VERGERONT, J.

William R. Peterson appeals a judgment of conviction of homicide by intoxicated use of a vehicle in violation of § 940.09(1)(a), Stats.; four counts of causing injury by intoxicated operation of a motorboat in violation of § 30.681(2)(a), Stats.; and failing to render aid in a boating accident that involved an injury of great bodily harm in violation of § 30.67(1), Stats. On appeal, he contends that the trial court erroneously excluded from evidence a videotape that demonstrated the conditions on the river at the time of the accident. We conclude that the trial court erred in requiring expert testimony as a foundation for the videotape and erred in relying on prior personal experience. We therefore reverse and remand to the trial court to redetermine the admissibility of the videotape.

BACKGROUND

On August 16, 1996, Frederick and Connie Sandvik took seven of their friends for an evening cruise on the Black River in their boat. Shortly after nine o'clock, their boat was struck by a boat driven by Peterson, who was allegedly intoxicated at the time. As a result of the collision, one passenger on the Sandvik boat died and four were injured. Peterson was charged with homicide by intoxicated use of a vehicle, four counts of causing injury by intoxicated operation of a motorboat, and one count of failing to render aid.

At trial Peterson raised the affirmative defense that the accident would have occurred even if he had not been under the influence of an intoxicant, calling witnesses who testified that the Sandvik boat was not lighted and other witnesses who testified that it was very hard to see an unlighted boat on the river on a night with similar conditions to that of the accident. See § 940.09(2), Stats.

In anticipation of this defense, Peterson's investigator prepared a videotape for the purpose of demonstrating to the jury the approximate conditions at the site of the collision immediately prior to the accident, including the visibility on the river at the time. The video camera was on Peterson's boat, which was positioned as it was at the time of the accident, moving towards another boat, positioned as the Sandvik boat was at the time of the accident. The videotape was made at night on September 17, 1996, by investigator James Brieske. At a pretrial evidentiary hearing, Peterson moved that the videotape be admitted into evidence at the trial. To support this motion, Peterson offered evidence that the boat being viewed (and moved toward) was similar, but not identical, to the Sandvik boat and that September 17 was a night with similar, but not identical, cloud cover and lunar illumination as that of the night of the accident. Peterson showed the videotape to the court and Brieske testified that the videotape represented what he saw on September 17.

The trial court denied the motion, finding there was "a lack of adequate foundation" and the videotape "would mislead and confuse the jury." The court gave these reasons for its decision: there was a lack of expert testimony establishing that a video camera sees as the human eye does; based on the judge's personal experience of being on the river at night, the videotape was not an accurate representation of what one can see; and the Sandvik boat was "substantially larger" than the boat used in the demonstration and would have been easier to see. The court also referred to its observations of the videotape in explaining its decision.

ANALYSIS

Whether to admit a demonstrative videotape into evidence is a discretionary decision, which we review with deference to the trial court. Gieseke v. DOT, 145 Wis.2d 206, 210, 426 N.W.2d 79, 81 (Ct.App. 1988). We affirm discretionary determinations if the trial court applied the correct law to the facts of record and reached a reasonable result. See Loy v. Bunderson, 107 Wis.2d 400, 414-15, 370 N.W.2d 175, 184 (1982).

In Maskrey v. Volkswagenwerk Aktiengesellschaft, 125 Wis.2d 145, 165, 370 N.W.2d 815 (Ct.App. 1985), we reviewed a trial court's decision to admit a motion picture of crash experiments that showed how a certain vehicle reacted in a particular type of accident. We stated:

Pretrial experiments may be admitted into evidence if their probative value is not substantially outweighed by prejudice, confusion, and waste of time. Similarity of the movie to replicate the issues in the case is required. Motion pictures of experiments are proper, if no distortion occurs, but if it does exist, then the films should be excluded.

Id. (footnotes omitted). In other words, before a videotape of a demonstration may be admitted into evidence, there must be a foundation for the videotape — that it is a fair and accurate representation of what was seen — and for the demonstration — that it was conducted under conditions reasonably similar to conditions existing at the actual event. Even if this foundation is established, the trial court may, in its discretion, exclude the videotaped demonstration upon a finding that the probative value of the videotape is outweighed by its prejudicial effect. See § 904.03, Stats.

Other jurisdictions have suggested that the following factors are appropriate for trial courts to consider when determining the admissibility of demonstrative evidence: the degree of accuracy in the recreation of the actual prior conditions; the complexity and duration of the demonstration; other available means of proving the same facts; the risk that the demonstration may impact on the fairness of the trial; and whether the exhibit will aid the jury or confuse it. See Lambert v. State of Ind., 643 N.E.2d 349, 353 (Ind. 1994); reh'g granted, rev'd in part on other grounds, 675 N.E.2d 1060 (1996); Potlatch v. Mo. Pac. R.R. Co., 321 Ark. 314, 326, 902 S.W.2d 217, 224 (1995). This is not an exhaustive list, but it is consistent with Maskrey while providing somewhat more guidance to trial courts than Maskrey.

One reason the trial court found the foundation inadequate in this case is that there was no expert testimony to support the proposition that the videotape was an accurate representation of what the human eye would see. Peterson argues that no expert testimony was required and Brieske's testimony was adequate to establish a foundation for the videotape. The State replies that the trial court was properly "concerned with the distorted perception of Peterson's vision recreated by the videotaped picture" and correctly adhered to Maskrey in concluding the foundation was inadequate.

The State provides no authority to support the trial court's imposition of a requirement that, as a matter of law, expert testimony is necessary to establish a foundation for video images, and we are aware of none. Wisconsin case law does not impose such a requirement for the admission of still photographs. In State v. Sarinske, 91 Wis.2d 14, 44, 280 N.W.2d 725, 739 (1979), the court held that for still photographs, the photographer's testimony that the pictures accurately portray what they purport to portray is a sufficient foundation under §§ 909.01 and 909.015(1), Stats. Other jurisdictions have routinely held motion pictures admissible with the same foundation as that held adequate for still photographs in Sarinske. See, e.g., Long v. Gen. Elec. Co., 213 Ga. 809, 809, 102 S.E.2d 9, 9 (1958); see also 2 McCormick on Evidence § 214, at 17 (John William Strong ed., West Publ'g Co., 4th ed. 1992). As one commentator explained:

If a witness can testify based on personal knowledge that the movie or videotape is a fair and accurate representation of what is depicted, it is admissible subject to W.S.A. 904.03 considerations. Testimony about the technical details of the process is necessary only in the absence of testimony by a witness with personal knowledge that the videotape or movie fairly and accurately depicts the events shown. . . .

7 Daniel D. Blinka, Wisconsin Practice: Evidence § 401.4, at 74 (West Publ'g Co., 1991).

We see no reason to require a more stringent foundation for videotapes than photographs. As with still photographs, the photographer's/witness's testimony that the videotape fairly and accurately portrays what he or she saw is sufficient. In this case, Brieske, the photographer, did testify that the videotape was "a very good reproduction" and it "closely resembled" what he saw. The State offered no evidence to the contrary. Under these circumstances, we conclude it was improper for the trial court to exclude the videotape based on a lack of expert testimony regarding its accuracy.

A second reason the trial court determined that the foundation for the videotape was inadequate was the judge's opinion that, based on his personal experience on the river at night, "you can certainly see a lot more than that video represents." Although there was no evidence offered to counter Brieske's testimony that the videotape was an accurate representation, the trial judge found Brieske's testimony incredible based on the judge's own experience. Peterson argues that it was improper for the court to rely on its own experiences over that of a witness who was present at the demonstration, and we agree.

A trial court sitting as fact-finder may derive inferences from the testimony and take judicial notice of a fact that is not subject to reasonable dispute, but it may not establish as an adjudicative fact that which is known to the judge as an individual. Hoeft v. Friedli, 164 Wis.2d 178, 189, 473 N.W.2d 604, 607-08 (Ct.App. 1991). In Hoeft the trial judge knew, from personal experience, that the author of a particular letter in evidence had a sense of humor. Based on that personal knowledge, the court discounted the letter's evidentiary value. We held this was an erroneous exercise of discretion because the author's "sense of humor was neither part of the evidence nor a fact generally known." Id. at 189-90, 473 N.W.2d at 608. Similarly, here the trial judge's opinion of what one can see on the river at night is neither part of the record nor a generally known fact suitable for judicial notice. As we stated in Hoeft, "[w]e recognize that the trial judge's opinion was guided by good faith reliance upon his past experience and personal knowledge." Id. at 189, 473 N.W.2d at 608. However, the trial judge may not rely on his own experience on the river at night to determine whether the videotape was an accurate portrayal of the demonstration. We conclude that the court erroneously exercised its discretion in doing so.

The court also stated that the Sandvik boat was "substantially larger" than the boat used in the demonstration and therefore gave a better opportunity to be seen. It was proper for the court to consider whether conditions in the demonstration were different from conditions existing at the actual event. See Maskrey, 125 Wis.2d at 165, 370 N.W.2d at 825. However, not every difference in condition warrants exclusion. "If enough of the obviously important factors in the case are duplicated in the experiment, and if the failure to control other possibly relevant variables is explained, and if the jury is aided, the court should let the evidence in." Id. Based on our review of the record, including a viewing of the videotape, we are unable to conclude that the different size of the boat was "an obviously important factor" that was, in itself, enough to warrant excluding the videotape.

In addition to finding an inadequate foundation, the trial court concluded that the videotape would mislead and confuse the jury. Even if a proper foundation is laid for the videotape and the demonstration, a trial court may properly exclude a videotaped demonstration upon a finding that its probative value is outweighed by its unfairly prejudicial effect. See § 904.03, Stats. See also Maskrey, 125 Wis.2d at 165, 370 N.W.2d at 825. The State contends the court's conclusion was proper because the court's ruling that the videotape did not fairly represent nighttime visibility conditions on the river made its probative value very low. However, we are unable to determine to what extent the court based this conclusion on its observation of the videotape (such as the glow from the dashboard lights, which it mentioned), and to what extent the court based this conclusion on its personal experience of being on the river at night. The former is proper and the latter, as we have already held, is not.

Peterson argues that the ruling excluding the videotape was not harmless error, and the State does not respond to that argument. When a respondent does not refute an appellant's argument, we may assume it is conceded. We accept the concession that the error was not harmless. State ex rel. Sahagian v. Young, 141 Wis.2d 495, 500, 415 N.W.2d 568, 570 (Ct.App. 1987).

We are unable to determine whether the trial court would have excluded the videotape if it had not imposed a requirement of expert testimony and had not considered its prior personal experience on the river at night. When the trial court has made an error that underlies the exercise of its discretion, we may not exercise the trial court's discretion for it, but are to remand to permit the trial court to exercise its discretion. See Wis. Ass'n of Food Dealers v. City of Madison, 97 Wis.2d 426, 434-35, 293 N.W.2d 540, 545 (1980). See also Hoeft, 164 Wis.2d at 191-92, 473 N.W.2d at 608. We therefore reverse and remand to the trial court to redetermine whether the videotape is admissible consistent with this opinion, based on the evidence in the record and the court's viewing of the videotape. If the court determines the videotape should have been admitted, Peterson is entitled to a new trial. If the court determines that the videotape should have been excluded, the judgment of conviction shall stand.

By the Court. — Judgment reversed and cause remanded with directions.