*19In this DUI case, the State appeals from an order of the Bibb County State Court finding that a law enforcement officer lacked probable cause to arrest Antonio Mikeal-Austin Culler and granting Culler's motion to suppress the evidence resulting from that arrest. The State contends that the trial court erred when, in assessing probable cause, it: (1) failed to consider the totality of the circumstances; and (2) gave no weight to the results of a field sobriety test administered to Culler. For reasons explained more fully below, we vacate the trial court's ruling and remand for further proceedings consistent with this opinion.
In a hearing on a motion to suppress, the trial court sits as the trier of fact and its findings are analogous to a jury verdict. Watts v. State , 334 Ga. App. 770, 771, 780 S.E.2d 431 (2015). Accordingly, we defer to the trial court's credibility determinations and will not disturb its factual findings in the *437absence of clear error. Id. And "[w]hen reviewing the grant or denial of a motion to suppress, an appellate court must construe the evidentiary record in the light most favorable to the trial court's factual findings and judgment." Caffee v. State , 303 Ga. 557, 557, 814 S.E.2d 386 (2018). Additionally, as a general rule, appellate courts must limit their "consideration of the disputed facts to those expressly found by the trial court." Id. (citation and punctuation omitted). See also Hughes v. State , 296 Ga. 744, 746 (1), 770 S.E.2d 636 (2015) (on an appeal from the grant or denial of a motion to suppress, appellate courts must "focus on the facts found by the trial court in its order") (citation, punctuation, and emphasis omitted). "An appellate court may, however, consider facts that definitively can be ascertained exclusively by reference to evidence that is uncontradicted and presents no questions of credibility, such as facts indisputably discernible from a videotape." Caffee , 303 Ga. at 559 (1), 814 S.E.2d 386 (citation and punctuation omitted). Finally, although we defer to the trial court's fact-finding, we owe no deference to the trial court's legal conclusions. Hughes , 296 Ga. at 750 (2), 770 S.E.2d 636. Instead, we independently apply the law to the facts as found by the trial court. Id.
Viewed in the light most favorable to the trial court's judgment, the record shows that at approximately 2:30 a.m. on November 25, 2017, Officer Thomas Burns of the Georgia State Patrol observed a Nissan Murano driving in downtown Macon without its headlights illuminated. Based on the lack of headlights, Burns conducted a traffic stop of the vehicle, which was driven by Culler. The traffic stop, in turn, led to Culler's arrest for DUI. Following his arrest, and after *20being read Georgia's implied consent notice, Culler agreed to provide a breath sample, which was "positive."1
Culler was charged by accusation with a single count each of DUI per se,2 DUI less safe,3 and driving without headlights. Prior to trial, Culler filed a motion to suppress the results of his breath test, asserting that the officer lacked probable cause to arrest him and therefore the evidence was obtained illegally. The trial court held a hearing on the motion to suppress, at which the State presented the testimony of the arresting officer and the video recording of the traffic stop.4
The video shows that once the officer initiated the traffic stop, Culler responded by immediately pulling the car off of the roadway and into what appeared to be an adjacent parking lot, out of the way of traffic. Culler exited the vehicle and waited for the officer to approach him. Culler provided Burns with his driver's license and responded to Burns's questions. According to Burns, during this exchange, the officer noticed that Culler's eyes appeared bloodshot and watery and his speech was slurred. Burns also detected a "strong odor" of alcohol emanating from Culler's person. Based on these observations, Burns asked Culler how much alcohol he had consumed that evening. Culler initially responded, "not that much," and then clarified that he had consumed 2 to 3 mixed drinks over the course of the evening, and that he had last consumed alcohol approximately 40 minutes earlier. Burns then asked Culler if he would be willing to submit to field sobriety tests, and Culler agreed. The *438officer performed three field sobriety tests on Culler: the walk-and-turn, the one-leg stand, and the horizontal gaze nystagmus ("HGN").
According to Burns, he performed these standardized tests in accordance with his training.5 Two of the tests (the HGN and the *21one-leg stand) were performed in view of the patrol car's camera and were captured on the video recording of the traffic stop. Burns explained, however, that because of uneven pavement in front of the patrol car, he had Culler perform the walk-and-turn test in a different part of the parking area, out of the camera's range.6 Burns testified that Culler tested positive on six out of six clues (three in each eye) on the HGN test and on one clue on the walk-and-turn test.7 Both Burns's testimony and the video of the traffic stop show Culler exhibited no clues on the one-leg stand test. Indeed, the video shows that Culler was able to stand straight, arms at his side, with one leg raised, and while counting out loud for approximately 23 seconds, during which time he exhibited no problems with his balance, coordination, or speech. He ceased performing the test only when Burns directed him to do so.
After Culler completed the field sobriety tests, Burns used the Alco-Sensor, which gave a positive reading for the presence of alcohol on Culler's breath. Burns then placed Culler under arrest for DUI. Burns testified that he relied on six facts to find probable cause to arrest Culler. These facts included: the odor of alcohol coming from Culler's person; Culler's bloodshot and watery eyes; Culler's slurred speech; the fact that Culler was positive for six out of six clues on the HGN test; the fact that Culler was positive on one clue on the walk-and-turn test; and the results of the Alco-Sensor, showing the presence of alcohol on Culler's breath.
On cross-examination, Burns acknowledged that Culler was cooperative, communicated clearly, and was easy to understand throughout the traffic stop. Burns further testified that Culler responded appropriately to all questions and directions and exhibited no problems with his balance. Additionally, Burns stated that he did not observe Culler driving erratically, and that the traffic stop was based solely on the lack of headlights, and not on Culler's operation of the vehicle.
Based on this evidence and its independent review of the video recording of the traffic stop, the trial court granted the motion to suppress. In doing so the court acknowledged the six facts on which Burns had relied to find probable cause for Culler's arrest, as well as the evidence showing that Culler was operating the car without *22headlights and that he had admitted consuming alcohol approximately 40 minutes before driving the car. The trial court found that this evidence provided the officer with a reasonable basis for believing that Culler was operating the car after having consumed alcohol, but noted that the relevant question for probable cause purposes was whether Burns had a reasonable basis for believing that Culler was impaired.
In conducting its probable cause analysis, the trial court declined to give any weight to Culler's failure to turn on his headlights, noting that the video showed that the streets were well lit, such that the absence of headlights would not have been readily apparent to a driver. Additionally, the court rejected Burns's testimony that Culler was slurring his speech, explaining that the video showed the exact opposite. Specifically, the court found that the video showed that Culler "spoke very clearly," "asked appropriate questions to clarify" the officer's directions to him, and "provided prompt, clear responses when questioned." The trial court further *439found that Culler exhibited no outward signs of being an impaired driver, observing that Culler operated the vehicle safely, using "his blinker for both turns shown in the video." The court also noted that Culler "stopped promptly when blue[-] lighted and parked in a safe area off the street. He stepped out of his vehicle immediately. He did not sway or stumble. He was cooperative and appropriately responsive in all his interactions with [the officer]."
The trial court also rejected the results of the HGN test, stating that "[t]he video shows that [the officer] did not perform the HGN test substantially in accordance with scientific procedures."8 Thus, the court concluded that the only credible evidence that Culler was impaired were the presence of alcohol and Culler's one positive clue on the walk-and-turn test. And in light of the fact that Culler exhibited no clues on the one-leg stand test and that he otherwise did not appear to be impaired, the court found that the evidence, viewed as a whole, did not support a reasonable probability that Culler was driving while impaired. The State now appeals from the trial court's order.
Before addressing the State's enumerated errors, we note the general legal principles that apply to the question of whether Burns *23had probable cause to arrest Culler for DUI less safe.9 Probable cause for an arrest exists where the objective facts known to the officer, taken together with the circumstances in which those facts exist, establish a probability that the suspect is or has been engaged in criminal activity. Campbell v. State , 313 Ga. App. 436, 438, 721 S.E.2d 649 (2011). See also Caffee , 303 Ga. at 561 (2), 814 S.E.2d 386 ("[a] probable cause inquiry ... is a flexible and practical assessment of probabilities given a particular factual context") (citation and punctuation omitted). A probability is "less than a certainty but more than a mere suspicion or possibility." Jaffray v. State , 306 Ga. App. 469, 473 (3), 702 S.E.2d 742 (2010). The standard is an objective one, meaning that the relevant inquiry is whether the known facts and circumstances would have led a reasonable officer to conclude that the suspect was engaged in criminal activity. Hughes , 296 Ga. at 749 (2), 770 S.E.2d 636 (because the standard "is an objective one ... the subjective thinking of the actual officers in a particular case is not important"). And in undertaking a probable cause analysis, a court must consider the facts and circumstances "altogether, for it is the totality of those facts and circumstances that matters, not any one fact or circumstance standing alone." Id. (Citation omitted). See also Campbell , 313 Ga. App. at 438, 721 S.E.2d 649 ("[p]robable cause need not be defined in relation to any one particular element, but may exist because of the totality of circumstances surrounding a transaction") (citation and punctuation omitted). Furthermore, "where the totality of the facts and circumstances known to an officer would permit reasonable officers to draw differing conclusions about whether the suspect probably has committed a crime, probable cause exists, and it is for the officer - not judges, trial or appellate - to decide which of the several reasonable conclusions to draw." Hughes , 296 Ga. at 749 (2), 770 S.E.2d 636.
Bearing the foregoing legal principles and the applicable standard of review in mind, we turn to the State's claims of error.
1. Probable cause to arrest an individual for DUI less safe exists where the facts and circumstances known to the officer would support a reasonable belief that "the suspect was actually in physical control *24of a moving vehicle, while under the influence of alcohol to a degree which render[ed] him incapable of driving safely ." *440Jaffray , 306 Ga. App. at 473 (3), 702 S.E.2d 742 (citation, punctuation, and footnote omitted; emphasis supplied). "Mere presence of alcohol is not the issue[,] ... [and] [a]ccordingly, the mere fact that a suspect admits to having consumed alcohol before driving does not provide the probable cause necessary to support an arrest for DUI [less safe]." Bostic v. State , 332 Ga. App. 604, 607, 774 S.E.2d 175 (2015) (citations and punctuation omitted).
According to the State, in finding the evidence did not support the existence of a reasonable probability that Culler was an impaired driver, the trial court failed to consider the "totality" of the facts and circumstances. Instead, the State contends, the trial court committed legal error by assessing each fact or circumstance individually, and in isolation. The State further contends that the trial court erred by failing to consider the presence of alcohol on Culler's breath, Culler's admission that he had consumed alcohol, and Culler's bloodshot and watery eyes in determining the existence of probable cause. And these failures, the State asserts, led the trial court to conclude erroneously that no probable cause existed for Culler's arrest. We disagree, as these arguments ignore the fact that the trial court chose to assign no weight to much of the evidence on which Burns relied for probable cause. Nor do the State's arguments acknowledge the evidence that would support the trial court's conclusion - i.e., the evidence that would lead a reasonable person to conclude that Culler was not impaired.
The trial court's order reflects that in assessing probable cause, the court considered all of the facts and circumstances on which Burns claimed he based his decision to arrest Culler. The trial court also considered that Culler was operating his vehicle without headlights and admitted to consuming alcohol before driving. The court then found the evidence as to several of those facts was either not credible or not reliable. Specifically, the court explicitly rejected Burns's testimony that Culler was slurring his speech and implicitly rejected Burns's testimony concerning Culler's bloodshot and watery eyes. See Hughes , 296 Ga. at 747 (1), 770 S.E.2d 636 (appellate courts "must presume that the absence of a finding of a fact that would tend to undermine the conclusion of the trial court reflects a considered choice to reject the evidence offered to prove that fact, especially where there were grounds upon which the trial court properly could have assigned no weight to such evidence") (emphasis omitted). Additionally, the trial court gave no weight to the fact that Culler's headlights were not illuminated. Nor did the trial court give any weight to the results of the HGN test, based on its finding that at least a portion the test was not administered properly. The court's weighing of the evidence *25and assessing the credibility of the witnesses is clearly within the province of the court, as the trier of fact, and therefore did not constitute legal error. See Tate v. State , 264 Ga. 53, 56 (3), 440 S.E.2d 646 (1994) ("[c]redibility of witnesses and the weight to be given their testimony is a decision-making power that lies solely with the trier of fact"). See also State v. Damato , 302 Ga. App. 181, 183 (1), 690 S.E.2d 478 (2010) (as the trier of fact on a motion to suppress, a trial court is free to reject an officer's testimony, even where that testimony is uncontradicted).
Based on its determination as to the weight and credibility to be assigned the evidence, the trial court concluded that the only facts which could have provided Burns with probable cause for Culler's arrest were the presence of alcohol, Culler's admission that he had consumed alcohol, and Culler's positive result on one out of eight clues on the walk-and-turn test. The court then considered these facts in conjunction with the additional facts known to Burns that would lead a reasonable person to believe Culler was not impaired. These additional circumstances included the fact that Culler exhibited no outward signs of impairment, completed the one-leg stand test with the ease, and was observed driving the car in a safe manner. Finally, the trial court found that all of these facts, considered as a whole, failed to establish a probability that Culler was driving while impaired. Thus, the record shows that, despite the State's assertion to the contrary, the trial court properly considered the totality of the circumstances in assessing probable cause.
*441Finally, in light of the facts as found by the trial court, we cannot say that the trial court erred as a matter of law in concluding that Burns lacked probable cause to arrest Culler for DUI less safe. See Bostic , 332 Ga. App. at 607, 774 S.E.2d 175 (no probable cause existed to support an arrest for DUI less safe where, although the defendant admitted drinking, alcohol was present on his breath, and his eyes were bloodshot and watery, the traffic stop was not premised on his unsafe operation of the vehicle and the video of the traffic stop showed that "his gait was steady and otherwise normal"); Ojemuyiwa v. State , 285 Ga. App. 617, 620 (2), 647 S.E.2d 598 (2007) (although two officers detected "a strong odor of alcohol" coming from the defendant's person, that fact did not support an arrest for DUI less safe where "there was no evidence presented whatsoever as to [the defendant's] manner of driving, that her speech was slurred, her gait unsteady, or her eyes red and glassy"); Shaheed v. State , 270 Ga. App. 709, 710 (1), 607 S.E.2d 897 (2004) (probable cause did not support an arrest for DUI less safe where the facts showed that the traffic stop was not predicated on a traffic violation, there was a significant odor of alcohol coming from the driver's person, and he admitted to drinking *26and refused to submit to either field sobriety tests or chemical testing of his blood). As discussed in Division 2, however, we cannot determine, based on the current record, whether the trial court erred in finding that none of the results of the HGN test were reliable. On remand, should at least some of those results be found reliable, that new factual finding would impact the probable cause analysis.
2. For the results of an HGN test10 to have any evidentiary value, the State must prove "that the person performing the test substantially performed the [same] in an acceptable manner" - i.e., in such a manner that the results are reliable. Walsh v. State , 303 Ga. 276, 280, 811 S.E.2d 353 (2018). The trial court found the results of the HGN test performed on Culler "to be unreliable" because the video showed that Burns did not perform a portion of the test within the applicable guidelines. The court stated: "Specifically, in relation to the 'lack of smooth pursuit' portion of the test, Trooper Burns performed the passes significantly more quickly than the required two seconds out and two seconds back, which may induce nystagmus."
On appeal, the State challenges the trial court's findings that the smooth pursuit component of the HGN test was not administered according to the applicable guidelines and that this failure could produce a false positive result. It also challenges the trial court's refusal to give any weight to those portions of the test that were performed correctly.11
(a) The State argues that no evidence supported the trial court's finding that the relevant guidelines require that an officer perform the passes involved in the lack of smooth pursuit portion of the HGN test at a pace of "two seconds in and two seconds out." We agree with the State that the record contains no testimony or other evidence regarding the proper administration of an HGN test. With respect to this issue, however, the State concedes in its brief that the relevant guidelines for the smooth pursuit component require an officer "to move the stimulus 'steadily at a speed that takes approximately *27two seconds to bring the eye from the center to the side.' " Thus, it appears *442that the trial court and the State are in agreement about the guidelines governing the time element of the smooth pursuit component.12 Unfortunately, given the lack of testimony concerning HGN tests and their proper administration, we cannot readily determine from the video recording whether Burns complied with the guidelines at issue. Put another way, this is not a question of fact that is "indisputably discernible from" the video recording. Caffee , 303 Ga. at 559 (1), 814 S.E.2d 386.
The State also asserts that no evidence of record supports the trial court's finding that taking less than two seconds to perform the smooth pursuit component could induce nystagmus. We agree that there was no testimony on this issue. We note, however, that it is "the responsibility of the trier of fact ... to draw reasonable inferences" from the facts. Kiser v. State , 327 Ga. App. 17, 18, 755 S.E.2d 505 (2014) Accordingly, in an appeal from an order granting or denying a motion to suppress, "the standard of review require[s] [an appellate court] to defer to" any such inferences the trial court draws. Mayberry v. State , 312 Ga. App. 510, 512, 718 S.E.2d 822 (2011). And here, the trial court's finding on the issue of a possible false positive result is a logical inference to draw from the undisputed fact that guidelines exist for administering the HGN test. Those guidelines exist for a reason, and as a matter of logic, at least part of that reason would be to ensure the reliability of the test results.
Again, however, the lack of testimony about the specifics of the HGN test and the proper technique for performing the same prevents us from determining whether the trial court properly inferred that a failure to administer the test in accordance with the prescribed guidelines could trigger a false positive result. Given this fact, and given that we cannot discern from the record whether Burns administered the test properly, we vacate the trial court's order and remand for further proceedings on these questions. See Williams v. State , 301 Ga. 60, 62, 799 S.E.2d 779 (2017) (vacating an order granting a motion to suppress and remanding to the trial court "for further clarification on the specific findings that form the basis for its legal conclusions"); Stroud v. State , 344 Ga. App. 827, 834 (3), 812 S.E.2d 309 (2018) (reversing denial of motion to suppress and remanding for determination of factual questions necessary to decide that motion).
*28(b) The State further contends that even assuming that Burns failed to administer the smooth pursuit component of the test within the scientific guidelines, the trial court erred in concluding that the positive results on the other two evaluative components were unreliable. To support this argument, the State cites to the fact that an officer can conclude that a suspect is impaired if the suspect has a positive result on four out of the six clues tested for on the HGN. See Parker v. State , 307 Ga. App. 61, 64 (2), 704 S.E.2d 438 (2010). ("[u]nder law enforcement guidelines, a score of four out of six clues on an HGN test constitutes evidence of impairment"). Thus, the State argues that even assuming one portion of Culler's HGN test was tainted, the trial court erred in refusing to give any weight to the remaining four positive results. Given the current record, however, we cannot determine whether the trial court erred in finding that no part of the HGN test was reliable. It could be that the trial court concluded that, based on Culler's appearance, demeanor, and conduct as shown on the video recording, none of Burns's testimony on this issue was credible.13 Alternatively, *443it could be that the trial court found that any induced nystagmus on the smooth pursuit portion of the test rendered the entire test unreliable, especially if that portion of the test was performed first. Because it is unclear as to how the trial court arrived at its finding of unreliability, we vacate the trial court's order and remand for further proceedings on this issue. Williams , 301 Ga. at 62, 799 S.E.2d 779 ; Stroud , 344 Ga. App. at 834 (3), 812 S.E.2d 309.
For the reasons set forth above, we vacate the trial court's order granting the motion to suppress and remand for further proceedings, including a hearing on the questions of: (1) whether Burns performed the HGN test properly; (2) whether failure to perform the test within the applicable guidelines could trigger a false positive result; and (3) whether, if one portion of the HGN test was performed incorrectly, that fact renders the entire test unreliable. Any new order entered by the trial court should address these issues, as well as any credibility determinations affecting the court's decision to accept or reject the results of any field sobriety tests.
*29Judgment vacated and case remanded with direction.
Dillard, C. J., and Hodges, J., concur.
Other than the testimony of the arresting officer that Culler's breath sample was "positive," the State provided no evidence concerning the results of that test. The accusation filed against Culler, however, charged him with DUI per se by having a breath alcohol concentration of ".08 grams or more."
See OCGA § 40-6-391 (a) (5) (prohibiting any person from being "in actual physical control of any moving vehicle while ...[t]he person's alcohol concentration is 0.08 grams or more at any time within three hours after such driving or being in actual physical control from alcohol consumed before such driving or being in actual physical control ended").
See OCGA § 40-6-391 (a) (1) (prohibiting any person from being "in actual physical control of any moving vehicle while ... [u]nder the influence of alcohol to the extent that it is less safe for the person to drive").
The DVD was not played during the hearing, nor was the arresting officer questioned about the same. Instead, because of time constraints in the court's schedule, the parties agreed that the DVD would be introduced into evidence at the hearing, and that the trial court would review the DVD independently, following the hearing.
Burns testified that at the time of trial he had been a state patrol officer for 10 years and was POST certified. He had received training in standardized field sobriety testing, as well as training in "drugs and impaired driving."
The audio associated with the walk and turn test was captured on the recording.
Although Burns provided no testimony as to how many clues of impaired driving one would look for on the walk-and-turn test, the trial court's order states that Culler was positive for one out of eight clues on that test. We assume that the court's statement that there are eight possible clues on the walk-and-turn test is based on the court's experience. In any event, the State does not challenge this finding on appeal.
In his motion to suppress, Culler did not challenge the HGN test specifically. Rather, he challenged only his arrest. The State, however, does not enumerate as error the trial court's decision to assess whether the HGN test was administered correctly. Instead, as discussed in Division 2, the State asserts that the trial court's assessment contains legal and/or factual errors.
Given that police did not perform the breathalyzer test showing that Culler's blood alcohol concentration was above the legal limit until Culler was in custody, Burns did not have probable cause to suspect Culler was DUI per se at the time he conducted the arrest. See Spencer v. State , 302 Ga. 133, 136, 805 S.E.2d 886 (2017) (holding that evidence of positive test results on an HGN test, standing alone, is not evidence of DUI per se).
As its name suggests, the HGN test is designed to detect a nystagmus, which "is an involuntary jerking of the eye [that] can occur as a result of impairment by depressants (including alcohol), inhalants, or dissociative anesthetics." Walsh v. State , 303 Ga. 276, 277, 811 S.E.2d 353 (2018). The HGN test apparently includes three "evaluative components," one of which is called "smooth pursuit." See Parker v. State , 307 Ga. App. 61, 64 (2), 704 S.E.2d 438 (2010). A person who exhibits positive clues on all three of these evaluative components in both eyes tests positive for six out of six clues of impairment. Id.
In its brief, the State refers to the trial court's "exclusion" or "suppression" of the HGN test results. And both parties cite cases dealing with the admission or exclusion of the results of such a test at the trial of a defendant's DUI case. The trial court, however, neither suppressed nor excluded Culler's HGN evaluation. Instead, it suppressed the evidence resulting from Culler's arrest - i.e., Culler's breath test. In reaching the conclusion that no probable cause supported the arrest, the court declined to give any weight to the HGN evaluation.
Notably, the State does not contend that the video shows that Burns performed the test within the applicable guidelines - i.e., it does not contend that the video shows that it took Burns approximately two seconds to move the stimulus from the center to the side.
Logically, such a finding would mean that the trial court did not believe Burns's testimony regarding the positive result on the remaining four clues. The trial court is, of course, entitled to make such a credibility finding. "[T]he trier of fact is not obligated to believe a witness even if the testimony is uncontradicted and may accept or reject any portion of the testimony ... Factors such as demeanor, contradictory or inconsistent statements[,] and evidence that an officer had 'ulterior motives' can all lead a finder of fact to disregard testimony by an officer." Walsh , 303 Ga. at 282, 811 S.E.2d 353 (citation and punctuation omitted). Here, however, the trial court did not make such a finding expressly, and we hesitate to infer such a significant credibility determination from the court's order.