Following a jury trial, Appellant Timothy Boothe was found guilty of malice murder and other offenses in connection with the death of Geneva Strickland.1 Appellant contends, among other things, *286that the trial court erred by admitting into evidence photocopies of police sketches of two men that were based on descriptions from one of the State’s witnesses. We conclude that, even assuming that the admission of the sketch copies violated the “best evidence” rule of Georgia’s old Evidence Code, see former OCGA § 24-5-4 (a),2 any error was harmless. Appellant’s other enumerations of error lack merit. Accordingly, we affirm his convictions.
1. Viewed in the light most favorable to the jury’s verdict, the evidence presented at trial showed as follows. On October 31, 2007, Clayton County police officers and firefighters responded to a fire at the home of Geneva Strickland around 11:00 p.m. After the fire was extinguished, police found Strickland dead in one of her bedrooms. Her wrists and legs had been bound with the kind of long plastic zip ties ordinarily used to secure ductwork or plumbing pipes, and her mouth was covered with an Ace bandage that had been wrapped around her head seven times. When the medical examiner unwrapped the Ace bandage, he found, in the fifth turning of the wraps, a blue latex glove.3 Strickland died from carbon monoxide poisoning coupled with suffocation and affixial restraint.
Earlier that night, Torie Gertsch, who was riding her bike in front of Strickland’s house, smelled smoke coming from the area and saw a white man and a black man hanging around the house. When the two men saw Gertsch, they chased her, but she was able to escape on her bicycle. The next day, Gertsch described the two men to GBI agents and a GBI sketch artist, who drew a pencil sketch of each man. At trial, the State did not account for the original sketches; over Appellant’s objection, the trial court admitted photocopies of the sketches into evidence.
*287DNA testing identified Appellant’s nuclear DNA (nDNA) inside the blue latex glove that was found wrapped in the Ace bandage covering Strickland’s mouth.4 The only other DNA evidence found on the glove was a partial profile consistent with the victim’s DNA. In addition, a firefighter found a black hoodie-type mask in the front yard of Strickland’s home and placed the mask on a utility box, where the GBI recovered it. One testable strand of head hair was found on the mask. Microscopic analysis indicated that the hair came from Appellant or someone whose hair possessed the same microscopic characteristics, and mitochondrial DNA (mDNA) testing of the hair showed that it matched Appellant’s mDNA. Evidence also showed that Appellant had previously worked as a handyman at the victim’s house and knew that she kept large amounts of cash there.
After receiving the results of the DNA testing of the glove, police officers secured an arrest warrant for Appellant and went to his mother’s house to try to locate him. They had to knock on the door for an extended time before anyone answered. Appellant’s mother eventually opened the door, and she and his sister told the officers that no one else was home. When the officers searched the house, however, they found Appellant hiding in the attic, which could be accessed only by pull-down stairs. The police told him to come down, but Appellant instead stood in the attic opening hiding one of his hands like he had a gun and yelling at the officers to shoot him. He surrendered after a short standoff.
Viewed in the light most favorable to the verdict, the evidence presented at trial and summarized above was sufficient to authorize a rational jury to find Appellant guilty beyond a reasonable doubt of the crimes of which he was convicted and sentenced. See Jackson v. Virginia, 443 U. S. 307, 319 (99 SCt 2781, 61 LE2d 560) (1979). See also Vega v. State, 285 Ga. 32, 33 (673 SE2d 223) (2009) (“ £It was for the jury to determine the credibility of the witnesses and to resolve any conflicts or inconsistencies in the evidence.’ ” (citation omitted)).
2. Appellant argues that the trial court erred by admitting into evidence photocopies of the two police sketches drawn from the descriptions provided by Gertsch. He contends that the admission of copies of the sketches, as opposed to the original sketches, violated the “best evidence” rule set forth in former OCGA § 24-5-4 (a). We need not decide if this contention is correct, because it is clear that any error regarding the admission of the sketch copies was harmless.
*288(a) Georgia’s archaic “best evidence” rule stated: “The best evidence which exists of a writing sought to be proved shall be produced, unless its absence shall be satisfactorily accounted for.” Former OCGA § 24-5-4 (a). Under that provision, “when the contents of a writing are material, the original of the writing must be produced or its absence accounted for” before the writing could be admitted into evidence. Clark v. State, 271 Ga. 6, 11 (515 SE2d 155) (1999). Only when the original writing was unavailable for some reason other than the fault of the proponent did the trial court have discretion to admit a duplicate or copy of the original.5 See Merrill Lynch, Pierce, Fenner & Smith v. Zimmerman, 248 Ga. 580, 580 (285 SE2d 181) (1981). Here, the “writings” at issue were two pencil sketches, and it is undisputed that the State did nothing to account for the whereabouts of the original sketches when it offered the copies of them at trial.
Former OCGA § 24-5-4 (a) did not define the term “writing,” and the State argues that the sketches do not constitute “writings” for purposes of the old best evidence rule.6 This appears to be a question of first impression for Georgia’s appellate courts. In cases where former OCGA § 24-5-4 (a) has been applied, the “writing” appears to have been a document containing words. See, e.g., Norris v. State, 289 Ga. 154, 158 (709 SE2d 792) (2011) (applying the best evidence rule to a letter); Baptiste v. State, 288 Ga. 653, 655-656 (706 SE2d 442) (2011) (applying the rule to search warrants and affidavits). See also Merrill Lynch, 248 Ga. at 581 (explaining that the old best evidence rule “recognize [d] the central position writings have in the law. In many instances significant rights turn on the precise language used in a writing. When that is the case, the law prefers the writing itself to be produced.”). Photographs were not considered writings. See Smith v. State, 236 Ga. 5, 8 (222 SE2d 357) (1976) (holding that the best evidence rule did not apply to a photograph of a police lineup). *289And the probative value of a police sketch is not in any language or words, but rather, like a picture, in the image depicted.
On the other hand, a sketch, like a handwritten document, is produced by a writing implement and is based on the artist’s subjective interpretation of information supplied to the artist as she draws on the page. And as Appellant points out, a California court has concluded that a police sketch was a writing for purposes of that state’s best evidence rule, although the statute there defined the term “writing” expansively. See People v. Garcia, 201 Cal. App. 3d 324, 328 n. 1 (1988).7
Thus, whether a police pencil sketch is a “writing” under the old best evidence rule is a close question. And we need not decide that question to decide this case, because even assuming that the sketch copies were inadmissible under former OCGA § 24-5-4 (a), any error in admitting them was clearly harmless.
(b) In determining if an error is harmless, we “review the record de novo.” Arizona v. Fulminante, 499 U. S. 279, 295 (111 SCt 1246, 113 LE2d 302) (1991). In doing so, we weigh the evidence as we would expect reasonable jurors to have done so, as opposed to assuming that they took the most pro-guilt possible view of every bit of evidence in the case.8 Applying the test for determining whether a non-constitutional, evidentiary error was harmless, we have no difficulty concluding that it is “highly probable” that any error regarding the *290admission of the police sketch copies did not contribute to the jury’s guilty verdict in this trial. Lindsey v. State, 282 Ga. 447, 450 (651 SE2d 66) (2007).9
Review of the trial record reveals that the sketches of the two men Gertsch saw in the vicinity of the crime scene were not important inculpatory evidence at trial. The two pencil sketches are in the record. One depicts a black man with close-cut hair; the other is of a white man with long, straight sandy-brown hair and no mustache. Also admitted into evidence at trial was Appellant’s September 21, 2007, driver’s license, the photograph on which shows that Appellant is a white man who, in the time period of the crimes, had long dark brown or black hair, a thick mustache, and a beard. Thus, one of the sketches that may have been admitted erroneously — the one of the black potential perpetrator — was entirely exculpatory of Appellant; indeed, the record offers no other evidence of this man’s identity or of a second person being involved in the crimes. And comparing the other sketch to Appellant’s license photo, the jury easily could have considered that sketch to be exculpatory as well; at best it served only to not exclude Appellant as a potential perpetrator, to the extent he is a white man who had long hair.10
Moreover, the probative value of that sketch depended entirely on the accuracy of Gertsch’s description to the sketch artist, and the sketch did not gain any credence from Gertsch’s testimony at trial. To the contrary, her testimony was inconsistent and heavily impeached; like the sketches, Gertsch at best did not entirely exclude Appellant as a potential perpetrator. Gertsch testified that one of the men she saw around (but not entering or leaving) the victim’s house that night (a night she repeatedly described as “very dark”) was African-American — so he was not Appellant. As for the second man she saw, he was white and had long hair, like Appellant, but she said the man’s hair was dark blonde and admitted that she had said earlier that it was red. Gertsch also testified that the white man had a mustache — *291but then she testified, and she told the GBI before trial, that he did not have a mustache (as the sketch depicts).11 Before trial, Gertsch told the GBI that the white man was 5'6" to 5' 10" tall, but at trial she specified that the man was about three to four inches taller than her five-foot-three-inch height (that is, 5'6" to 5'7"); Appellant’s driver’s license shows him to be six feet tall and 185 pounds. Gertsch said at one point that the white man was wearing a mask, but she also testified that the black man was the one with the mask, and she told the GBI before trial that she did not see a mask that night. Gertsch did not identify Appellant at trial as one of the men she had seen near the victim’s house; she testified only that the copy of the sketch looked like the sketch that was made when she gave her description to the artist.
Thus, the sketches, as well as the witness who provided their content, did little if anything to identify Appellant as a perpetrator of the crimes. So why did the jury find him guilty? Because the other evidence was so powerful.
To begin with, the State’s evidence showed that Appellant had a prior relationship with the victim and a motive to commit the crimes; he had worked at her house and knew that she kept a significant amount of cash there. The State then proved beyond any reasonable doubt, primarily through solid forensic evidence, that Appellant was directly involved in the crimes.
Evidence showed that during the month before the crimes, Appellant worked for a construction company that sometimes stocked long zip ties of the size used to bind the victim’s wrists and legs. Much more significantly, the nuclear DNA testing of the blue latex glove, which was found stuck in the fifth turning of the Ace bandage wrapped around the victim’s mouth, showed that Appellant’s DNA was inside the glove, tying him conclusively to an implement used in committing the murder. The only other DNA found on the glove was a partial profile consistent with the victim’s DNA. Thus, the evidence indicated that Appellant — and no one else — wore a glove used while muzzling the victim before she was killed. And as with the zip ties, there was evidence that Appellant had access to such gloves. Bobby Joe Taylor, Appellant’s friend and sometimes work partner, testified that Appellant used latex gloves to mix paint and clean paint brushes.
*292The dissent seeks to discount the import of this DNA evidence by-suggesting that Appellant’s DNA might have been left on the glove when he did repair and painting work at the victim’s house. But full consideration of the evidence shows that to be an unrealistic possibility. Appellant’s sister, who testified for him, said that he had worked at the victim’s house six or seven times, but the last time was in either late 2006 or early 2007. Taylor also testified that Appellant had done work at the house on several occasions, but he dated that work to 2005-2006 or before. Thus, the evidence was that Appellant had not been in the victim’s house for many months before the October 31, 2007 crimes.12 In addition, a picture of the blue latex glove was admitted into evidence. It is a surgical-style latex glove, the type of thin, flimsy glove that is normally used once and discarded, not kept and reused. Moreover, the glove is pristine, with no sign that it had been used for anything like mixing or cleaning up paint.
Thus, to conclude that Appellant did not leave his unique nDNA on the glove while using it in gagging and murdering the victim, one would have to believe that Appellant’s DNA was left inside the glove when he used it to clean or mix paint or do other such work, but without leaving a mark on it; and that the glove had not been discarded after Appellant used it but instead remained in the house for an intruder to find many months later; and that Appellant’s DNA remained in the glove those many months; and that the intruder entered the house, found the glove, and decided to use it while binding the victim with an Ace bandage, without leaving his own DNA on the glove, only Appellant’s DNAand a partial DNAprofile consistent with the victim’s. No rational juror would accept that as a reasonable explanation for how Appellant’s nDNA came to be on the glove.
But the defense had to do more than just explain away the glove containing Appellant’s unique nuclear DNA. On the black mask discarded in the victim’s front yard was a head hair that was subjected not only to microscopic comparison to Appellant’s hair but *293also to mitochondrial DNA testing. Although the results of this testing are not as conclusive as nuclear DNA testing (nothing is), the State’s experts explained that not only was the hair found on the mask microscopically indistinguishable from Appellant’s hair, but the chances of the mitochondrial DNA from a Caucasian person like Appellant matching the mDNAin the hair was 1 in 588. That is a far more definitive identification than a pencil sketch based on a shaky witness’s recollection of two men she saw on a dark night.
The dissent also suggests that the probative value of the hair is undermined because the mask might have been used by another intruder and then contaminated, before it was collected by the GBI, by a hair that Appellant shed on the ground or on the utility box when he worked at the house. But that would require the jury to have believed that Appellant’s hair remained in the victim’s yard for many months after he last worked at the house in early 2007, in quantities sufficient that the mask happened to pick up a hair after being discarded. And the jury also would have needed to ignore the GBI expert’s testimony that a hair left outside for a long time would show signs of weathering, and there were no such signs on the hair found on the mask. And of course the hair mDNA evidence stands alongside the glove nDNA evidence, requiring a theory that someone other than Appellant was fortunate enough not only to have the mask he wore (without leaving any of his own hair on it) come into contact with a hair that Appellant shed many months before but also to find and use (without leaving any of his own nDNA) a latex glove that Appellant had used many months before. There is no reason to believe that the jury relied on such an unreasonable theory. And finally, Appellant’s action of hiding from and engaging in a standoff with police officers when they came to arrest him was evidence of his consciousness of guilt.
To sum up, even assuming that copies of the police sketches of the two men Torie Gertsch saw outside the victim’s house were erroneously admitted in violation of the old best evidence rule, there is no doubt that such error did not contribute to the jury’s decision to find Appellant guilty. One sketch was entirely exculpatory, and the other resembled Appellant only in that he is a white male who had long hair — even though his hair color, mustache, and height appear inconsistent with the description Gertsch gave. By contrast, the other evidence supporting the jury’s verdict was overwhelming, in particular the DNA evidence linking Appellant to both the latex glove found in the Ace bandage wrapped around the victim’s mouth and the black mask found in her yard just after the murder. The “main witness” in this case was not Gertsch or the sketches based on her descriptions, Dis. Op. at 297; the main witness against Appellant was the biological *294evidence containing his DNA that he left at the crime scene. It was clearly that evidence, not the sketch, that directly identified Appellant as a perpetrator of the crimes. Accordingly, we conclude that it is “highly probable” that any error in admitting the sketches did not contribute to the guilty verdict. See Lindsey, 282 Ga. at 450.
3. Appellant argues that the trial court erred by prohibiting him from questioning Torie Gertsch about her prior drug use. Appellant does not contend that at trial he had evidence of any prior drug convictions of Gertsch that he could have used for impeachment purposes under former OCGA § 24-9-84.1. Instead, he contends only that Gertsch’s prior drug use could have affected her memory on the night that she saw the two men near Strickland’s home. At the motion for new trial hearing, Appellant offered expert testimony that long-term cocaine use may cause brain damage and affect memory, but no such evidence was proffered at trial.
In the absence of such evidence, the trial court allowed Appellant to question Gertsch only about whether she was under the influence of any drugs at the time that she saw the two men and at the time that she gave her testimony at trial — the only time periods that were relevant to Gertsch’s testimony. To the extent that Appellant wished to question Gertsch about drug use in general, the trial court did not abuse its discretion in determining that such inquiries were irrelevant and improper. See, e.g., Garcia v. State, 240 Ga. 796, 801 (242 SE2d 588) (1978) (holding that the trial court did not abuse its discretion in limiting the cross-examination of a State witness’s use of drugs to the time of the crime, as opposed to his general abuse of drugs); Lancette v. State, 151 Ga. App. 740 (4) (261 SE2d 405) (1979) (holding that the trial court properly sustained an objection to the defendant’s asking a witness on cross-examination, “Now, do you use marijuana?,” because that issue was irrelevant and the defendant could not, “under the guise of attacking the witness’ credibility, ask questions suggesting illegal or immoral conduct in areas other than that before the court”); former OCGA § 24-9-62 (“It shall be the right of a witness to be examined only as to relevant matter and to be protected from improper questions . . . .”).13 And in any event, as discussed in Division 2 above, Gertsch was heavily impeached in other ways, and her testimony was not very useful to the State’s case, so any error regarding her potential further impeachment with prior drug use was harmless. See Lindsey, 282 Ga. at 450.
*2954. Finally, Appellant contends that his trial counsel provided constitutionally ineffective assistance of counsel by failing to discover that Gertsch had four prior felony convictions that could have been used to impeach her. To prevail on this claim, Appellant must show that his trial counsel’s performance was professionally deficient and that, but for the deficiency, there is a reasonable probability that the outcome of the trial would have been more favorable to him. See Strickland v. Washington, 466 U. S. 668, 687, 694 (104 SCt 2052, 80 LE2d 674) (1984). “This burden, although not impossible to carry, is a heavy one.” Young v. State, 292 Ga. 443, 445 (738 SE2d 575) (2013). And the reviewing court need not “address both components of the inquiry if the defendant makes an insufficient showing on one.” Strickland, 466 U. S. at 697.
Appellant asserts that Gertsch had four prior felony convictions that trial counsel should have discovered, but only one such conviction (for criminal damage to property) was admitted into evidence at the motion for new trial hearing. And even assuming that trial counsel performed deficiently in failing to discover that conviction,14 considering Gertsch’s weakness as a witness and the strength of the other evidence against Appellant, we conclude that, had trial counsel discovered that one conviction and used it to further impeach Gertsch, there is no reasonable probability that the outcome of the trial would have been different.
Judgment affirmed.
All the Justices concur, except Benham and Melton, JJ., who dissent.The crimes occurred on October 31, 2007. On April 30, 2008, Appellant was indicted by a Clayton County grand jury for malice murder; three counts of felony murder (with burglary, robbery, and arson as the underlying felonies); three counts of burglary; and one count each of robbery, aggravated assault, arson, false imprisonment, and kidnapping. Before his first trial *286began, a nolle prosequi was entered on the felony murder and burglary counts based on robbery and the robbery count. At his first trial on January 31-February 7,20Í1, the trial court directed a verdict in Appellant’s favor on the kidnapping count, and the jury acquitted him of aggravated assault and arson. The jury could not reach a verdict on the remaining counts, however, and a mistrial was declared. Following a July 5-8, 2011, retrial, Appellant was found guilty of all remaining charges: malice murder, felony murder based on burglary, two counts of burglary, and false imprisonment. The trial court sentenced Appellant to life in prison for malice murder and ten consecutive years for false imprisonment. The burglary charges merged into the malice murder count for sentencing purposes, and the felony murder count was vacated by operation of law. See Malcolm v. State, 263 Ga. 369, 373 (434 SE2d 479) (1993). Appellant filed a motion for new trial on July 11, 2011, which he amended twice with the assistance of new counsel, and which was denied on April 27, 2012. Appellant then filed a timely notice of appeal, and his appeal was docketed in this Court for the January 2013 term and submitted for decision on the briefs.
Because this case was tried before January 1, 2013, our new Evidence Code does not apply. See Ga. L. 2011, pp. 99, 214, § 101.
The glove was not itself used to bind the victim. Instead, it may have come off the perpetrator’s hand as he wrapped the Ace bandage repeatedly around the victim’s face.
The nuclear DNA expert noted that the only other human being this DNA could have come from was an identical sibling of Appellant’s, hut there is no suggestion that he has an identical twin.
Under our new Evidence Code, which does not use the term “best evidence,” the presumption that the original document must be offered instead of a duplicate is reversed, in accordance with the corresponding Federal Rule of Evidence:
A duplicate shall be admissible to the same extent as an original unless:
(1) A genuine question is raised as to the authenticity of the original; or
(2) A circumstance exists where it would be unfair to admit the duplicate in lieu of the original.
OCGA § 24-10-1003. See Fed. R. Evid. 1003.
The new Evidence Code defines the term “writing” as follows: “A ‘writing’ or ‘recording’ means letters, words, or numbers, or their equivalent, set down by handwriting, typewriting, printing, photostating, magnetic impulse, or mechanical or electronic recording or other form of data compilation.” OCGA § 24-10-1001 (1). Unlike the old rule, the new rule applies not only to “writings”but also to “recordings” and “photographs,” which are defined in OCGA § 24-10-1001 (2). See OCGA § 24-10-1002.
The court relied on Cal. Evid. Code § 250, which says:
“Writing” means handwriting, typewriting, printing, photostating, photographing, photocopying, transmitting by electronic mail or facsimile, and every other means of recording upon any tangible thing, any form of communication or representation, including letters, words, pictures, sounds, or symbols, or combinations thereof, and any record thereby created, regardless of the manner in which the record has been stored.
Thus, harmless-error analysis differs from the analysis done to determine if the evidence presented at trial was sufficient to support a conviction as a matter of due process, which requires the reviewing court to view all the evidence in the light most favorable to the jury’s verdict, see Jackson v. Virginia, 443 U. S. at 319. See Dixon v. State, 173 Ga. App. 280, 282 (325 SE2d 893) (1985) (“On appeal we do not weigh the evidence, but afford the jury verdict every available presumption of correctness.... But where error is committed in the trial of the case in the form of the admission of prejudicial evidence, the verdict cannot carry the same presumption of validity, but the evidence is examined to determine whether the legal evidence is so strong or overwhelming that it is highly probable the illegal evidence did not contribute to the verdict.” (citations omitted)). See also Al-Qaadir v. Gallegos, 56 F3d 70, 1995 WL 330628, *3 n. 5 (9th Cir. 1995) (unpublished) (“This standard [for analyzing sufficiency of the evidence] should not be used in conjunction with harmless error review. It is impossible to determine whether an error was harmless beyond a reasonable doubt by construing evidence in the light most favorable to the prosecution. Such construction gives the benefit of the doubt to the prosecution. The two standards are mutually exclusive.”).
Contrary to the dissent’s view, see Dis. Op. at 297, n. 19, harmless error must be assessed based on the review of the record of the trial whose verdict is challenged on appeal, not another jury’s inability to reach a verdict based on the evidence, arguments, jury instructions, and other proceedings in a different trial. Thus, we have not compared the record of the previous trial to the record of the trial at issue here, and the dissent gives no indication that it has either. We note that the jury in this trial did not indicate any difficulty in reaching the guilty verdict that is the subject of this appeal.
The dissent says that “the jury would consider the sketch of a Caucasian male seen on a dark night to be another important link connecting [Appellant] to the crime.” Dis. Op. at 298. As just discussed, a charitable view of the sketch would make it a link between Appellant and the area near the crime scene, but the assertion that evidence that merely puts a person of the same race and gender as the defendant near a crime scene is importantly probative is untenable.
It is of course possible that Appellant shaved his mustache and beard in the month between his driver’s license photo and the crimes, see Dis. Op. at 298; it is also possible that he dyed his hair a lighter color. But there is no evidence or suggestion of that in the record; the driver’s licence photo is the only evidence the jury had as to Appellant’s appearance around the time of the crimes, and thus it is relevant in considering how the jury reasonably viewed the sketch.
The dissent asserts that “the record does not show definitively that [Appellant] ceased working at the victim’s home after [the 2005-2006] time period and did not work at her home closer to the date of the murder.” Dis. Op. at 299, n. 21. That is true to the extent that Appellant’s sister’s testimony placed him at the house as late as “the first part” of 2007, hut neither she nor Taylor testified or suggested that Appellant had ever heen to the house after that. As the dissent indicates, Taylor did discuss “other times” that Appellant had gone to the house with him to work on the victim’s car or do painting and remodeling work, but in context the “other times” referred to times other than during the one painting project in 2005-2006 that the State had asked Taylor about on direct examination. Taylor indicated that the other occasions were before or during the 2005-2006 time frame and, when specifically asked if he knew whether Appellant “continued to work [there] after [Taylor] left,” he responded, “I don’t know.” In sum, nothing but unfounded speculation puts Appellant in the victim’s house between early 2007 and the night of the murder many months later.
OCGA § 24-6-611 (a) (3) of Georgia’s new Evidence Code says, “The court shall exercise reasonable control over the mode and order of interrogating witnesses and presenting evidence so as to . . . (3) Protect witnesses from harassment or undue embarrassment.”
During discovery, the State’s witness list mistakenly gave as Gertsch’s date of birth the birth date of her mother. Trial counsel used that incorrect birth date in seeking criminal history information regarding Gertsch. Another document provided in discovery showed Gertsch’s and her mother’s correct birth dates.