “more likely explanation” for the residue distribution on only the right palm is that Applicant “touched a gun or object contaminated with gunshot residue versus having fired a gun. However, the effect of post-shooting activity on the distribution of gunshot residue is difficult to predict.” 11 Applicant’s attorney filed a FIA request with the Dallas County District Attorney’s Office on February 23, 2009. The police reports at issue were not contained in the *661 records of the District Attorney’s files and were, therefore, not provided to Applicant’s attorney. V. SUBSEQUENT APPLICATIONS FOR WRIT OF HABEAS CORPUS On September 18, 2009, Applicant filed the present applications and raised three grounds: (1) The State failed to produce police reports which identified other potential suspects for the shooting. These police reports should have been provided pursuant to Brady. (2) Gunshot analysis was introduced at Applicant’s trial. The scientific analysis of gunshot residue has evolved since 1994. The testing approach and standards used in Applicant’s case are no longer deemed sufficiently reliable. (3) Applicant is factually innocent of the charges for which he was convicted. Applicant filed a supporting memorandum and attached exhibits, which included the two undisclosed police reports, the affidavits of Edward Gray and James McClos-key, and Springer’s gunshot-residue analysis. On October 6, 2009, the State responded to Applicant’s writ applications and conceded that the DPD had the two police reports that had not been turned over to the DA’s office. Consequently, the State did “not oppose the granting of this habeas application for this reason.” VI. FINDINGS AND CONCLUSIONS OF THE HABEAS COURT AND OTHER DEVELOPMENTS On October 12, the habeas court adopted stipulated findings of facts and conclusions of law proposed by Applicant and the State. The habeas court concluded that a Brady violation resulted from the prosecution’s failure to disclose the anonymous phone call and Garland reports. It did not address Applicant’s other claims because “the State is still conducting its investigation regarding” such claims. Then, on January 6, 2010, Marcus Thurman recanted his in-court identification of Applicant as the shooter. 12 In his affidavit, Thurman contended that, at the time of trial, he was unable to identify the shooter, and when he told the State of this fact, the prosecutor directed him to identify Applicant by showing him where Applicant would be seated in the courtroom. Consequently, on February 4, 2010, the habeas court adopted supplemental stipulated findings and conclusions of law, which were agreed to by Applicant and the State. The court reiterated its October 12 findings and conclusions and added that Thurman’s affidavit, if true, supported its previous findings and conclusions, which concluded that Applicant is entitled to a new trial. On February 24, 2010, before receiving the supplemental findings and conclusions, we remanded Applicant’s subsequent writs to determine whether his Brady claim was barred under Article 11.07, Section 4 of the Texas Code of Criminal Procedure. Ex parte Miles, Nos. WR-64,325-03 & WR-64,325-04, 2010 WL 653581, 2010 Tex. CrimApp. Unpub. LEXIS 130 (Tex.Crim. App. Feb. 24, 2010) (not designated for publication); see Tex.Code Crim. Proc. art 11.07, § 4. We directed the trial court to (1) determine whether Applicant’s claim concerns the State’s failure to disclose the same police report as Applicant’s previous Brady claim, which was litigated in his first application; (2) if a different report is at issue, determine whether Applicant *662 could have discovered the police report at issue here, before he filed his prior application if he had exercised due diligence; and (3) make any other findings of fact and conclusions of law deemed relevant and appropriate to the disposition of Applicant’s claims for habeas corpus relief. On March 2, 2010, we received the supplemental findings and conclusions that the habeas court had adopted on February 4. Because this supplement contained no findings or conclusions regarding Section 4, on May 5, we again remanded these applications and directed the trial court to determine whether Applicant’s Brady claim was barred under Section 4. Ex parte Miles, Nos. WR-64,325-03 & WR-64,325-04, 2010 WL 1795172, 2010 Tex. Crim.App. LEXIS 269 (Tex.Crim.App. May 5, 2010) (not designated for publication). We also asked the trial court to determine whether (1) the affidavit Thurman submitted was credible, (2) Applicant had established actual innocence, and (3) the expert testimony on gunshot residue is no longer reliable. On May 24, 2010, the habeas court entered its second supplemental stipulated findings of fact and conclusions of law, also stipulated to by Applicant and the State, which responded to our original February 24 remand orders. The habeas court determined that Applicant’s new Brady claim “concerns two completely different police reports concerning completely different subject matter than the one litigated in Applicant’s first writ.” The court also concluded that the factual basis of Applicant’s new Brady claim was not available when he filed the -01 and -02 applications and that Applicant used due diligence in trying to obtain the two police reports at issue. Hence, his claim was not procedurally barred by Article 11.07, Section 4. Subsequently, on July 27, 2010, Vicki Hall signed an affidavit related to her trial testimony. After reviewing the transcript of her trial testimony, the original SWIFS file on testing, and Springer’s affidavit, she concluded, “If I were to testify in this case today, I would testify differently than I did in [Applicant’s] trial in August of 1995.” She continued, “[a]ccording to SWIFS standards, both in 1995 and now, the levels of barium and antimony detected on the swabbings/wipings taken from [Applicant’s] right palm are ‘below threshold’ meaning that under SWIFS standards, both then and now[,] the results would be reported as negative for [gunshot residue].” Hall explained that she did not testify about and was not asked about SWIFS standards. Instead, her testimony focused on the FBI standards. She stated that she has “not found any published materials as to what the FBI threshold levels were at the time,” but she does have the handwritten notes relied upon at trial; they were from a 1993 gunshot primer residue course at the FBI Academy. 13 In her final paragraph, Hall asserted, “I would not testify now, as I did back then, that the levels of antimony and barium detected on the swab taken from [Applicant’s] right palm were ‘unique’ to gunshot residue.” On August 29, 2010, Applicant was administered a polygraph examination. He answered “No” to the questions “Did you shoot two men sitting in a car at a Texaco Station?” and “Did you shoot two men in their car in May 1994?” The test results indicated no deception, and the professional opinion reported that Applicant’s answers were considered truthful. 14 The results were confirmed by two DPD po-lygraphers, who reviewed the polygram. *663 On October 7, 2010, Applicant filed a supplemental memorandum. He claimed that the State violated Brady by failing to disclose that “trace amounts of antimony and barium found on Applicant’s hand as a result of the gunshot residue test did not meet the Southwest Institute of Forensic Sciences (SWIFS) standards for gunshot residue determination.” On the next day, October 8, James Hammond, a Texas Peace Officer employed as an investigator in the Criminal District Attorney’s Office of Dallas County with the task of reinvestigating Applicant’s case, signed an affidavit stating that AFIS identified the source of a previously unknown fingerprint found on the victims’ car. The fingerprint at issue was found during the initial investigation by DPD officers in a location on the victims’ car where a shooter might have left prints if he were shooting down into the car as per witnesses’ recounts. Hammond stated that a DPD homicide detective interviewed the source of the now-known fingerprint. The individual acknowledged that he “lived approximately one half mile from the scene of the crime in the early 1990s,” “he frequented the club next to the location of the offense,” and “in the early 1990s he owned a white Cadillac convertible, the same type of car described as the get-away car used by the assailant.” Additionally, the detective asked the individual to take a polygraph, and the DPD polygrapher reported that “he failed the polygraph, as to his involvement in the crime.” On October 11, 2010, the habeas court adopted a third set of findings of fact and conclusions of law agreed to by Applicant and the State, responding to our second remand order regarding Applicant’s actual innocence claim. The trial court found that Applicant’s actual innocence claim was not proeedurally barred by Section 4 because Hall’s affidavit, the results of Applicant’s polygraph examination, and the identification of the before-unknown fingerprint were new or previously unavailable factual bases. The court also concluded that Applicant had established that he is actually innocent. Finally, the court concluded that the gunshot-residue standards, as testified to at trial, are no longer reliable, so “Applicant should also be granted relief, independently, on the ground of flawed forensic testimony.” In a post-conviction review of a writ of habeas corpus, this Court is the ultimate fact finder. We are not bound by the findings and conclusions of the convicting court, but we generally defer to such if they are supported by the record. See, e.g., Ex parte Chabot, 300 S.W.3d 768, 772 (Tex.Crim.App.2009); Ex parte Thompson, 153 S.W.3d 416, 417-18 (Tex.Crim.App. 2005). VII. SECTION FOUR Article 11.07, Section 4, restricts habeas applicants to “one bite of the apple.” Tex.Code CRim. Proc. art. 11.07, § 4(a); Ex parte Santana, 227 S.W.3d 700, 703 (Tex.Crim.App.2007). It provides that this Court cannot consider the merits of a subsequent application unless it contains sufficient specific facts establishing that (1) the current claims and issues have not been and could not have been presented previously in an original application or in a previously considered application filed under this article because the factual or legal basis for the claim was unavailable on the date the applicant filed the previous application; or (2) by a preponderance of the evidence, but for a violation of the United States Constitution no rational juror could have found the applicant guilty beyond a reasonable doubt. Tex.Code Crim. Proc. art. 11.07, § 4(a). A factual basis of a claim is “unavailable” under Subsection (a)(1) “if the factual basis *664 was not ascertainable through the exercise of reasonable diligence on or before that date,” 15 and reasonable diligence “suggests at least some kind of inquiry has been made into the matter of the issue.” Ex parte Lemke, 13 S.W.3d 791, 794 (Tex. Crim.App.2000). We agree with the habeas court that Applicant’s actual innocence claim meets the statutory requirements of Section 4(a)(1). That claim relies on the “new” evidence of the two undisclosed police reports (the anonymous phone call report and the Garland report), Thurman’s recantation of his in-court identification of Applicant as the shooter, the identification of the source of the previously unknown fingerprint, and Hall’s affidavit stating that her trial testimony was incorrect, in addition to the facts discovered during the investigation of the “new” evidence. The record supports that the factual bases for the claim were unavailable and not ascertainable through the exercise of reasonable diligence on or before October 13, 2005, the date Applicant filed his first applications. First, although defense counsel filed a Brady motion prior to trial and Applicant’s father filed a FIA request in November 2004, not until information was received from the DPD pursuant to Centurion Ministries’ March 2007 FIA request did Applicant become aware of the two additional police reports. Second, Thurman’s affidavit recanting his in-court identification was not signed until January 6, 2010. 16 Third, the source of the previously unknown fingerprint was not identified until some time after the filing of Applicant’s initial writ applications, as indicated by Hammond’s October 8, 2010 affidavit. Finally, Hall’s affidavit was not signed until July 27, 2010. 17 Accordingly, because Applicant’s actual innocence claim fulfills the requirements of Section 4(a)(1), it is not procedurally barred. For the same reasons, we also agree with the habeas court that Applicant’s Brady claim, which relies on the two undisclosed police reports, and Applicant’s gunshot-residue-analysis claim, which relies on Hall’s affidavit, meet the requirements of Section 4(a)(1). Hence, those claims are not procedurally barred. Turning to the merits of the writ applications, we note that we are presented with a hybrid case, in which the facts and standards underlying Applicant’s claims are so intertwined that we cannot fully analyze one without the benefit of the other. At a minimum, Applicant is entitled to a new trial pursuant to Brady; however, when we combine the Brady evidence with other newly discovered evidence and view it in light of the habeas court’s findings and conclusions, it is apparent that Applicant is entitled to relief on actual innocence grounds. Therefore, because of the intertwined nature of Applicant’s claims, we must initially discuss the Brady violation to fully understand Applicant’s actual innocence claim. 18 VIII. BRADY As Applicant argues, “[t]he State failed to produce police reports which iden *665 tified other potential suspects for the shooting. These police reports should have been provided pursuant to Brady.” Specifically, the evidence at issue includes the two undisclosed reports: the anonymous phone call report and the Garland report. 19 The Supreme Court in Brady v. Maryland held “that the. suppression by the prosecution of evidence favorable to an accused upon request violates due process where the evidence is material either to guilt or to punishment, irrespective of the good faith or bad faith of the prosecution.” Brady, 373 U.S. at 87, 83 S.Ct. 1194. To establish a claim under Brady, a habeas applicant must demonstrate that (1) the State failed to disclose evidence, regardless of the prosecution’s good or bad faith; (2) the withheld evidence is favorable to him; [and] (3) the evidence is material, that is, there is a reasonable probability that had the evidence been disclosed, the outcome of the trial would have been different. Hampton v. State, 86 S.W.3d 603, 612 (Tex.Crim.App.2002). Additionally, we require that the evidence central to the Brady claim be admissible in court. Ex parte Kimes, 872 S.W.2d 700, 703 (Tex.Crim. App.1993). A. The State failed to disclose the evidence Applicant must initially show that the State failed to disclose evidence “which had been known to the prosecution but unknown to the defense.” United, States v. Agurs, 427 U.S. 97, 103, 96 S.Ct. 2392, 49 L.Ed.2d 342 (1976). Even if the prosecutor was not personally aware of the evidence, the State is not relieved of its duty to disclose because “the State” includes, in addition to the prosecutor, other lawyers and employees in his office and members of law enforcement connected to the investigation and prosecution of the case. Kyles v. Whitley, 514 U.S. 419, 437, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995); Ex parte Reed, 271 S.W.3d 698, 726 (Tex. Crim.App.2008) Because the DPD had possession of the reports, the State was responsible for their disclosure. However, those reports were not present in the State’s file nor were they produced pursuant to defense counsel’s pretrial requests for production or the FIA request filed by Applicant’s father. It was not until the records were obtained from the DPD via Centurion Ministries’ FIA request that Applicant became aware of the documents. B. Evidence withheld is favorable to Applicant Applicant must also demonstrate that the evidence withheld by the State is “favorable” to his case. Favorable evidence is that which, if disclosed and used effectively, “may make the difference between conviction and acquittal.” United States v. Bagley, 473 U.S. 667, 676, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985). Favorable evidence includes exculpatory evidence and impeachment evidence. Exculpatory evidence is that which may justify, excuse, or clear the defendant from fault, and impeachment evidence is that which disputes, disparages, denies, or contradicts other evidence. Harm v. State, 183 S.W.3d 403, 408 (Tex.Crim.App.2006); Thomas v. State, 841 S.W.2d 399, 404 (Tex. Crim.App.1992). The two undisclosed police reports are exculpatory and could have constituted im *666 peachment evidence within the purview of Brady. In its arguments and questioning, the State emphasized the absence of any other suspect or theory for the crime. The police reports identified other potential suspects for the crime (i.e., Richard, Deuce, and Garland). Moreover, subsequent investigation of those allegations could have led to other exculpatory evidence. As defense counsel Gray noted in his affidavit, the reports “would have allowed [him], at a minimum, to develop an alternate theory for the shooting that occurred at the Texaco station.” This may have made the difference between conviction and acquittal. The police reports could also be employed as impeachment evidence in response to the testimony of Detective Hooker and victim Johnson. At trial, Hooker testified that there were no other potential suspects or possible theories that would explain the shooting, other than the altercation at Redbird Mall testified to by Johnson. The police reports could have negated Hooker’s statements because they identify other possible suspects (i.e., Richard, Deuce, and Garland), and the Garland report describes a confrontation between the two victims and a third party, allowing for a possible motive for the shooting by someone other than Applicant. With respect to victim Johnson, his testimony portrayed him as a law-abiding citizen who kept a sawed-off shotgun for personal protection. He denied that he and Williams were having trouble with anyone around the time of the shooting. The Garland report disputes Johnson’s testimony because, according to it, just five days before the Texaco shooting, Johnson and Williams had been involved in a dispute with Garland in which Johnson pointed the sawed-off shotgun at Garland and threatened to shoot him. C. The evidence is material Next, Applicant must show that the undisclosed, favorable evidence is “material” to guilt or punishment. “The mere possibility that an item of undisclosed information might have helped the defense, or might have affected the outcome of the trial, does not establish ‘materiality’ in the constitutional sense.” Agurs, 427 U.S. at 109-10, 96 S.Ct. 2392. Hence, Applicant must show that, “in light of all the evidence, it is reasonably probable that the outcome of the trial would have been different had the prosecutor made a timely disclosure.” Hampton, 86 S.W.3d at 612; see Smith v. Cain, 565 U.S. -, 132 S.Ct. 627, 630, 181 L.Ed.2d 571 (2012); Bagley, 473 U.S. at 682, 105 S.Ct. 3375. “A ‘reasonable probability’ of a different result is accordingly shown when the Government’s evidentiary suppression ‘undermines confidence in the outcome of the trial.’ ” Kyles v. Whitley, 514 U.S. 419, 434, 115 S.Ct. 1555, 131 L.Ed.2d 490 (1995). When evaluating whether the materiality standard is satisfied, the strength of the exculpatory evidence is balanced against the evidence supporting conviction. Hampton, 86 S.W.3d at 613; see Smith, 132 S.Ct. at 630-31. The suppressed evidence is considered collectively, rather than item-by-item. Kyles, 514 U.S. at 436, 115 S.Ct. 1555. Here, the two police reports are material to Applicant’s case. The probability of a different outcome had the information been timely disclosed to the defense is sufficient to undermine confidence in the outcome of the case. Applicant asserts that the undisclosed reports could have led to further investigation of other suspects and theories for the shooting, which he could have brought to light at trial. In the anonymous phone call report, the caller knew significant information about the shooting. She stated that *667 her ex-boyfriend, Richard, told her that he had shot two black males at a Texaco near Bachman Lake. She described that a nine-millimeter gun had been used and insisted that the wrong person had been arrested for the offense. The subsequent investigation of “Richard” in fact revealed that he had been at the New York, New York night club with his cousin on the very night of the Texaco shooting, the same club where the victims had been. It also showed that Richard’s height, body type, and complexion are consistent with the description of the shooter. Henry Evans, the only witness to provide a physical description of the shooter, stated that the shooter was 6'2" to 6'4", slim, and well built (like he “worked out”). Richard is 6'6", and he told McCloskey of Centurion Ministries that he was a “ ‘great’ basketball player who could play with the best of them,” even playing against the players of Navarro College and Notre Dame. Richard’s physique resembles that of the shooter much more closely than Applicant’s 5'9", 190-pound frame. Moreover, all of the witnesses described the shooter’s complexion as being very dark, and according to McCloskey, Richard’s complexion is dark, much darker than Applicant’s. Several witnesses (including Johnson, Evans, and Knight) 20 indicated that the shooter’s skin tone was darker than that of Applicant. The investigation also revealed that Richard has a violent and aggressive history. Richard’s record includes three arrests and convictions for the unlawful carrying of a weapon, and just three weeks prior to the Texaco shooting, Richard assaulted his ex-girlfriend, hitting her repeatedly and pointing a nine-millimeter gun at her. He also told McCloskey that he had been a drug dealer who made a lot of money. This can be contrasted with Applicant’s history — although he has a misdemeanor conviction for theft under $200 and a felony possession of cocaine, Applicant received deferred adjudication sentences, and notably, there is no evidence of Applicant possessing a weapon or committing assaultive or violent conduct. The disclosure of all of this information to the jury could have significantly undermined the confidence in the State’s case. Further, Applicant argues that the undisclosed reports would have allowed him to attack the State’s argument that no other suspects or theories for the shooting existed. As mentioned previously, Hooker denied that the criminal investigation revealed any other potential suspects or theories to explain the shooting (besides the confrontation with the white Cadillac at Redbird Mall), and in its closing argument, the State emphasized, “If nothing else — I mean, there’s no other suspects.” However, with the introduction of the two reports, Applicant could have attacked those statements. He could have pointed to Richard, Deuce, and even Garland as potential suspects. In addition, he could have developed possible alternate theories for the crime based upon both the victims’ altercation with Garland and Richard’s history as a drug dealer and his physical characteristics which were consistent with the description of the shooter. The significance of the police reports becomes even more obvious when considered in the context of the trial record, including the questionable gunshot-residue analysis and relevant testimony, the suggestive eyewitness identifications, the disparities between the descriptions of the shooter and Applicant, and Applicant’s alibi. *668 The gunshot residue was the only physical evidence purportedly connecting Applicant to the crime scene, but Hall’s trial testimony supporting that connection was of limited value as it was highly qualified. For example, she testified that she found elevated levels of antimony and barium while also conceding that the levels were very low. In addition, she recognized that there have been cases of known non-shooters who had presumptive levels of residue on both hands, and she stated that the existence of residue does not necessarily mean that the individual fired the gun because he could have encountered it from other sources. Yet, she still testified that the residue on Applicant’s palm may not be from those sources, and the residue ratios were unique to gunshot residue. Similarly, Thurman was the only witness out of twenty to thirty individuals who could identify Applicant as the shooter. From a distance of about twenty feet, Thurman viewed the shooter’s face as he ran by — the shooter was wearing a “floppy hat” and Thurman’s view lasted for only a second or two. Subsequently, Thurman identified Applicant twice before trial, but those identifications are arguably suggestive in nature. See Tillman v. State, 354 S.W.3d 425 (Tex.Crim.App.2011). Thurman’s first identification occurred when he observed Applicant in the squad car, as it returned to the crime scene. Applicant was removed from the car while handcuffed and surrounded by police officers. This situation essentially resulted in a one-on-one showup. Although Officer Lund-berg testified that they parked in the back of the Texaco so that Applicant would not be observed, Thurman did observe him. Thurman’s second identification occurred when he viewed a photographic lineup and chose Applicant’s photograph as the shooter. Thurman repeatedly said that the shooter was wearing a white tank top, and Applicant was the only individual in the photo lineup wearing a white tank top. See New Jersey v. Henderson, 208 N.J. 208, 27 A.3d 872, 897-98 (2011) (stating that mistaken identifications are more likely to occur when the suspect stands out from the other members of a photographic lineup). This viewing occurred within hours of Thurman observing Applicant in the squad car. It is unclear whether Thurman was recalling the shooter’s image from the time of the offense or Applicant’s image from the previous identification. The other eyewitnesses, some closer in distance and some with longer viewing periods, were later shown the photo lineup, and not one of them could select a photograph. Unlike Thurman, these witnesses had not seen Applicant taken out of a squad car while handcuffed and surrounded by officers. In addition, the many disparities between the descriptions of the shooter and Applicant weigh in favor of Applicant. Although trial testimony showed that the witnesses consistently stated that the shooter wore shorts, Applicant was wearing long blue Dickie pants when he was arrested. This fact even gave Detective Hooker pause as he admitted that the thought that he had “the wrong guy ... had crossed his mind.” Similarly, the witnesses had varying descriptions of the shooter’s shirt. 21 Further, as mentioned previously, several witnesses indicated that Applicant’s skin tone was significantly lighter than that of the shooter. And the only witness to testify to the shooter’s physical structure stated that the shooter was at least 6'2" to 6'4" with a slim, athlet *669 ic build while Applicant is 5'9" and 190 pounds. Additionally, witnesses stated that the shooter was carrying the gun in his right hand, but Applicant is left-handed. Finally, Applicant provided a detailed alibi that was corroborated by the trial testimony of Yarborough and Clark. Although the State attempted to discredit the alibi, Detective Hooker testified that it “[s]omewhat, yes, it did” check out, and Applicant testified that an officer told him that his story “added up.” One discrepancy emphasized by the State was Yarbor-ough’s mistaken belief that Miles called around 2:00 a.m., rather than 3:00 a.m. However, we note that Yarborough had been asleep when Miles called, and he was estimating the time based on the television show that he had been watching when he went to sleep. Yarborough heard the helicopter overhead within three minutes of that phone call, and it is undisputed that a police helicopter was flying over the area once Officer Lundberg had radioed that they had spotted the suspect around 3:00 a.m. We also note that McCloskey’s time line, which reflected the events of the evening, rationally supports Applicant’s contention that he was walking from the location where he exited Clark’s car to Yarborough’s house when he was arrested. Overall, the two police reports are material to Applicant’s case. D. The evidence would be admissible Finally, Applicant must show that the evidence would have been admissible at trial. The State does not have a duty to disclose favorable, material evidence if it would be inadmissible in court. Kimes, 872 S.W.2d at 703. At a minimum, the undisclosed reports would have been admissible as impeachment evidence. 22 See Tex.R. Evid. 607. Generally, a defendant complaining of the nondisclosure of information that he believes could have been used for impeachment must make clear the legal basis on which the evidence would be admissible to impeach a specific witness. Ex parte Richardson, 70 S.W.3d 865, 871-72 (Tex. Crim.App.2002); Kimes, 872 S.W.2d at 703. In Richardson, we held that a police officer’s diary was exculpatory evidence that should have been disclosed to the defendant. Richardson, 70 S.W.3d at 871-72. The diary could have led the defense to testimony of the officer that, in her opinion, the State’s key eyewitness was not a truthful person. The police officer had served on the witness’s security detail. She testified that she maintained the diary of the time spent guarding the witness because she believed the witness was not a truthful person and hoped to protect herself from false accusations. In addition, the diary identified other officers in the security detail, so its disclosure could have also led to five other police officers who could testify similarly. Id. at 872. In contrast, in Kimes, we held that the nondisclosure of police reports and accompanying affidavits did not violate Brady because the record failed to establish that they could have been used to impeach a witness by showing that she was “lying.” Kimes, 872 S.W.2d at 703-04. *670 Here, the record is clear that the undisclosed reports could have been used to discredit the accuracy of Hooker’s “no other suspect” testimony and Johnson’s “no other conflicts” testimony, which were critical aspects of the State’s case. Consequently, like the diary in Richardson, the reports would have been admissible for impeachment purposes against Johnson and Hooker. E. Article 39.14 does not exempt evidence from disclosure Pursuant to Article 39.14 of the Code of Criminal Procedure, offense reports and investigative reports prepared by the police are protected from discovery as work product of the police. Tex. Code Crim. Proo. art 39.14(a); Washington v. State, 856 S.W.2d 184, 187 (Tex. Crim.App.1993); Brem v. State, 571 S.W.2d 314, 322 (Tex.Crim.App.1978). However, the privilege derived from the work-product doctrine is not absolute, and the duty to reveal material exculpatory evidence as dictated by Brady overrides the work-product privilege. See also Hampton, 86 S.W.3d at 612 (discussing that the State has a duty to disclose police reports containing material exculpatory information); Thomas v. State, 837 S.W.2d 106, 113-14 (Tex.Crim.App.1992) (“Denial of access to information which would have a reasonable probability of affecting the outcome of a defendant’s trial abridges a defendant’s due process rights.... ”). While Article 39.14 “makes it clear that the decision on what is discoverable is committed to the discretion of the trial court,” the trial court must permit discovery if “the evidence sought is material to the [djefense of the accused.” Quinones v. State, 592 S.W.2d 933, 940-41 (Tex.Crim.App.1980). The materiality standard for purposes of Article 39.14 is the same as that applied in our Brady analysis above. Id. at 941 (citing Agurs, 427 U.S. at 108-10, 96 S.Ct. 2392). Therefore, because the two undisclosed reports were material to Applicant’s defense, Article 39.14 does not exempt the reports from discovery. Because the two undisclosed reports contain favorable evidence material to Applicant’s case and the State failed to disclose such evidence, the State violated Applicant’s constitutional right as expressed in Brady. At a minimum, Applicant is entitled to a new trial. But the evidence in this case requires us to take a step further, and with a clear understanding of the significant implications of the Brady evidence, we now turn to Applicant’s actual innocence claim. IX. ACTUAL INNOCENCE Applicant contends that he “is factually innocent of the charges for which he was convicted.” 23 A. Caselaw A claim of actual innocence is cognizable in a post-conviction habeas corpus proceeding. Ex parte Elizondo, 947 S.W.2d 202, 205 (Tex.Crim.App.1996). Two types of actual innocence claims may be raised. Schlup v. Delo, 513 U.S. 298, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995); Herrera v. Collins, 506 U.S. 390, 113 S.Ct. 853, 122 L.Ed.2d 203 (1993). A Herrera-type claim is a substantive claim in which the applicant asserts a “bare claim of innocence based solely on newly discovered evidence.” Ex parte Franklin, 72 S.W.3d 671, 675 (Tex.Crim.App.2002). In contrast, a Schlup-type claim is a procedural claim in which the applicant’s claim of *671 innocence “‘does not by itself provide a basis for relief,’ but is intertwined with constitutional error that renders a person’s conviction constitutionally invalid.” Ex parte Brown, 205 S.W.3d 538, 544-45 (Tex.Crim.App.2006); see Schlup, 513 U.S. at 315, 115 S.Ct. 851. While Applicant raises procedural claims attacking his conviction (i.e., his Brady claim), he asserts a Herrera-type innocence claim, which requires him to show by clear and convincing evidence that no reasonable juror would have convicted him in light of the newly discovered evidence. Elizondo, 947 S.W.2d at 210. Establishing a Herrera-type claim is a Herculean task. Brown, 205 S.W.3d at 545. Because it attacks the conviction directly, the applicant must make “an exceedingly persuasive case that he is actually innocent.” Elizondo, 947 S.W.2d at 206. Accordingly, the applicant must “unquestionably establish” his innocence by proving “by clear and convincing evidence that no reasonable juror would have convicted the applicant in light of the new evidence.” Id. at 209. A reviewing court must assess “the probable impact of the newly available evidence upon the persuasiveness of the State’s case as a whole, [so] we must necessarily weigh such exculpatory evidence against the evidence of guilt adduced at trial.” Id. at 206; see Thompson, 153 S.W.3d at 417. B. Application Applicant presents a bare innocence claim. Applicant must show that the evidence he is presenting is “newly discovered” or “newly available” and that such is affirmative evidence of his innocence. Ex parte Spencer, 337 S.W.3d 869, 878 (Tex. Crim.App.2011); Franklin, 72 S.W.3d at 678. Only when that is satisfied do we “proceed with a determination of whether the applicant can prove by clear and convincing evidence[ 24 ] that no reasonable juror would have convicted him in light of the newly discovered evidence.” Franklin, 72 S.W.3d at 678. While this case differs from many of our actual innocence cases in which we rely on a single piece of evidence (e.g., DNA evidence or the recantation of the only victim or witness), we believe that the multiple pieces of newly discovered evidence presented here (including the Brady evidence) amount to affirmative evidence that unquestionably establishes Applicant’s innocence. Initially we note that the “new” evidence upon which Applicant relies is in fact “newly discovered” or “newly available.” Such terms have been defined as evidence that was not known to the applicant at the time of trial, plea, or post-trial motions and could not be known to him even with the exercise of due diligence. Brown, 205 S.W.3d at 545. As discussed thoroughly in the context of Article 11.07, Section 4, supra, the two undisclosed police reports (the anonymous phone call report and the Garland report), Thurman’s recantation of his in-court identification of Applicant as the shooter, and the identification of the source of the previously unknown fingerprint were factual bases that were unavailable and not ascertainable through the exercise of reasonable diligence on or before the date Applicant filed his first writ applications. For the same reasons, the evidence also satisfies the “newly discovered” or “newly available” standard. *672 We further hold that we are convinced by clear and convincing evidence that no rational jury would have convicted Applicant in light of the new evidence. Applicant’s actual innocence claim relies primarily on the Brady evidence (the two undisclosed police reports) and the resulting investigation. In the anonymous phone call report, the caller related significant information about the shooting including the approximate date, the location, the number of victims, the victims’ race and gender, and the caliber of gun used, and she identified her ex-boyfriend, Richard, as the alleged shooter. Subsequent investigation of Richard placed him at the New York, New York night club at the same time as the victims, on the very night of the shooting. An investigation also revealed that Richard’s height, body type, and complexion are consistent with the description of the shooter. Richard was a drug dealer who had a violent and aggressive history, and he had used a nine-millimeter gun during prior assaultive conduct. Similarly, the Garland report shows that, just five days before the Texaco shooting, the victims had been involved in a dispute with a third party in which threats were made and a sawed-off shotgun was displayed by victim Johnson. Another possible suspect for the shooting (Deuce) was identified in the Garland report. As discussed previously, this Brady evidence alone undermines the confidence in the outcome of the trial. When combined with the other newly discovered evidence (ie., Thurman’s recantation, Hall’s affidavit, and the recently identified source of the previously unknown fingerprint), Applicant’s claim of innocence is affirmatively supported. Thurman was the only eyewitness to identify Applicant as the shooter, and he has now recanted his (and the only) in-court identification. He asserted that he did not recognize Applicant or remember him as the shooter and that Applicant did not look like the shooter. Although the trial court made no findings regarding the credibility (or lack thereof) of Thurman’s recantation, at a minimum the recantation undermines the value of his earlier in-court testimony. Hall testified that the palm of Applicant’s right hand tested positive for gunshot residue. However, she would now testify differently. Instead, she would report the test results as negative for gunshot residue under SWIFS standards. A fingerprint was found in a location on the victims’ car consistent with where a shooter might have left prints if he were shooting down into the car as per witnesses’ recounts. Although the identity of that individual has not been disclosed, the source of the previously unknown fingerprint has been identified. The habeas court described the individual identified as having “a violent history.” At the time of the shooting, that individual lived half of a mile from the Texaco station, and he frequented the New York, New York club where the victims had been on the night of the shooting. Moreover, the individual identified owned a white Cadillac convertible, the same type of car described as the get-away car used by the Texaco shooter. It is also the same type of car driven by a number of black males with whom Johnson had had a confrontation at Redbird Mall earlier on the day of the Texaco shooting, an incident emphasized by the defense in its closing arguments as a motive for the shooting. The individual identified denied any involvement in or knowledge of the Texaco shooting, but he failed a polygraph administered by the DPD polygrapher “as to his involvement in the crime.” The newly discovered evidence is of heightened importance when viewed in the context of the entire record. It is particu *673 larly clear that this new evidence would have weighed in favor of an acquittal of Applicant when it is balanced against all of the evidence presented at trial and the facts uncovered through post-trial investigation. After all, the State’s case relied on the gunshot-residue analysis, Thurman’s identifications, the consistencies between the descriptions of the shooter and Applicant, and the fact that there were no other suspects; each piece of that evidence is questionable or has since been undermined or completely invalidated. The gunshot residue was the only physical evidence purportedly connecting Applicant to the crime scene. However, Hall has now changed her position, stating that she would have testified that the hand washings were negative for gunshot residue. She now firmly asserts that the levels of antimony and barium detected were so low that they are below the threshold levels and would be reported as negative for gunshot residue under SWIFS standards. Thus, this physical evidence no longer supports Applicant’s connection to the crime. Similarly, Thurman was the only witness out of twenty to thirty individuals who could identify Applicant as the shooter, and he has now recanted his (and the only) in-court identification. We recognize that Thurman also identified Applicant as the shooter on two other occasions, but those identifications are arguably unreliable for the reasons stated above. Moreover, there are great disparities between the descriptions of the shooter and Applicant. At the time of his arrest, Applicant was wearing long blue Dickie pants and a white tank top that had faded writing on it. This contrasts with the witnesses’ descriptions of the shooter as wearing shorts and a “T-shirt,” without mention of writing. This fact even caused Detective Hooker to admit that the thought that he had “the wrong guy ... had crossed his mind.” Further, several witnesses stated that Applicant’s skin tone was significantly lighter than that of the shooter. And Applicant’s 5'9", 190-pound frame contrasts with the shooter’s physique, which was described as 6'2" to 6'4" with a slim, athletic build. Also, Applicant is left-handed, but the shooter was carrying the gun with his right hand. In addition, the State emphasized the absence of any other suspect or theory for the crime, but such an argument is contrary to the evidence as it now exists. To illustrate, both the anonymous phone call report and the Garland report identified other possible suspects (ie., Richard, Deuce, and Garland); the source of the previously unknown fingerprint could also be considered a potential suspect. Possible alternate theories for the shooting could have been developed as well, based upon both the victims’ altercation with Garland and Richard’s history as a drug dealer. Finally, it cannot be ignored that Applicant provided a detailed alibi that was corroborated by the trial testimony of Yar-borough and Clark. Although the State attempted to discredit the alibi, Detective Hooker testified that it “[s]omewhat, yes, it did” check out, and Applicant testified that an officer told him that his story “added up.” Moreover, McCloskey’s time line, which reflected the events of the evening, is consistent with Applicant’s contention that he was walking from the location where he exited Clark’s car to Yarbor-ough’s house when he was arrested. When we balance the newly available evidence (including the Brady evidence) with other exculpatory evidence and the evidence of guilt presented at trial, we are satisfied that Applicant has shown by clear and convincing evidence that no rational jury would convict him in light of the new evidence. *674 X. CONCLUSION The habeas court concluded that Applicant is entitled to actual innocence relief. We have reviewed the record, and we agree that Applicant has established that he is actually innocent. Habeas corpus relief is granted. The judgments in cause numbers W94-54687-S(B) and W94-54688-S(B) in the 282nd Judicial District Court of Dallas County are set aside, and Applicant is remanded to the custody of the Sheriff of Dallas County to answer the charges as set out in the indictments. Copies of this opinion shall be sent to the Texas Department of Criminal Justice-Correctional Institutions Division and Pardons and Paroles Division. MEYERS and PRICE, JJ., not participating. Footnotes [1] A Dallas County medical examiner, Dr. Charles Odom, testified that he performed the autopsy on Williams. There were a number of gunshot wounds caused by bullets consistent with a nine-millimeter bullet, and the injuries sustained were consistent with Williams being seated in the passenger seat and the shots coming from the driver's side of the car. [2] Dodge was not instructed to or aware of the need to check the bushy area behind the Texaco for evidence. [3] The next day, Hooker returned to the scene to search for a nine-millimeter weapon in the area directly behind the Texaco where the shooter had been seen, but he was not able to find a weapon after looking for an hour. [4] The individual in Exhibit 23 also wore a tank top, but it was not white. [5] Detective Hooker showed the photo lineup to Johnson while he was in the hospital. The first time, he was on medication and heavily sedated. Hooker waited several days, and when he returned again, Johnson was unable to identify anyone. [6] When asked about the particular reason for testing for antimony and barium, Hall explained, "The reason is that antimony and barium is contained in the primer and it's unique. It’s not as common in the environment as some of these nitrates that the defense attorney has referred to, so it’s more a unique combination or a unique chemical that would be found.” [7] Clark testified that he was supposed to be at his girlfriend’s home by 3:00 a.m. He dropped Applicant off within a "ten-minute frame” of that hour. [8] A call placed to CAPERS is done so without suggestion of a reward or other benefit to the caller. [9] We also note that handwritten notes on a copy of the incident report relating to the shooting of Johnson describe the Garland confrontation. [10] Garland was murdered on October 29, 1998. Per the murder investigative report, his body was found in a parked car in a residential area. He sustained multiple gunshot wounds from a nine-millimeter handgun. [11] Springer also discussed the progress of research focusing on contamination issues. However, those studies relied on analyses by SEM/EDS, rather than by AAS (the process used in this case). Those “two analytical techniques are fundamentally different.... Therefore, an evaluation of the likelihood of law enforcement contamination of gunshot residue samples analyzed by AAS would have to be studied using AAS as the analytical technique.” [12] The Dallas County District Attorney’s Office conducted a post-conviction investigation to determine if Applicant was innocent, and as part of its investigation, it interviewed and obtained this affidavit from Thurman. [13] We note that Hall has not recanted her trial testimony related to the FBI standards. [14] Because polygraph exams are not admissible evidence, we do not rely on these results as evidence of Applicant’s innocence. [15] Tex.Code Crim. Proc. art. 11.07, § 4(c). [16] Nothing in the record indicates that Thurman would have recanted earlier. [17] We note that Springer’s “Review of Gunshot Residue Analysis by Atomic Absorption,” also indicating that Hall's trial testimony was incorrect, was dated August 29, 2009. [18] We recognize that addressing both the Brady and actual innocence claims may result in some repetitiveness, but a thorough discussion of the impact of the Brady evidence is necessary to fully comprehend Applicant’s actual innocence claim. [19] The convicting court recommended, and the State does not oppose, that we grant relief based upon Applicant’s Brady claim. [20] We note that the eyewitnesses were African-American, so this is not a situation involving cross-ethnic descriptions and identifications. Hence, the fact that the witnesses were insistent that the shooter "was dark” is significant. [21] The shooter’s "floppy hat” was a key identifier included in Tolleson’s broadcast and witnesses' descriptions, and we find it significant that the hat worn by Applicant at the time of his arrest was lost and not introduced into evidence for the jury's observation and comparison. [22] We note that the undisclosed reports could have also led to other admissible evidence favorable to Applicant. While the State usually does not have a duty to turn over inadmissible evidence, the analysis might not end there. The Fifth Circuit has held that, if inadmissible evidence would give rise to the discovery of other admissible evidence or witnesses, the State does have a duty to disclose that evidence. United States v. Brown, 650 F.3d 581, 588 (5th Cir.2011); United States v. Sipe, 388 F.3d 471, 485 (5th Cir.2004); Sellers v. Estelle, 651 F.2d 1074, 1077 n. 6 (5th Cir. 1981). [23] The convicting court recommended, and the State does not oppose, that we grant relief based upon Applicant's Brady claim. [24] Clear and convincing evidence is defined as that "degree of proof which will produce in the mind of the trier of fact a firm belief or conviction as to the truth of the allegations sought to be established.” Young v. State, 648 S.W.2d 2, 3 (Tex.Crim.App.1983) (quoting State v. Addington, 588 S.W.2d 569, 570 (Tex. 1979)); see Lackey v. State, 819 S.W.2d 111, 117 (Tex.Crim.App.1989).