delivered the opinion of the court:
James E. Lippert appeals from his conviction on one count of murder following a bench trial in the circuit court of Madison County. The trial court found that Lippert had set fire to his house knowing that there was a strong probability that the fire would cause death or great bodily harm to his wife, Ruth. During the fire, Ruth died of as*491phyxia caused by smoke inhalation. Lippert was sentenced to 40 years in prison.
At about 2:30 a.m. on October 28, 1980, fire broke out in the Lip-pert home. Lippert said he heard the sound of breaking glass and some sort of thumping sound. When he went into the living room he saw flames near the picture window. The house filled with smoke. Lippert returned to the master bedroom he shared with his wife but was unable to find her there. He looked in other rooms and in the garage but did not find her. Lippert left the burning house through the garage.
A neighbor saw the flames and called the fire department at about 2:34 a.m. Deputy Sheriff Gibson was on the scene at 2:39 a.m. He saw Lippert running nude from the garage to the back of the house, where he put a ladder up to the spare room window. Gibson heard Lippert screaming his wife’s name and observed that Lippert was visibly upset. Gibson thought Lippert was not dirty from smoke. Shortly after his arrival, Gibson saw flames shooting out of the master bedroom as well as from the living room window.
Volunteer firemen began arriving within five minutes of the first call. They found an unusually hot fire which produced dense smoke and reignited after it had been hosed down, most notably in the living room and stairwell area. Firemen noted that the top of the front door and the top of the door leading to the garage were burned away. After the firemen put a ladder up to the window of the master bedroom, they saw Ruth’s body on the bed. When the firemen were finally able to get into the master bedroom, they found the body at an angle on the bed, the detached left foot on the floor next to the bed and the right leg over the edge, with the right foot resting on the floor.
While the firemen struggled with the fire, Deputy Gibson placed Lippert in the police car and covered him with a jacket. Later Lippert was taken to a neighbor's home, where he was given clothes. The neighbors who observed Lippert testified that he was obviously upset, smelled of smoke and was covered with a black film which came off on a washcloth. He was seen coughing and expelling “black stuff” from his mouth. He had first degree burns across his shoulders and back and his hair was singed on top. At the hospital he was treated with oxygen and Valium for smoke inhalation and an anxiety reaction. He refused to be admitted.
Police and fire investigators were notified of the fire very quickly. A police photographer arrived at 3:48 a.m. State Fire Marshal Bux-ton, East Alton Fire Chief Shewmaker, Deputy Sheriff Crockarell and Alva Busch of the Illinois Department of Law Enforcement conducted *492the initial investigation of the cause and origin of the fire. Once they were able to view the interior of the house and examine the debris, the investigators quickly concluded that the fire resulted from the deliberate ignition of a flammable liquid which had been poured in the living room, trailed into the fatal bedroom, down the stairs to the basement and over to the door leading to the garage. Buxton ruled out a malfunction in either the electrical or heating system as the cause of the fire. Shortly after 6:35 a.m., the deputies who were interviewing Lippert at the hospital informed him that he was suspected of setting the fire. The sheriff’s department already knew that on October 24, 1980, Lippert had been convicted of indecent liberties with a minor and had been released on bond the day before the fire.
Lippert was interviewed several times in the hours immediately after the fire. He denied his own involvement in starting the fire but said to officers as well as to friends that he thought the house had been firebombed by someone seeking revenge in connection with the indecent liberties conviction. Lippert agreed to take a voice stress test as a check on his veracity. At trial, the result of the test was not admitted, but admissions made by Lippert at the time of the testing were ruled admissible. At that time Lippert stated that he had taught chemical and psychological warfare during the 90 days he was in the Air Force, but Air Force records show only that he was honorably discharged as a private first class for medical reasons. In describing his actions during the fire, Lippert said he thought someone, perhaps Ruth, was beside him in the smoke but that the only way out for that person was through an upstairs window because “the fire was coming down [the stairs] behind me.” Deputy Sheriff Fischer testified that Lippert described to him the fire “trickling down the stairs *** following a line down the steps” and so blocking his return to the upper floor of the split level house.
On November 7, 1980, Lippert was indicted on two counts of murder and one count of aggravated arson. Trial was preceded by extensive discovery and vigorous argument of pretrial motions. Of particular concern to the defense was a statement given to the State by Earl Abney, a prisoner who had shared a cell with Lippert on March 11 and 12, 1981. Abney stated that Lippert admitted stuffing rags in the furnace and tampering with the wiring but was confident he had left no incriminating evidence. Abney reported that Lippert talked freely because they were both in jail for indecent liberties with a minor. On May 1, 1981, Lippert filed a motion in limine requesting that the State delete from Abney’s statement all references to sex offenses and prevent any reference to such offenses from being made in the *493presence of the jury.
On May 29, 1981, the court ruled that Abney could not refer to sex offenses in his testimony except when he repeated Lippert’s own references to Abney and Lippert having the common experience of being in jail on sex charges. Ruling on other motions, the court ordered the State to refrain from any mention of Lippert’s conviction except for impeachment purposes. Lippert was permitted to show that his conviction was on appeal. On June 8, 1981, Lippert waived a jury trial because he thought that the trial court, which had conducted Lippert’s previous trial, was best qualified to weigh the prior conviction and to correctly construe Lippert’s decision not to testify if he were to make that choice at trial. He also thought the trial court could best evaluate the “highly technical” nature of expert testimony.
Although most of the testimony at trial was that of expert witnesses and was technical in nature, Abney also testified for the State as expected. He stated that Lippert had showed him sketches of the furnace and explained in detail how he had arranged it so that wires would appear to be melted together and there would “be something melted inside [the furnace].” Abney claimed his conscience was shocked by the coldness with which Lippert spoke of killing his wife and going into the obscene picture business. Abney denied that he had made a deal with the prosecution in exchange for his testimony. On cross-examination Abney admitted that he had previously lied in court and had signed false affidavits concerning his own criminal record in order to gain release from jail. During post-trial motions, the court learned that Abney had been placed on probation, which the prosecution explained had occurred because the minor plaintiff in the indecent liberties case against Abney had failed to identify him. Another prisoner, who had been in jail with Lippert in November 1980, a few days after the fire, testified for the State that Lippert drew a rough diagram of the furnace to show how black people had caused the furnace to overheat and catch fire.
The State’s case was primarily based on burn patterns observed and interpreted by its expert witnesses. These witnesses testified that fire normally burns up and out from its origin and that it usually burns in a continuous path as it spreads across a surface such as a carpet. In the Lippert house the patterns of bums on the floors and stairs were said to be “spotty” and “erratic,” but when viewed altogether they formed a network connecting the master bedroom, living room, and stairs. A deep char at the top of the stairs, isolated from other deep burning, was part of a “sporadic and irregular” pattern of burning on the center of the stair treads. This pattern was considered *494abnormal because a fire which had begun upstairs would not have moved down except as carried by a fuel source at floor level. A fire which began in the basement should have consumed the entire lower portion of the stairs before the upper stairs were burned. Intermittent burning of the studs supporting the stairs indicated fire traveling up and down the stairs simultaneously. An aluminum door which opened to the outside from the landing of the stairs was melted at both the top and bottom. The top melting was considered the result of normal venting of the fire at the top of an exit. The melted bottom indicated “a very abnormal heat at floor level” in the foyer. The garage door also burned at the bottom.
In the master bedroom the burn patterns were localized in the northwest portion. The burns appeared underneath and on both sides of the bed, which was two or three feet from the west wall with its head against the north wall. Buxton described a heavy irregular burn around the north vent on the west wall, although he conceded that photographs showed unconsumed carpet in the area where he thought accelerant had been poured. He also described a crescent-shaped burn beneath the bed as the possible result of a sloshing action with a gas can which itself would not fit under the bedframe. While one expert testified that low burning along the north wall had not entirely consumed the baseboard and carpet at floor level, another expert pointed out that the paneling on the northwest corner of the master bedroom was burned away above a line level with the mattress, indicating to him that the fire emanated from the surface of the bed. The bedroom ceiling showed the same localized burning in the northwest portion of the room as the burn patterns on the floor. The paneling on the east wall above a heat duct showed no sign of an open flame venting through the duct although the wall showed the effects of great heat in the room.
The State’s witnesses pointed out that the burns were not only irregularly placed but also unexpectedly deep in certain spots. The most severely burned areas upstairs were the west wall of the bedroom near floor level, that is, within two or three feet of the body, and the area in the living room along or near the west wall where there was a picture window. The State’s experts considered it important that the burn in the living room consumed the carpet and padding and penetrated the wood flooring, since deep burning could indicate accelerant holding the heat to the floor. Buxton explained that a burning chair could also create deep burns by holding heat to the floor. Charring beneath a living room bookcase, not attributable to burning furniture, was also offered as evidence of a liquid accelerant. The vertical sup*495ports and the underside of the stairwell were heavily and regularly charred, evidence that a hot, fast fire had sprung up from floor level in the basement and burned up the stairwell, fed by the natural flue effect of the stairs as well as by household items stored beneath the stairs. Buxton pointed out that the accelerant could have been poured beneath the steps or could have dripped through the stairs from above. He also testified that gasoline vapors which feed a fire are heavier than air and accumulate in low places such as stairwells, heating ducts, or surface depressions.
The State’s experts considered two small areas of broken concrete floor in the basement indicative of pooled liquid burning for an extended period. This breaking of the concrete, called spalling, is caused by the expansion of water vapor within the concrete when heat is held against the surface. The State did not determine that the points of spalling were in fact low spots in the concrete; a defense expert testified that water drained away from the spalls. The defense also countered with a learned treatise which stated that spalling was often erroneously taken as the sign of an accelerant. A concrete expert testified that spalling could occur where burning wood or metal fell on the concrete and no liquid was present.
The State’s investigation of the basement provided nothing to alter its theory that an accelerant was the cause of the fire. The investigators checked the electrical system as well as the controls and burn chamber of the furnace but saw nothing to identify either as the source of the fire. They saw no evidence that fire had come out of the combustion chamber and moved to the joists above the furnace. Bux-ton did not look into the top of the furnace because the paint on the outside indicated to him that the furnace was not “relevant to the fire.” Another expert “pretty well eliminated that furnace area as a source of a tremendous amount of heat or flame” because he observed that the copper solder on the A-coil was unmelted. He could not connect the floor level burning on the north end of the basement near the garage door with the burned joists in the south end of the basement. No one found rags stuffed into the fire chamber in a manner consistent with Abney’s story. The State’s experts did not note that ductwork leading from the furnace to the west side of the house was missing.
A variety of physical evidence was gathered for testing. Among other items, the State took carpet and flooring from the west side of the bed, carpet and pad from the living room, a section of the vinyl couch, debris from under a bookcase in the living room where a hydrocarbon detector had given a positive reading, and baseboard from *496the basement hallway. Tongue-and-groove flooring from the living room was pried apart to show burning in the grooves. The mattress from the master bedroom was not tested. A sample of carpet taken from outside the “pour area” was used as a control sample in the testing process. A t-shirt was taken from a laundry basket in the basement because it smelled of petroleum. Testing of these items by the State’s forensic scientist failed to show any trace of a definitely identified accelerant. The t-shirt gave a positive reading of some kind of solvent or hydrocarbon in the midrange of flammability. The State’s experts testified that the gas chromatograph used to test for acceler-ants could fail to register a flammable liquid even when a person could still smell it in the sample and that fuels could be completely consumed by fire or washed away by the firefighters. The carpet taken from beneath the bookcase gave the same test result as the clean sample of carpet. Three items contained gasoline: two gas cans taken from the garage and a sample of gasoline taken from the car in the garage.
The State contended that the condition of Ruth’s body confirmed the presence of an accelerant. Dr. Harry Parks, the forensic pathologist who performed the autopsy, testified that the lower torso and internal organs were charred. Both hands were also burned. Parks’ examination revealed no burning of the respiratory system, no trauma to the brain or skull. He testified that while it was common to see severe burns “in gasoline fires in cars and planes,” he could not say “this degree of burning means it had to have been gasoline involved ***.” Parks considered a distance of three or four feet between Ruth’s face and an accelerant fire sufficient to account for the lack of burning in her trachea and lungs shown by his microscopic examination.
Over defense objections, George E. Gantner, Jr., chief medical examiner of St. Louis City and St. Louis County, testified concerning the cause of the burning to the body. Gantner based his evaluation on Parks’ autopsy report and pictures of the body and master bedroom. He likened the condition of the body to that of other victims he had seen where fire had consumed “the flooring and structure beneath the body.” While the hands and feet commonly burn first, Gantner pointed out that “severe injury to the one extremity without complete consumption of the other parts” suggested that the body had been subjected to uneven temperatures. He believed either a hot fire or prolonged exposure or a combination of the two was necessary to produce the injury, but he did not see in the photographs enough damage to the bedroom to explain the severity of the burning. Asked directly *497whether gasoline poured on the leg could cause the injury, he answered affirmatively.
Taking into account the physical evidence as well as the intensity and stubbornness of the fire, the State argues that a flammable accel-erant had been intentionally poured in the living room, in the bedroom, on the bed and victim, and down the stairwell to the basement, where it was ignited by Lippert as he stood in safety behind the door leading to the garage. They contended that the fire erupted simultaneously throughout the house, instantly superheating the air wherever there was accelerant. Small clear fragments of the picture window suggested that it was broken out by extreme heat or the concussion of ignition before a film of smoke could form on the glass. An ample supply of oxygen entering through the broken window would have encouraged the fire.
Lippert’s motion for a directed verdict at the close of the State’s case in chief was denied. Lippert chose not to testify and the defense which followed the denial was based on the theory that the burn patterns corresponded to areas where highly flammable furniture stood and where heavy use had worn away the nap of the carpet, making it susceptible to quick burning after a malfunction in the heating system propelled heat or flame into the rooms. Neighbors and relatives testified that two or three chairs and a couch had stood in the area of the heavy burning in the living room. The carpet was described as heavily worn in that area as well as at the top of the steps. The burns centered on the .stair treads were said to correspond to the points of heaviest wear. The vertical bums on the risers were explained as the natural effect of fire burning upward. A path worn in the carpet in prior years when the bed was placed on the south wall of the bedroom accounted for the crescent burn under the bed.
A major premise of the defense was that a furnace malfunction started the fire. A furnace expert, examining the furnace in March or April of 1981, observed that the insulated control wiring was melted in the lower part and charred by open flame above. To him, melted wires meant the temperature controls were ineffective and air moving through the furnace could be heated beyond the normal limits. He saw no melted metal or plastic in the blower, evaporation coil or A-coil on top of the furnace, although a wooden framing just above the furnace was burned. The State’s rebuttal witness, an electrical engineer who was an arson investigator, testified that the cause was not a superheated furnace because the whole top of the furnace was not burned out, and damage in the furnace room was not the result of direct fire but heat from the hallway. James A. Flynn, a forensic chem*498ist testifying for the defense, cut out a section of flooring from the guest room closet just above the furnace. He found the top of the furnace “partially melted in two areas” and the flooring charred on the side nearest the furnace but not discolored on top.
Defense experts differed on the point at which the fire began. The furnace expert would not say the fire started in the furnace, only that “there was a fire in the furnace.” Others testified that the fire might have begun in the ceiling joist and plywood just above the furnace, in an opening between a heat duct and a cold air return or “at the bottom of the basement stairs at a point at about ceiling level adjacent to a hot air duct ***. At least one of those joists is completely burned through. It’s the deepest burning that was found anywhere in the house.”
Pyrophoric carbon, a condition caused by exposing wood to continuous heat, was put forward to explain the origin of the fire. The ignition point of such dried wood is lowered, so that it will burn more easily than normal wood. The furnace expert testified that the duct-work was improperly installed, so that hot air escaping through faulty joints circulated in cold air returns where there was exposed wood. One expert explained that a furnace explosion in the Lippert home in 1977 could have caused bulging of the ductwork so that heat could escape, but the remaining ductwork joints showed corners which were “square” and “intact.” A forensic chemist, testifying for the State, stated that pyrophoric carbon would produce a smoldering, smoky fire.
Several defense witnesses associated the path of the fire with the ductwork of the heating system. They testified that floor joists in the basement were burned where they paralleled ductwork leading to vents in the master bedroom, that the north wall of the bedroom was more severely burned near the cold air return than at floor level, and that wall studs in the master bedroom and living room burned only in the cold air return system. However, the furnace expert testified that the cold air return was on the east wall of the bedroom, which did not burn. Descriptions of the heat register's on the west wall of the living room were inconsistent, although they agreed that one register was burned on one end only.
A sizeable section of aluminum duct distributing heat to the west side of the house was missing. Some melted aluminum was found on top of the main beam in the basement and a section of partially melted duct remained in the garage. Neither the missing duct nor its melted residue was found. That portion which remained was described as “severely smoke stained” and “black on the inside” while the east *499duct was “clean inside.” Defense experts testified that the furnace blower propelled heat or even “flaming bits” out into rooms filled with highly flammable draperies, paneling, carpeting and furniture, which supported a hot, intense fire.
The State’s case was considered defective because it did not recognize that the combustion of these furnishings would explain the damage to the house and body. Several witnesses described a process in which draperies ignited by heat or flame from the vents could have fallen on the carpet and couch, leading to a fire which would have made the upper floor untenable in five or 10 minutes. Although no evidence was offered to show that a fan in the south end of the attic was operating, the defense contended that the effect of such a fan and the brisk wind from the north would be to drive the flame south from the living room to the master bedroom.
John F. Connell, a New York expert on the cause and origin of fires, viewed the Lippert home three times shortly before trial and reviewed police reports and photographs. He testified that the fire was caused by a defect in the heating system and directly contradicted major points forwarded by the State’s experts. He did not accept their assertion that carpet doused with a flammable liquid might not register on the gas chromatograph. He testified that the gas chro-matograph “would pick it up ***” even if it could not be identified. He stated that the mattress also would have retained traces of a flammable liquid. Neither party tested the mattress. He considered the fourth degree burns on the body fully consistent -with the length of time it was in the fire and not necessarily the result of an accelerant. After clearing and photographing the floors in the dining area and guest bedroom where the State had not alleged any accelerant had been used, he testified that burning material falling on those floors had produced burns similar to those in the living room and bedroom. Connell did not see any difference between the burns at the top of the stairs and those caused by the dropping of burning materials in the other rooms. He characterized the damage to the steps as burning of “the leading edge of the tread ***” because of wear.
Photographs and samples of flooring from the kitchen and guest room were offered by the State to show that the floor burns in those rooms were related to the falling of ceiling tiles rather than acceler-ant and that the tongue-and-groove joints did not show the same deep burning as samples taken from the living room.
In Connell’s opinion, a deliberate pour of accelerant in sufficient quantity to explain the burning under the steps would have destroyed the steps entirely and a person pouring enough gasoline to explain the *500charring of the floor would have been blown away from the house when the furnace pilot light ignited the gasoline fumes. He assumed the pilot light and blower were on because the weather was cool and there was soot and smoke stain in the ductwork.
In rebuttal, Barker Davie, a forensic chemist and fire investigator, described four areas in the living room not associated with melting furniture, which indicated that an accelerant had been used: charring around the lip of a heat register, charring under the bookcase and two isolated hot spots in the hallway. He did not see ceiling damage above the living room registers commensurate with rising hot gas and no evidence that rubberbacked draperies had fallen on the floor or near where the couch was supposed to be. He testified that the mattress was not a good fuel source because of its density but the wall joists behind the bed were heavily charred about 18 inches from the floor. Davie contradicted Connell’s assertion that a sizeable pour of gasoline would have caused a power explosion. He explained that an explosion depended on the presence of a particular ratio of vapor to air and that if there were too much or too little vapor there could be fire without explosion and that the fire might burn along the path of a “pour” but not necessarily consume nearby materials.
After complimenting both sides on their preparation and presentation, the court retired to consider the evidence. On August 31, 1981, the trial court found Lippert guilty of one count of murder. A post-trial motion, a motion to arrest judgment and a motion for a new trial to admit additional evidence were denied. Judgment and sentence were entered on August 17,1982. This appeal followed.
On appeal Lippert presents the following issues: (1) whether his conviction was “against the manifest weight of the evidence”; (2) whether statements made by Lippert during a voice stress test were improperly admitted; (3) whether it was reversible error for the trial court to permit Earl Abney to testify that Lippert had been convicted of a sex offense; (4) whether the denial of Lippert’s motion in limine to bar impeachment with a prior conviction was error; (5) whether unsupported insinuations by the prosecution denied Lippert a fair trial; and (6) whether he was denied a fair trial because the prosecution delayed in making available an insurance report including pictures of the house and furnace as they appeared in 1977.
The review of testimony presented at the trial shows that virtually every factual issue was a matter of dispute between the experts. Most of the evidence was circumstantial, although Lippert’s statements to the police provided direct evidence of the beginning of the fire and his own actions. Where the evidence is conflicting, it is *501the duty of the trier of fact to weigh the evidence and judge the credibility of the witnesses. A reviewing court will not reverse a criminal conviction unless the evidence is so improbable that it creates a reasonable doubt of the defendant’s guilt. (People v. Manion (1977), 67 Ill. 2d 564, 367 N.E.2d 1313.) A conviction can be sustained only if it rests on the strength of credible evidence presented by the prosecution. (People v. Morgan (1970), 121 Ill. App. 2d 196, 257 N.E.2d 554.) Circumstantial evidence is sufficient to support a conviction if it leads to “a reasonable and moral certainty that the accused and no one else committed the crime.” (People v. Williams (1977), 66 Ill. 2d 478, 484-85, 363 N.E.2d 801, 804.) We find that the record contains sufficient evidence to support the trial court’s judgment that Lippert deliberately set the fire which caused his wife’s death.
The thrust of the State’s case was that the fire burned abnormally and that all the abnormalities indicated the use of an accelerant. The fire was hotter, faster and more stubborn than the usual house fire. It burned in a discontinuous pattern, with three major sites of severe burning connected by a spotty trail of lesser burns. While some of the burn pattern could be associated with melting furniture, fallen tile, or worn carpet, there was also burning at isolated points in the hallway and in comers where there was no vinyl furniture, tile or walkway. The floor level burning of baseboard and lower step in the basement and the “alligator charring” in the stairwell sides indicated a fast fire coming up from the basement. The low burning in the basement was beneath and away from the point in the joists where the defense indicated that pyrophoric carbon could have ignited. The thin pattern of bums on the upper half of the stairs indicated fire moving down, contrary to its normal tendency to bum up and out. Lippert’s own description of the fire was a “trickling *** following a line down the steps.” It was not clear whether the floor of the foyer which separated the upper and lower steps showed a continuation of burning between the upper and lower levels, although enough heat was generated at floor level on the foyer to melt the aluminum door at the bottom, contrary to the normal tendency of fire to vent itself at the highest point possible. The garage door in the basement was also burned from the bottom.
In the master bedroom the severest burning was localized in the northwest corner opposite the door. The burning at floor level on the west side of the bed left an irregular pattern and burned up toward the side of the bed. A defense witness testified that the registers in the bedroom, one of which was near the bed, appeared to have been closed at the time of the fire. The burning on the north wall of the *502bedroom appeared to have burned up and down from a line parallel to the mattress top, charring the top of the baseboard but leaving the bottom edge and wooden tack strip beneath it intact. There was testimony that a mattress is not a good fuel because of its density and that the mattress was not much damaged. Nevertheless the result of the localized burning in the master bedroom was a severely charred body on the mattress. Although the condition of the body was consistent with either the presence of an accelerant or a long exposure to fire, other flammable objects in the room did not ignite, although they were subjected to an intense heat and were in the house for the same period as the body. A pillow on the floor next to the closet on the east side of the room did not burn. The east wall had been brought near its ignition point but did not catch fire.
Two gas cans owned by Lippert were taken from the garage. Lip-pert said they were usually kept in a shed and he did not know why they were in the garage. They contained small amounts of gasoline. Although none of the samples taken from the burned areas showed the presence of an identifiable accelerant, a t-shirt recovered from the basement was repeatedly tested. A forensic chemist determined that the vapors on the shirt had 40 points of similarity with gasoline, although he felt more comfortable calling the substance a flammable hydrocarbon. State investigators looked for evidence of forced entry and firebombing but found nothing to suggest the presence of a third party at the fire scene. There was no evidence of rags in the furnace or of deliberately crossed wires on its controls. Lippert and his wife were the only people in the house when the fire started. We find that the evidence left no reasonable doubt of Lippert’s guilt.
Lippert claims that the court erred in denying his motion in limine to prohibit all statements made during a voice stress test. On October 28, 1980, at 10:05 a.m., Detective Eck of the Granite City police department interviewed Lippert prior to Eck's administering a Psychological Stress Evaluation Test (P.S.E.) to which Lippert had consented. A polygraph, which uses cardiovascular and respiratory patterns as a measure of a person’s reactions, was not available on October 28 and Lippert preferred to take the P.S.E., which measures reactions by changes in vocal patterns. Lippert was twice given his Miranda warnings during the testing procedures. During the interview Lippert described his discovery of the fire in the living room and his efforts to rescue his wife. He went from the fiery living room to the smoky bedroom and then down the steps to the front door, which would not open. The fire was coming down the steps so he went to the basement “and there was no fire down there.” After checking the *503garage for his wife, he found he “couldn’t get [back] through [the basement hallway] *** that was full of fire already.” After carrying an aluminum ladder to the master bedroom window he was surprised to find the room “ablaze” in only four or five minutes. He emphasized the speed with which the fire spread. He thought the furnace blower, which he kept on 24 hours a day, could have spread the fire. At the time of the interview, he was convinced that “it was an arson job.”
The court granted a defense motion in limine to bar evidence of the results of the P.S.E. The court’s decision was in full accord with Illinois Polygraph Society v. Pellicano (1980), 83 Ill. 2d 130, 138, 414 N.E.2d 458, 463, which determined that “recordings of cardiovascular and respiratory patterns, at the least, are needed” to' “assure reasonably reliable” examinations. Even where a polygraph is used correctly, the results are inadmissible because of their unreliability. (People v. Baynes (1981), 88 Ill. 2d 225, 430 N.E.2d 1070.) The court denied a defense motion in limine to prohibit all testimony and admissions resulting from the test situation. Lippert received Miranda warnings before each session with Eck. On the first day he acknowledged the warnings by signing a copy of them. The transcript of the second day’s interview contains the warning as well as Lippert’s statement that he wished to go on with the test, although his attorney had not approved of the first test and did not know of the second. The circumstances are similar to those in People v. Varney (1978), 58 Ill. App. 3d 70, 373 N.E.2d 1033, where the court found that statements made to police officers after a voluntary polygraph were admissible because the defendant had been given Miranda warnings, had signed a form to that effect, and had waived presence of counsel. We find that the trial court properly admitted Lippert’s statements. Admissions by a defendant, even if intended to be exculpatory, are direct evidence which may sustain a conviction. People v. Brown (1974), 56 Ill. 2d 312, 307 N.E.2d 356, rev’d and rem. on other grounds (1975), 422 U.S. 590, 45 L. Ed. 2d 416, 95 S. Ct. 2254.
Lippert sought to bar any reference to his prior conviction by filing two motions in limine, one to prohibit impeachment and the other to bar direct testimony by Abney or other prosecution witnesses. The trial court denied the first motion but specifically allowed Lippert to show that he was appealing his conviction. For impeachment purposes, a prior conviction punishable by more than one year in prison is presumed to be relevant to the witness’ credibility. (People v. Miller (1981), 101 Ill. App. 3d 55, 427 N.E.2d 987.) Asked to reconsider the denial, the trial court reaffirmed its decision, commenting “I gave it considerable thought, considerable research.” The original or*504der cited People v. Montgomery (1971), 47 Ill. 2d 510, 268 N.E.2d 695, People v. Bey (1969), 42 Ill. 2d 139, 246 N.E.2d 287, and People v. Kellas (1979), 72 Ill. App. 3d 445, 389 N.E.2d 1382. Lippert asserts that the trial court refused to exercise its discretion in balancing the prejudicial and probative effects of the prior conviction as required by Montgomery. We do not agree that the trial court cavalierly ignored its duty. The record contains considerable commentary by the trial court on the requirements established by Montgomery. Where the court was aware of the Montgomery standards we will assume it gave proper consideration to the relevant factors. (People v. Washington (1973), 55 Ill. 2d 521, 304 N.E.2d 276; People v. Hovanec (1979), 76 Ill. App. 3d 401, 394 N.E.2d 1340.) We find no abuse of discretion.
Although the court denied the defense motion in limine to bar all testimony to the prior conviction by prosecution witnesses, it ordered the State to limit its direct evidence of the conviction to the testimony of a single deputy circuit clerk, who could state only that Lip-pert had been in custody and had posted bond shortly before the fire. No reference could be made to the nature of the offense, which the court considered “too inflammatory and prejudicial” to go to the jury. The court acted to prevent “a parade of witnesses running through here going into [the indecent liberties conviction.]” Abney was limited to repeating a single statement made to him by Lippert in which Lip-pert recognized that they both had been in jail on “sex charges.” Ab-ney’s testimony did not in fact repeat this statement and contained no reference to a “sex charge” except that pending against Abney. Ab-ney stated that Lippert’s “eyes looked like a demon’s eyes to me” while he explained his plot to kill his wife and go live with “a young black boy.” Under vigorous cross-examination, Abney misidentified one of the defense counsel as his own attorney and admitted that he had lied in court several times. He denied that the prosecution had offered leniency in exchange for his testimony and stated that the State’s Attorney was being very hard on him.
Lippert seeks reversal because the denial of his two motions in limine allowed in evidence so prejudicial that he decided to waive a jury and to decline to testify. Lippert’s decision not to testify was a matter of trial strategy and standing alone is not sufficient evidence of prejudice to justify reversal. (People v. Leonard (1980), 83 Ill. 2d 411, 415 N.E.2d 358; People v. Hovanec (1979), 76 Ill. App. 3d 401, 394 N.E. 1340; People v. Ganter (1977), 56 Ill. App. 3d 316, 371 N.E.2d 1072.) We are directed to no case which holds that a defendant’s decision to waive the jury is proof that the ruling which influenced the decision was error. Testimony of a cellmate has been admit*505ted where the conversations occurred before the informant went to the State’s Attorney (People v. Taylor (1982), 110 Ill. App. 3d 112, 441 N.E.2d 1231) and held “at most, harmless error” when the witness was the only one to mention prior convictions and the prosecutor did not emphasize the conviction in his argument to the jury. (People v. Dumas (1977), 49 Ill. App. 3d 756, 761, 364 N.E.2d 616, 620.) In the present case, Abney was the only prosecution witness authorized to mention the prior conviction but did not do so. The State’s closing argument did not mention a prior conviction. Abney volunteered information about Lippert nearly 10 days after the conversations.
With regard to the rulings allowing the introduction of the prior conviction, we find at most harmless error because the trial court deciding the case was already well aware of Lippert’s conviction, having presided over the previous jury trial. We note also that defense counsel and witnesses referred to Lippert’s “trial,” “conviction,” and “guilty verdict,” apparently fearing no prejudice from those repetitions before the judge.
Lippert contends that the trial court was prejudiced by the State’s insinuations that someone acting for the defense had tampered with the fire scene and removed ductwork. The State sought discovery of the ductwork. The defense moved for discovery of State witnesses who would testify to tampering. Vigorous argument followed the defense motion, and it is the State’s comments during this argument to which defendant now objects. During argument, the court warned both sides against making personal accusations and insinuations. The court ordered the State to provide the names of all witnesses who would testify to tampering. A list was provided but no witnesses so testified. Lippert points to the court’s comments that the defense’s chain of custody left “something to be desired” as evidence of prejudice. The court made these comments while admitting defense exhibits over the State’s objections, and the references did not involve the missing ductwork.
Lippert argues that a pattern of sexual innuendo prejudiced the court and denied him a fair trial by forcing him to waive the jury. The State’s motion for production of “sexual devices, i.e., dildoes ***” found after the fire was not acted upon until after Lippert had already waived the jury. The court had previously ruled that the State could not introduce evidence of the so-called sex devices to the jury until he had ruled on it during trial. After jury waiver, the State attempted to elicit testimony about the presence of the devices in a picture of the bathroom, but the court cut off the testimony, stating that the picture could speak for itself. All other sexual innuendos com*506plained of occurred even later in the trial.
Although Lippert relies on People v. Nuccio (1969), 43 Ill. 2d 375, 253 N.E.2d 353, in seeking reversal on the basis of prejudicial insinuations, we find the present case clearly distinguishable. Here the trial court recognized that counsel were engaging in insinuations and personal accusations related to “tampering” and warned them against it. He pressed the State to define “tampering” and to provide the names of those who would testify. The names were provided but no one testified. The court repeatedly acted to limit the introduction of sexual references in order to reduce unfair prejudice. We do not see any evidence that his judgment was affected by any pattern of insinuations. We therefore presume that the trial court considered only competent evidence. People v. Pelegri (1968), 39 Ill. 2d 568, 237 N.E.2d 453.
Nor do we agree with appellant that he was prejudiced by the prosecution’s comments that one of the defense counsel was a member of the Madison County Board. When the prosecution first notified the court of this matter, the court immediately informed Lippert and gave him the opportunity to decide whether to proceed with the same counsel. The court stated that he saw no impropriety in the representation. When the State attempted to raise the same issue in post-trial motions, the court’s comment showed that he gave no weight to such charges. There is no affirmative showing that the court was misled or improperly influenced by prosecutorial remarks. People v. Wallenberg (1962), 24 Ill. 2d 350, 181 N.E.2d 143; People v. Grodkiewicz (1959), 16 Ill. 2d 192,157 N.E.2d 16.
Although Lippert claims he was denied a fair trial because of the prosecutor’s delay in making known an insurance report and photographs of the Lippert house as it appeared in 1977, no authority in support of this proposition is provided. We therefore decline to consider the issue. People v. Ortiz (1980), 91 Ill. App. 3d 466, 414 N.E.2d 1072.
For the reasons given above, we affirm the judgment of the circuit court of Madison County.
Affirmed.
JONES and KASSERMAN, JJ., concur.