By Mark Gutglueck
The defense in the McStay Family murder trial rested yesterday, having put on its case over a period of ten weeks, one week longer than the prosecution had taken to lay out its basis for alleging that Charles Merritt was the perpetrator of the brutal 2010 killings. The final barrage of testimony represented a major breakthrough for the defense, during which it was able to turn the testimony of a highly reputable expert witness originally retained by the prosecution to its cause to not just refute but virtually obliterate a primary element of the narrative of the defendant’s guilt. On a separate issue, however, the defense suffered a setback when the prosecution located and brought in a surprise rebuttal witness who undercut the alternate narrative the defense had heretofore successfully propounded to the effect that it was not Merritt but rather another of Joseph McStay’s business associates who had killed him  along with his wife and two young sons.
Outside the jury’s view there were additional elements of drama, as the defense made a sally against one of the prosecutors while calling for a discontinuation of the proceedings before the current jury or the outright dismissal of the case, with Judge Michael Smith ultimately rejecting the defense’s motion for a mistrial together with its accusation that the case’s senior prosecutor, Supervising Deputy District Attorney Britt Imes, had hidden exculpatory evidence.
According to the prosecution, Charles “Chase” Merritt, driven by financial desperation that grew out of his unbridled gambling addiction and utter lack of fiscal discipline, engaged in a series of thefts from Earth Inspired Products, a company owned and operated by his business associate, Joseph McStay, Sr. Through that company, Joseph McStay was selling high end decorative water features – artificial waterfalls and fountains – which he and Merritt designed and which Merritt manufactured out of steel, glass, rock and other components based upon the specifications outlined by the company’s customers and passed along to Merritt by McStay. The prosecution’s theory, presented during the initial nine weeks of the trial that began on January 7 of this year, is that Merritt fraudulently obtained access to the QuickBooks accounting system McStay had set up for the Earth Inspired Products enterprise and on February 1 and February 2, 2010 wrote himself $7,495 worth of checks.
When he learned of what Merritt was up to, either shortly before or perhaps even on February 4, 2010, Joseph McStay traveled to Rancho Cucamonga, where Merritt was then living, and confronted him about his larceny, threatening to alert authorities, prosecutors allege. After Joseph McStay later that afternoon returned to the home he shared with his family in Fallbrook in north San Diego County, the prosecution’s theory continues, Merritt that evening drove to the McStay residence, where he slaughtered the McStays, using a three-pound sledgehammer to bash their skulls in.
Merritt then secreted the bodies for two days, in the meantime again fraudulently accessing Joseph McStay’s QuickBooks account for Earth Inspired Products, the prosecution maintains, and on February 4, 2010, February 5, 2010 and again on February 8, 2010 issued himself checks made out to himself for a total of $23,855.
On February 6, 2010, Merritt transported the corpses up into San Bernardino County’s High Desert, an area with which Merritt was familiar from having grown up in Hesperia where he attended Apple Valley High School for three years in the 1970s, according to prosecutors. There, in a remote area between Victorville and Oro Grande he buried all four of his victims along with the murder weapon in two six-foot long, two-foot wide and 18-inch deep graves he dug in a wash off a rarely-traveled dirt road, according to the prosecution.
To confuse the situation, throw authorities off his track and delay a serious investigation into the matter, the prosecution maintains, Merritt then drove the McStay family’s 1996 Isuzu Trooper, which yet contained the child seats for Gianni and Joseph, Jr., to San Ysidro, where he left the vehicle in a shopping center parking lot roughly a quarter of a mile from the Mexican border.
Placing Merritt at the McStay Home located at 3473 Avocado Vista Lane in Fallbrook on the evening of February 4, 2010 is a central and crucial element of the highly circumstantial case against Merritt. The prosecution team, consisting of Supervising Deputy District Attorney Britt Imes, Supervising Deputy District Attorney Sean Daugherty and Deputy District Attorney Melissa Rodriguez, have consistently sought to sell to the jury that at 7:47 pm on the night of February 4, 2010 a security video camera at the home of Jennifer Mitchley, who lived one door up and across the street from the McStay home in the 3400 block of Avocado Vista Lane, caught fleeting images of a portion of Merritt’s truck. This implicates Merritt in the murders, the prosecution maintains.
The security video, however, is problematic on multiple score. Black and white and of relatively low quality which is exacerbated by the consideration that its video data was compressed in order to be digitally stored, reducing the quality further, as the action captured occurred some two hours and 21 minutes after that day’s 5:26 pm sunset. Moreover, the primary video camera was mounted on the Mitchley home’s front porch relatively high on the Mitchley home’s front porch and angled in a way that was intended to take in activity in the front yard. Thus the camera’s visual field only incidentally captures the street area. What is more, because of the height at which the camera is mounted, the roof overhang above the porch cuts what would have been a slightly further extension outward from the visual field, such that the top part of the camera’s panorama ends roughly half way out into the street. In this way, the entirety of the vehicle, shrouded in darkness, is not entirely visible, with something approaching one third or more of its top portion left out of the video.
Despite the video’s limitations, the prosecution believed an exacting forensic examination and analysis would bear out its theory and assertion that the vehicle was in fact Merritt’s truck, damning evidence that places him at the scene where and when the prosecution maintains the murders took place.
To carry the day, the prosecution had been banking on Dr. Leonid I. Rudin, one of the world’s leading forensic photogrammetrists, being able to offer testimony that would leave no doubt that Merritt was at the McStay residence that fateful night.
Dr. Rudin’s credentials are impeccable.
Possessing a PhD and a master of science degree in computer science and imaging science from Cal Tech as well as a master of science degree in applied mathematics, he is a faculty member at the California Institute of Technology, UCLA, and University of Paris IX École Normale Supérieure, France. Rudin is the founder of Pasadena-based Cognitech, Inc., the world’s first forensic video processing and photogrammetry software and hardware design company. Arguably the world’s leading authority on image/video processing and analysis, Rudin holds over 20 patents, including ones for a system and method for pattern detection and camera calibration, registration and comparison of three-dimensional objects, a video data compressor with a very high data rate, a real-time three-dimensional videoing system, a system and method for determining geo-location(s) in images, a system and method for image and video search, indexing and object classification, registration and comparison of three dimensional objects in facial imaging, video demultiplexing based on meaningful modes extraction, a system and method for three-dimensional estimation based on image data, object recognition based on two dimensional images and three dimensional models, image frame fusion by velocity estimation using region merging and an apparatus for enhancing signals such as images, speech, remotely sensed data, medical, tactile, radar and audio, and other patents issued and in continuance by the US Patent Office.
Recognized during the 2000 Queen Elizabeth II Forensic Committee Conference for his work in furthering the science of human identification and the recipient of the 2010 American Technology Award for his fundamental contribution to three dimensional geospatial photogrammetry, Rudin is of note within the legal community for having advanced the application of science in the context of criminal justice based upon his having derived the first usage of peer-reviewed and published fundamental techniques for imaging, video processing and forensic three dimensional photogrammetry ever approved by the U.S. Court of Appeals to obtain a conviction. Caltech singled him out for its De Prima Mathematics Award for applying mathematics to benefit the justice system.
Upon being contacted by the prosecution, which called upon him to assist it in nailing down a crucial element of the case against Merritt by using his expertise to make a determination that it was indeed the defendant’s truck on Avocado Vista Lane that winter evening in 2010, Rudin agreed to take on the case. In addition to bringing his talent to bear, Rudin went further, offering and making good upon that offer to do the work pro bono, for no pay whatsoever, which is his routine practice through which he seeks to contribute to the US Justice System.
Merritt’s defense team, consisting of James McGee, Rajan Maline and Jacob Guerard, anticipating that Dr. Rudin was on the verge of coming into the courtroom and offering a determination that would prove absolutely devastating to their client, requested an evidentiary hearing outside the presence of the jury to go over with a fine-tooth comb what Rudin was prepared to testify to and what the basis for that anticipated testimony would be. Judge Smith granted the request, and on February 5, 2019, without the jury present, the hearing was held. During that hearing, Rudin previewed his findings and the methodology he had used, together with the video, photographic and photogrammetric evidence he had evaluated along with the exhibits he had generated. During that hearing Rudin testified that, based upon the materials and data provided to him to that point, though he could not conclusively say that the image on the Mitchely video matched Merritt’s truck, he could not reject that they were one and the same either. At the conclusion of the hearing, Judge Smith, over the defense’s objection, ruled that Rudin could testify.
Rudin was scheduled to appear as a witness on February 19, but that date came and went and the prosecution did not call him to the witness stand. Instead, on March 4, the prosecution elected to have San Bernardino County Sheriff’s Department Sergeant Ryan Smith, who is no blood relation to Judge Michael Smith, testify with regard to his non-scientific observations and the conclusion he had personally reached that the image on the Mitchley video is a match to the Chevrolet work truck Merritt was driving in 2010. Smith, who as the case officer on the McStay family homicides and the sheriff’s department liaison with the district attorney’s office on the matter, was questioned by Imes as three dimensional depictions, known as FARO scans, of Merritt’s truck were displayed in conjunction with stills from the Mitchley video on the courtroom’s overhead visual monitors, some of which were side-by-side comparisons Sergeant Smth had prepared to accompany his testimony. He testified that there was similarity with regard to the front of the images of the vehicles as well as a close similarity with regard to the back of the vehicles, including the exhaust pipe.
Sergeant Smith also testified that a light visible on the side and near the rear of the vehicle in the image captured on the Mitchley video that was not present on the FARO scan of Merritt’s truck was probably not a light but rather the reflection of a light in the neighborhood mirroring off of a feature on the truck, quite possibly the latch handle to one of the storage compartments.
As the last witness before it rested on March 11, the prosecution called Eugenio Liscio, an authority on three-dimensional forensic reconstruction and an instructor at the University of Toronto and the president of the International Association of Forensic and Security Metrology, to offer testimony with regard to whether Merritt’s truck was a match to the vehicle seen on the Mitchley video. He replaced Rudin as the prosecution’s expert witness on the topic. The jury, which was in no way privy to what had taken place at the February 5 evidentiary hearing, had heard no mention of Rudin at all during the trial and was entirely unaware that the prosecution had substituted in Liscio, who was paid $14,000 for his report and his testimony, to take Dr. Rudin’s place.
Under direct examination by Imes, Liscio said that the low lighting conditions and low resolution of Mitchley’s video camera as well as glare from the bright lights of the vehicle limited his ability to make measurements down to a certain range. Nevertheless, Liscio testified, he found numerous similarities between the vehicle seen on the video and Merritt’s truck. Liscio said “The headlamps are consistent with the 2000 Chevrolet truck” and that the headlight pattern of the Merritt’s Chevrolet truck matched up with the headlight pattern of the vehicle in the video.
Liscio, just as Sergeant Smith before him had, said that the light toward the back of the vehicle seen in the Mitchley video was probably a reflection, most likely from the latch handle on one of the truck’s cargo storage bins. He said this was supported by the consideration that the brightness of the light changed in intensity, becoming more and more dim until it disappeared altogether for four frames of the video before reappearing and gradually growing brighter. “I’m calling it an illuminated spot,” said Liscio. “It doesn’t appear as a constant light.”
Other features of Merritt’s truck were, Liscio said, “consistent with what we see in the video.”
With regard to the profile of the vehicle, Liscio said, “The outline of the bottom of the vehicle seems to be consistent with the shadows and dark spots in the video.”
Because of the relatively poor quality of the video he had to work with, Liscio said that in making his comparisons to the much more precise and crisp images on the FARO scan the consistencies in the images he noted fell “within a reasonable margin of error.” Liscio nevertheless asserted that based upon the available data Merritt’s truck could not be excluded as a match to the vehicle captured in the February 4, 2010 7:47 pm drive-by on the Mitchley video.
Under aggressive cross examination by McGee, Liscio said he was not definitively stating that the vehicle seen in the Mitchley video was Merritt’s truck but merely recognizing that was a possibility that could not be eliminated from consideration by the jury. “I think I said that there are other vehicles that could be, or fall into, this category, so vehicles of a similar nature, same form, geometry, same box at the back [match],” Liscio said. “If there is one that had exhaust pipe and brake lights and headlights and bumper, and somebody put a side light on there, that is possible. And it’s possible it’s another vehicle. So, I’m not saying that this is your client’s vehicle. All I’m saying is that the vehicle in question is consistent with my report and that if there’s another vehicle that looks similar, that is possible. I can’t make an identification and say it is.”
Called to the witness stand this week by the defense, Rudin said that during the course of the time from 2015 until 2019 when he was engaged in his forensic evaluation of the evidence in the case while working with the prosecution, he had consistently expressed that there was a critical need to supplement the computational forensic study he was doing of the available imagery with what he called a live reprojection validation experiment. The methodology he proposed to use in that experiment, he said, was the “gold standard” used to identify and validate the identification of three dimensional shapes of people and objects, including vehicles that are captured on a video. He said such an experiment was needed to cross-validate or reject any computational findings. The live reprojection validation experiment would normally have consisted, Dr. Rudin said, of driving and/or positioning Merritt’s truck through the same space occupied by the vehicle and posed in the same relative orientation to the original surveillance camera along the stretch of Avocado Vista Lane situated between the McStay and the Mitchley homes. The newly captured images would then be compared to the video frames in the Mitchley video. Dr. Rudin explained that live reprojection is usually done with the evidence video frames semi-transparently overlaying the live video, so that the match of the corresponding features is registered, if and when the proper position and orientation of the vehicle is achieved. Since the Mitchley video was made at night, and the captured video only depicts the vehicle front lights, some partial frontal details and a “mystery rear light,” the experiment would then consist of positioning the vehicle along its traveled path in such a way as to match or coincide with as closely as possible all features visible in the video and overlaying the “same features” visible in the Mitchley video, he said. This would entail simultaneously matching the two vehicles’ front lights and the rear latch, using the corresponding light pattern features in the evidence video overlay. If the corresponding features could not be matched simultaneously no matter how the vehicle was positioned/oriented, then the “best” position/orientation would be taken and the disparity error would be computed, Rudin said.  Rudin proposed conducting the experiment using specialized crime scene live reprojection software that has been developed by his scientific team and which was previously used in another criminal case investigation involving a vehicle. Rudin proposed that both the defense and prosecution be present and observe the experimental procedure. “I was shortly after told to ‘stand down’ by the prosecution and ‘to stop all case work,’ and later the defense stated that the prosecution controls the white truck [i.e., Merritt’s truck], and that the truck is not in moving condition,” Rudin told the Sentinel after his testimony was concluded. “Hence, only the prosecution can provide the white truck and the crime scene location availability for the experiment. I personally believe that had both sides sincerely cooperated in the interest of truth, the experiment I proposed is feasible, and it would yield the final word of forensic evidence to clarify: ‘Is the evidence video consistent with the white truck?’”
On the witness stand, under direct examination by McGee, Rudin went over how making a rock solid determination one way or the other whether the vehicle on the Mitchley video did or did not correspond with the images of Merritt’s truck eluded him. At the time of the February 5 evidentiary hearing, he said, he had yet to conclusively reject the possibility that they were one and the same. McGee asked him about his testimony in February.
“I said I could not reject at that time,” Rudin said.
“After that hearing, you kept working, didn’t you?” McGee asked.
“I kept thinking,” Rudin replied.
“And you kept doing further analysis?” McGee asked.
“I just realized mathematics doesn’t lie,” Rudin said. “We were getting short measurements, and I was trying to interpret it in my mind [how] there was the truck on the ground and I realized it’s not the truck on the ground. Then I went and confirmed it by other means.”
Dr. Rudin’s reference to the truck on the ground pertained to the consideration that the rear side running light on Merritt’s Truck was measured by Dr.Rudin to be 5.17 feet from ground  (taken from the available three dimensional truck model), while what appears to be the rear side running light on the vehicle in the Mitchley video is measured by Dr. Rudin to be 4.37 feet from the ground, if the truck is at the curb, or less then 4.37 feet if the rear of the truck is away from the curb. Thus, for the rear side running light of Merritt’s truck to be at the level of 4.37 feet above the street, the back tires of the truck would have needed to be  at least 0.8 feet underground, Rudin said.  This measurement ruled out that the identified light feature on the video could possibly be the rear side running light on Merritt’s truck.
“And that was after you testified in February?” McGee asked.
“It was within a week after that, roughly,” Rudin said.
“Did you inform the prosecution of this?” McGee asked.
“Objection, relevance,” Imes said.
“Sustained,” Judge Smith ruled.
“Violates the court’s previous ruling,” Imes added.
“Sustained,” Judge Smith said.
“Were you told to stop working?” McGee asked.
“Objection;” Imes again registered a protest, “relevance; violates the court’s ruling.”
“Sustained,” Judge Smith said.
“Were you ever called as a witness in the trial before us today?” McGee asked.
“Objection; relevance; violates the court’s ruling,” Imes said.
“I think it’s obvious he wasn’t called before today,” Judge Smith said, providing the jury with the information McGee was trying to get before them.
“You were following the trial after that hearing date, correct?” McGee asked.
“Objection; leading; assumes facts not in evidence,” Imes said.
“Sustained,” Judge Smith said.
“Did you follow what witnesses were called through the trial after February?” McGee asked.
“Not for about a month,” Rudin replied.
“Did you discover that a Mr. Liscio testified in this case?” McGee asked.
“Yes, I did,” Rudin said.
“Did you watch his testimony?” McGee asked.
“Originally, I just listened to it on YouTube,” Rudin said. “There was no visual part available of it.”
“After listening to that testimony, did you call me?” McGee asked.
“Yes, I did,” Rudin said.
“Why’d you call me?” McGee asked.
“For several reasons,” Rudin said.
“Like?” McGee prompted him.
“Number one, I was surprised that something called a live reprojection validation experiment was not done,” Rudin said. “I wanted to see. I had already been turned down by the prosecution on this request in writing and so I knew there was no point in asking them, so I called you hoping that maybe you would see the reason and do that.”
Earlier in Dr. Rudin’s testimony, McGee had sought to illustrate the methodology that Rudin had used by walking him through the analysis he had done of the McStay family’s Isuzu Trooper in seeking to determine if it matched the vehicle on the Mitchley video. That method involved Rudin making an exacting comparison of the two images, the first being a FARO scan of the McStay family’s vehicle. A FARO scan is a three-dimensional photograph which provides precise dimensions of the object depicted. On that FARO scan, Rudin demarked ten feature points of the vehicle, which included the four corners of the passenger side headlight, the passenger side taillight and the two interior corners of the driver side headlight. Rudin thereafter superimposed the Mitchley video on the FARO scan, using a software program created by his scientific team at Cognitech, Inc. that aligned the images in such a way that the lateral angle, attitude [that is, upward-or downward tilt] and distance from which the vehicle in the Mitchley video is reproduced in the angle, attitude and distance from which the Isuzu Trooper was viewed. Early in making the comparison, Rudin indicated, the features appeared to correspond with what he he termed minimal “error,” meaning that to that point, the Trooper could not be excluded as a match for the vehicle seen on the video. In measuring the wheel base, however, Rudin said Isuzu Trooper was  9.15 feet and the vehicle in the video pegged at 9.758  feet.
“So, 9.758 versus 9.15, so there is roughly 0.6 feet, which means just like that it’s not it,” Rudin said.
McGee also noted that the Isuzu’s tail pipe was on the driver’s side of the vehicle, which distinguished it from the vehicle on the Mitchley video. Having shown how Rudin’s methodology for making an exclusion worked, McGee turned to Merritt’s vehicle, which Rudin referred to as “the white truck,” based upon the color it had been painted by 2014 when the San Bernardino County Sheriff’s Department located it and temporarily seized it from its then-owner to subject it to a series of measurements as well as photographic, video and three dimensional surveys and recordation.
Rudin said that as he had done with the Trooper, he had demarked ten feature points on the truck. “That’s what I call a ten point configuration,” Rudin said. In making those demarcations Rudin marked near the top left point of the headlight on the passenger side, the top right point of the headlight on the passenger side, the bottom right of the headlight on the passenger side, the bottom left of the headlight on the passenger side, the passenger side of the front bumper, an electrical plug box on the left side of the bumper, the Chevrolet decal on the grill of the truck, the left top of the driver side headlight, the bottom left of driver side headlight and the light on the rear of the truck on the passenger side.
“You did not mark around the parking lights, correct? McGee asked. “You only marked around the headlight.”
Rudin said he had not “because we can’t see anything like this in the Mitchley image.”
“In the evidence video?” McGee sought to clarify.
“Yeah,” Rudin said.
Rudin said the running lights were not visible on the Mitchley video. Rudin testified that running lights were visible on a video the sheriff’s department made of Merritt’s truck while it was running at night.
On Merritt’s truck, the rectangular headlights are wider than they are tall and are positioned above and distinguishable from the running lights, which are roughly one third as tall as the headlights but somewhat wider than the headlights, extending toward the outer edge of the front of the truck and around the side an abbreviated distance. On Merritt’s truck, the running lights automatically illuminate when the headlights are turned on and cannot be shut off unless the headlights are also shut off. The headlights on the vehicle in the Mitchley video are also rectangular and wider than they are tall, but they have no accompanying running lights.
Rudin testified that the prosecution never showed him the video of Merritt’s truck running at night with its lights on when he had been working for the prosecution on an analysis of whether Merritt’s truck matched the vehicle shown on the Mitchley video.
“You were not provided this video, were you?” McGee asked.
“No,” Rudin said.
In his use of projection mapping, Rudin said he superimposed the rear marker light of Merritt’s truck with the light that is apparent at the back end of the vehicle in the Mitchley video. Rudin said it was possible to have the vehicles line up in the projection mapping and that having the two vehicles occupy the same space in correspondence with one “turns out not to be possible from the point of view of reality.” For Merritt’s truck and the Mitchley video vehicle to have their taillights in the same position on the street in the 3400 block of Avocado Vista Lane, Rudin said, the back end of Merritt’s truck would have to be “pushed underground. I can’t match it. I can’t match it on the ground so I have to push it down. So it turns out we made a measurement later pushing[the truck] one foot down into the ground.”
“If it has to be pushed into the ground for it to fit would it be excluded?” McGee asked.
“It should be excluded, yes,” Rudin said.
McGee explored with Rudin what had driven him to continue to examine the issues relating to the inclusion or exclusion of Merritt’s truck as a match to the vehicle on the Mitchley video after the February 5 evidentiary hearing, at which the prosecution had been given clearance by Judge Smith to proceed with Rudin’s testimony stating that Merritt’s truck could not be ruled out as a possible match to the vehicle that had driven out of the McStay home’s driveway the night of February 4, 2010.
“For many reasons, I was not satisfied with either computational result,” Rudin testified. “It was kind of a borderline on both ends. It was first borderline non-reject and then it was a borderline reject. The error can come computationally, even in the best possible method. There are complications. There are errors. So my data was not perfect. No computation is perfect. So the only way to resolve this in my mind is to do something very simple. Anybody can understand it. If you have a video with a truck and the camera is still there and it hasn’t moved – and if it’s moved we have enough math to fix that – so why not put the truck right there in just the way that we thought it was and find the position so the patterns match or don’t match and we exhaust all the possibilities? We record that and we know, if we’re lucky and we find it, okay then we cannot reject it, of course. So, maybe if we cannot find it [a match], it’s because our driver is not good enough but there are many good drivers out there. We have done those experiments before. That’s called live reprojection. We created a software that helps you find that position, helps the driver go where he needs to go because we know where our computation takes us. So, we can guide the driver. This takes an hour or two, max. In February I understood for the first time that a) the truck is available and b) the camera is still there and it’s possible to go there. Once I knew that, I am a scientist and the basis of science is experiment. The theory should follow by experiment, and in this case experiment would be the crown jewel of this whole process. That’s why I called you.”
“When you wanted to do the reprojection, you were just trying to be thorough, right?” McGee asked.
“Because that’s science,” Rudin said.
“You wanted to be thorough,” McGee said.
“I wanted to be logical,” Rudin said.
“You wanted to be accurate,” McGee said.
“Scientific,” Rudin responded.
“You wanted to be complete,” McGee said.
“Scientific” Rudin said. “That’s the word. If you can do an experiment, you have to do an experiment. That’s science.  It’s like the ABC of science.”
“And when you said you wanted to do that to be thorough and accurate and complete, you were told no,” McGee said.
“In slightly different words,” Rudin said. “’We don’t have time.’ This kind of thing. Somewhere in my emails [from the prosecution] it says. ‘We’re not going to do it.’”
Rudin testified he had listened to Liscio’s testimony and looked at the video element of Liscio’s presentation and concluded parts of it were inaccurate along with his conclusion that Merritt’s truck could not be ruled out as a match for the vehicle seen on the Mitchley video.
Dr. Rudin said that Liscio had created a projection matrix combining a three dimensional daylight image of the laser scanned Avocado Vista Lane streetscape and the daylight image from the Mitchley video and then did a calibration of the camera, which in part was finding the camera’s position and orientation relative to the three dimensional laser scan of the street and then used this camera position and orientation in his video animation simulation of Merritt’s truck positioning. Liscio then compared it, through projection, to the Mitchley security video camera recording. In doing that, Liscio’s projection matrix had failed to properly account for the camera position, evidenced by mismatch of roof overhang above the Mitchley security video camera that obscured the upper part of the vehicle in the video,  Rudin said.
“When he actually determines that position of the camera and orientation of the camera, and I think calibration parameters that go into that of which the main ones are focal length and something called optical center, then he shows the image and then he shows the overhang from the 3D scan and they don’t match,” Rudin said. “Now that’s like a warning flag to any photogrammetrist.  That means there is an error in our projection method. He should have matched those correctly. Somehow he didn’t. He used some software that wasn’t specifically for that.”
When Rudin asked the prosecution for the latest available three dimension laser scan of the street scene in the 3400 block of Avocado Vista Lane that he needed to run an examination with everything calibrated correctly, he was turned down, he said.
“I wanted to prove that and when I asked for the latest laser scan of the street, I was denied to get it,” Dr. Rudin said.
Of the foundation that Liscio had laid for his analysis, Dr. Rudin said, “When I look at this, that’s already wrong. If that is wrong, then everything else is questionable. That’s number one.
Upon going further “down the road” that Liscio had taken more problems revealed themselves, Rudin said. Liscio used a software program called 3D Studio Max to engage in a form of projection mapping that involved what Rudin said was a compromised form of single frame metrology, since 3D Studio Max was not designed for forensic single frame metrology comparison of precise spatial measurements but rather is a flagship three dimension animation software. Rudin said Liscio thereafter “re-created” a sequence of images using the projection matrix, the three dimensional laser scan of the vehicle, and the three dimensional laser scan of the street. “So he basically animates this vehicle and he moves it by hand. He moves it and moves it and moves it and this projection mapping at the same time projects it into the video, and he moves it until he thinks it matches the light. He then draws two circles [around the headlights], and the circles are to show how well you match the 3D model lights [on Merritt’s truck] with the visible pattern of the front lights [emanating from the vehicle in the Mitchley video]. Then he says, ‘Here is our best result’ and then he does some kind of analysis, which I also disagree with, and then he says, ‘Okay, it matches.’”
Rudin continued, “What’s wrong with this? First of all, the mapping is wrong. But let’s assume the mapping is right. What else is wrong? Moving this by hand doesn’t give you the best position possible. He doesn’t have the math. It wasn’t provided to him by his off-the-shelf software. His method does not provide a best possible position, therefore he cannot estimate the best possible error. But you say, ‘Why do I care? If I can position in such a way that everything perfectly matches, then it’s fine anyway.’ If you’re lucky enough in three dimensional space to find the right position that gives you zero error, that’s wonderful. But his error is not zero. He’s drawing these two circles. Nobody draws circles. People use points. The way to measure error is not by drawing circles and saying, ‘What’s the distance between them? Is it preserved?’ That’s what he’s doing. One can be putting circles on the other end of the car and still having the same distance. Then what? Is that a good thing? I don’t think so.”
Liscio’s analysis, Rudin opined was “not thought through very well.”
The discrepancy between the positioning of the lights in Liscio’s comparison of the video to the three dimensional scan of Merritt’s truck was significant, Rudin said. “His circle slided just halfway out of the driver side light,” Rudin said. “Now, that’s a huge error in my book. That would be like, I don’t know, a 15-20 pixel error. If you allow this error, you can match anything to anything.”
Beside finding fault with Liscio’s conclusion that the vehicle seen in the Mitchley video could not be ruled out as matching Merritt’s truck based on the Canadian scientist’s projection matrix error, Dr. Rudin said Liscio’s analysis from the standpoint of photogrammetry multi-point matching theory was deficient as well.
It was strikingly apparent to him, Rudin said, that Liscio never applied his analysis to a video frame that utilized three logical features – the two headlights and what Rudin referred to as the “rear mystery light” and which Liscio called “an illuminated spot” – simultaneously present. Rather, Rudin maintained, Liscio engaged in a matching effort that used a frame lifted from the Mitchley video in which the house’s overhang was blocking the illuminated spot and thus had only the two headlights visible within it. Defining a frame that had all three features as a “complete data frame,” Rudin said “Not using the best information available in the evidence data is in itself forensically indefensible.”
Rudin testified that the photogrammetric analysis that used just the two front lights was “under-constrained,” meaning that the computation lost two equations out of six equations in the photogrammetric matching procedure. With four equations, Rudin said, the two headlights on Merritt’s truck can be almost perfectly matched to the video evidence, with just a small error. By removing the third point at the rear of the vehicle from the analysis, Rudin said, such sparse data is left as to make any conclusion virtually meaningless. In seeking to match three points, Rudin said, there are six equations and three unknowns which are the vehicles’ on-the-ground positions, consisting of the x,y coordinates and the vehicle orientation angle. This system of equations, Rudin said, is far more robust mathematically and will yield a much larger error on the third point when a force-match is attempted with any two other “wrong” points. Using this analytical approach, Rudin said, is more suitable in determining whether Merritt’s truck matches the vehicle on the Mitchely video because if the vehicles are not the same, by forcing the two sets of headlights to match, the illuminated spot would “stick out.”
Rudin pointed out that Liscio in his testimony used the illuminated spot “by itself” to match it to Merritt’s truck’s back storage compartment latch. In doing that, Rudin said, Liscio used another incomplete data frame from the video in which the headlights were not visible.
“First, with one light you can match anything,” Rudin said. “Everybody knows this. With teeny tiny error, you can match any two points. It is bringing the third point” into correspondence simultaneously with the first two points that determines whether the objects actually correspond with one another, Rudin said. When the third point on Merritt’s vehicle – the taillight – was put into synchronization with the taillight on the Mitchley video vehicle, Dr. Rudin said, an impossibility occurred, accordingly to his now applied method of what he calls “height preserving projection with rigid constraints,” which is abbreviated HPP. On April 24, 2019 Rudin provided both the defense and the prosecution with his paper/brief entitled “Photogrammetric Comparison of Vehicles with Height Preserving Projection,” which contained an errors table.
Referring to Merritt’s truck as “the white truck,” Rudin said, “One can use standard photogrammetry method to project each chosen 3D point from the white truck 3D model back into the 3D street crime scene [i.e., the 3400 block of Avocado Vista Lane] through the evidence video, assuming its height is correct, as measured in the vehicle laser scan. If the vehicle holds its shape, we ‘don’t reject.’ If the vehicle ‘shape shifts,’ we ‘reject.’ Shape shift is observed when any pairwise distances change from what they were in the white truck 3D model.”
Rudin demonstrated computational results using his HPP method in a sequence of images with overlaid measurements, which were provided to the defense and prosecution prior to his testimony.
Though the prosecution did not provide him with an updated laser scan, Rudin said, he utilized the one available from 2015. Using height measurements of the “matching points” from the three dimensional laser scan of Merritt’s truck and then photogrammetrically computing where each of those points should land/project independently onto the street in front of Mitchley’s home, Rudin inserted that projection into the Mitchley video, while preserving the respective heights of the truck and the vehicle. This assured that all projected points were consistent, with the vehicle being placed on the road and not in the air or under the ground. All pairwise distances of the projections into the three dimensional street scene were measured and compared to their pairwise distances from the three dimensional laser scan of Merritt’s truck.
Had Merritt’s truck indeed matched the three dimensional scan of the vehicle visible in the night video, the pairwise distances would have preserved their length when they were “relocated” between the three dimensional model of the truck and the three dimensional model of the street, Rudin said, “up to some small computational error.”
Dr. Rudin selected four points on the three dimensional model of Merritt’s truck – three from the front, consisting of two headlights’ corner points and one ‘bumper box’ point, and a fourth point at the center of the ‘rear hinge’ on the three dimensional model of Merritt’s truck.  Since the prosecution stated the rear hinge corresponds to the bright point in the video described by Liscio as the “illuminated spot,” Rudin used this correspondence as the fourth matching three dimensional model video pair.
When the new pairwise distances were measured, the corresponding four pair lengths did not preserve, Rudin said, with disparities observed ranging from six percent to ten percent. Rudin said his method’s geometrical interpretation was that in order to maintain the given heights from Merritt’s three dimensional truck model and keep the truck on the road, the vehicle points “shape shifted” when reprojected to the street three dimensional model, which, Rudin pointed out, “is impossible in rigid objects like vehicles. The disparity range of six percent to ten percent is unacceptably high, exceeding an average expected computational error.”
Rudin provided his HPP measured rigidity disparity errors table to the court as proof of his analysis.
Accordingly, Rudin said “The decision is to reject the match between the white truck 3D model and the evidence video.”
Dr. Rudin’s testimony, which began on Tuesday and lasted to Wednesday, was interrupted twice while the court dealt with two other issues, one which involved testimony before the jury by a rebuttal witness for the prosecution along with another matter that was handled outside the presence of the jury.
Lauren Knowles, the one-time girlfriend of Dan Kavanaugh, was called to the witness stand on Tuesday prior to Dr. Rudin having finished with his testimony, largely because she had to travel some distance to comply with the subpoena for her to appear. Knowles’ relevance to the case stemmed from her relationship with Dan Kavanaugh.
Kavanaugh was another of Joseph McStay’s business associates who had assisted McStay in promoting Earth Inspired Products by utilizing his expertise with computer technology to ensure that the Earth Inspired Products website zoomed to the top of various search engines’ result lists in reaction to web surfers searching for information regarding water features or artificial waterfalls. Judge Smith, prior to the trial, had severely curtailed the way in which Merritt’s defense team was to reference Kavanaugh during the trial, rejecting its request to portray Kavanaugh as an alternate suspect in the killings and pursuing on Merritt’s behalf what in legal terms is called a third party culpability defense. Nevertheless, over the course of the trial, through seeming force of will, creative maneuvering in the questioning of witnesses as well as going right up to the limitations of what they were permitted to broach in terms of references to Kavanaugh and then ultimately proceeding outside the bounds Judge Smith imposed in granting the prosecution’s motions in limine greatly circumscribing what areas the defense could go into relating to Kavanaugh, both McGee and his colleague Raj Maline managed to drag in one piece of damning evidence after another implicating Kavanaugh. Among that evidence was that for roughly a year, Joseph McStay was making payments to Kavanaugh in an effort to buy out whatever interest Kavanaugh claimed he had in Earth Inspired Products so that McStay could reassume control of his company’s website; evidence consisting of internet activity on Kavanaugh’s personal computer in San Diego suggesting that Kavanaugh was actually in Southern California in the crucial late January/early February 2010 timeframe that included the weeks just before and the week of the McStay family murders; testimony from PayPal’s custodian of records augmented by digital data unearthed by a forensic computer analyst showing that six days after the McStay family’s disappearance Kavanaugh hacked into Joseph McStay’s PayPal account on February 10, 2010 to change the password to the account and over the next three days transfer from McStay’s account to his own $7,900; proof that by late February 2010 Kavanaugh commandeered control of Earth Inspired Products; financial records showing that between February 2010 and December 2010 Kavanaugh bled $206,064 out of Earth Inspired Products; and legal documents accompanied by the testimony of the lawyer who drew up those documents indicating that after running Earth Inspired Products into the ground Kavanaugh in July 2011 sold the company, which technically he did not own, for $20,000 and the purchaser’s agreement to indemnify him with regard to the lawsuits pending against the company that had been filed because of Kavanaugh’s failure to deliver projects for which he had received payment. In this way, the defense team had established that Kavanaugh had a motive for killing Joseph McStay. By further suggestions that Kavanaugh was in California at the time of the murders rather than in Hawaii as he had claimed, the defense also insinuated that he had the means and opportunity to have engaged in the murders as well.
On the witness stand Tuesday afternoon, Lauren Knowles was questioned by Supervising Deputy District Attorney Sean Daugherty. She testified that “Dan Kavanaugh was my boyfriend” between March 2009 and the end of their relationship in November 2011. Having established that, Daugherty asked if at any point the couple had traveled to Hawaii.
Knowles said they had gone to the Aloha State in early January 2010 “beginning right after New Years” and that they stayed there “a little over a month,” returning, she said, “in the middle of February.”
She said they were on the island of Oahu and initially stayed in a converted garage but after a week or so had rented a room on the North Shore from a man she identified as “Larry Haynes.”
Knowles testified that during their time in Hawaii they were together the whole time. “I was with him every day,” she said of Kavanaugh.
Daugherty asked about her postings on social media at the time, and thereafter displayed on the courtroom’s overhead monitors some of those postings, made under the name Lo Lo Mi Kno Kno, which included photos.
Knowles identified those as “a couple of pictures I uploaded on Facebook,” which she said were uploaded either on “the same day or the next couple of days” after they were taken.  One indicated a date of January 11, 2010 that referenced North Shore Oahu. She said she and Kavanaugh were depicted in the photo. Another photo posted on January 11 she said was taken by Kavanaugh.
Daugherty then displayed another posting made on February 4 which included a photo of herself, Kavanaugh and three others which she said was taken on the North Shore.
Daugherty asked her about her return from Hawaii. She said she “went with Dan on the same flight” and that Kavanaugh “paid for his own [ticket]. My mom paid for mine.” She said the trip home took place “after Valentine’s Day.”
Daugherty then displayed a boarding pass bearing her name with the date of February 17.
“Did Mr. Kavanaugh return on the same flight with you?” Daugherty asked.
“Yes,” Knowles said.
She said she had last seen Kavanaugh in person in 2012 and that she had blocked him from her Facebook page when he attempted to contact her last year.
During his brief cross examination of the witness, Merritt’s co-defense council Raj Maline pressed her on her claim to have been with Kavanaugh throughout their stay in Hawaii. When she said that was the case, Maline displayed an email she had sent to Kavanaugh on January 28, 2010 in which she inquired, “When are you coming back?”
Knowles said she did not recall the email, and when Maline asked her where Kavanaugh had gone, she said, “It was probably something like the store.”
Maline asked about the resistance the defense team’s investigator had encountered upon trying to communicate with her previously when she was living with her mother in Santa Rosa. She said her mother had told her a detective had come by but that she had not gotten back to anyone because she did not have contact information.
Maline used the opportunity of her being on the stand to ask her, “Did you tell your mother you were afraid of Dan Kavanaugh?”
“He makes me uncomfortable,” Knowles responded.
Maline then angled toward establishing that Kavanaugh’s economic outlook picked up considerably after the McStay family went missing.
“When you were in Hawaii, did you have a lot of money to spend?” Maline asked.
“I never had any money,” she said. “Dan was in charge of the funds, and I never asked him how much money he had.”
With the answers to Daugherty’s questions having already established that the couple was living in a garage during a portion of their Hawaiian vacation and that Kavanaugh did not have enough money to pay for her return ticket, Maline emphasized that further. “Do you recall telling the [sheriff’s] detective [who in 2014 was looking into the family’s deaths] that when you were in Hawaii… he [Kavanaugh] had no money when you lived out there?”
“Yes,” Knowles acknowledged.
Maline went on to Kavanaugh’s financial circumstance once they returned to California. “What about after you got back from Hawaii?” he asked. “Did Dan spend money frivolously?”
“I think so,” Knowles said. “Yes.”
“Did you tell the detective that as well?” Maline asked.
“Yes,” she said.
“Did you wonder where Dan Kavanaugh got all that money when you got back?” Maline asked.
“Objection; relevance,” Daugherty said.
“Sustained,” ruled Judge Smith. “Calls for speculation and conclusion.”
The second interruption of Dr. Rudin’s testimony consisted of a hearing outside the presence of the jury in which Judge Smith considered three motions by the defense which called for the declaration of a mistrial, the recusal of Imes to prevent him from prosecuting the current case further or any future cases against Merritt, and to strike Liscio’s testimony.
Forming the basis of the defense’s motions was an accusation that Imes had engaged in misconduct through the commission of a so-called Brady rule violation. The Brady rule references the 1963 U.S. Supreme Court decision in the case Brady v. Maryland, in which the Supreme Court ruled that the prosecution’s suppression of exculpatory evidence, that is evidence favorable to a defendant, violates the due process standard in the U.S. Constitution.  According to Merritt’s defense team, Imes suppressed information indicating that the vehicle on the Mitchley video was not Merritt’s truck by not explicitly informing the defense in February that Dr. Rudin had come to the conclusion that features and parameters of the video vehicle were inconsistent with those of Merritt’s truck.
Dr. Rudin, who at the conclusion of an evidentiary hearing on February 5 had been cleared by Judge Smith to testify on behalf of the prosecution that he could not rule Merritt’s truck out as being a match for the vehicle on the Mitchley video and was scheduled to go before the jury on February 19, had continued to work on that issue in the interim and was unable to validate a length correspondence between the video vehicle and Merritt’s truck. Rudin conveyed that to Imes, and it is the defense team’s contention Imes was less than straightforward in bringing that to the defense’s attention.
The prosecution, however, contends that in an email Imes sent to McGee between 10 am and 11 am on February 15, the prosecutor informed the defense attorney that Rudin was no longer prepared to testify that he could not reject the possibility of a match and had indicated some order of a problem with pixel space on the imaging of the truck. Imes also informed McGee that the prosecution was not going to call Rudin to the stand on February 19 as was originally planned, but that the scientist might be used as a rebuttal witness. Noting he did not have an understanding of much of what Rudin was saying, Imes invited the defense team to contact Rudin as it deemed appropriate.
Put on the witness stand, Rudin indicated that on February 13th and February 14th while he was traveling in Northern California, he had communicated with Imes by phone, indicating the degree to which he was migrating away from his earlier opinion that the video vehicle and Merritt’s truck could not be distinguished from one another. He testified as well that he sent a much more definitive text message to Imes after 11 am on February 15 in which he more explicitly and forcefully indicated that he had rejected a match between Merritt’s truck and the vehicle on the Mitchley video.
The defense contended that the February 15 email in no case was adequate disclosure to meet the Brady standard of disclosure in terms of what Imes knew at the time he sent the email. Further, the defense maintained, Imes’ February 15 email did not meet the disclosure requirements it was bound by in terms of what Imes was informed of by the text message that was sent to him by Dr. Rudin roughly an hour after he sent the February 15 email to the defense.
After hearing from the defense team, examining the email, seeing the contents of Rudin’s February 15 text message to Imes, hearing testimony from Rudin, and listening to Supervising Deputy District Attorney Mark Vos, who advocated on Imes’ behalf, Judge Smith laid out his analysis of the situation.
Referencing the California Supreme Court’s decision in what he referred to as the Morrison case which relates to the parameters of the Brady requirements in California, Judge Smith noted that “the purpose is for the defense to have that information” but that the California Supreme Court held that the prosecution has no duty to conduct investigation for the defense. “Brady’s purpose is to prevent the prosecution’s ability to suppress evidence, not to provide the accused the right to criminal discovery,” Judge Smith said. “Brady does not displace the adversary system as the primary means to discover the truth, and the defense has the obligation to follow up and obtain information. So, when information is available to the defense at the time of trial or during trial there is no Brady claims for the prosecution’s failure to provide that because the defense has the information if they follow up to get that information. More importantly, the critical aspects of Brady have to do with whether evidence, potentially exculpatory evidence, is suppressed, and on that issue the Supreme Court in Morrison said evidence is not suppressed if the defendant has access to the evidence – they said prior to trial here; it would be during trial – with the exercise of due diligence by the defense.  And they go on to say in any event evidence that is presented at trial is not considered suppressed, regardless of whether or not it had been previously disclosed during discovery. So here, the first indication that Dr. Rudin had changed his opinion or felt there was a potential problem with his opinion was on February 13th or 14th in the telephone conversations with Mr. Imes. Mr. Imes fulfilled his Brady obligations at that point by almost immediately informing the defense that, number one, Dr. Rudin had continued to do further work on the issue since his testimony in the 402 [evidentiary] hearing, that, number 2, Dr. Rudin’s opinions and conclusions changed.  Mr. Imes informed the defense Dr. Rudin indicated that he was unable to validate the overall length of the truck compared to the FARO scan length of the truck, therefore indicating that the length of the truck in the video was different than the length of Mr. Merritt’s truck, which would certainly be a factor which would suggest exclusion of Mr. Merritt’s truck. It did not match. Mr. Imes also informed the defense that in addition to that measurement issue, which Dr. Rudin indicated referenced the issue of the truck being, when it was placed in position where the software could match up with the video in order to make that match, that there was [a discrepancy]. Dr. Rudin indicated there was a problem with pixel space regarding the truck and Mr. Imes did not know the effect of that on the validity of the data. Mr. Imes said, ‘I only have a rudimentary understanding of this. I’m not the best conduit of all of that information, therefore you should feel free to contact Dr. Rudin if you want this explained further. And yes we’re not going to use him in our case in chief,’ which is certainly an indication that his opinion had changed, [and] ‘We may only use him in rebuttal.’ So, at that point the defense is on notice that Dr. Rudin, number one, had done additional work; number two, had changed his opinion; number three, that the change in opinion indicated that Mr. Merritt’s truck would likely be excluded as the truck in the video due to the length of the truck not matching.”
Judge Smith continued, “So the defense had that information. They were free to follow up on that. They were free to contact Dr. Rudin. In fact, there was contact. The reality is they did obtain all of that information, and have Dr. Rudin testifying. In terms of ‘Is there any prejudice?’ I don’t see any prejudice. ‘Was any evidence suppressed?’ No, the evidence was not suppressed. The defense had access to that information prior to resting case, shortly after they began their case. They had ample opportunity to present it, and did present it. That, in my mind, still does not change the fact that Exhibit O, the more detailed text message was, number one, still exculpatory, and number two, should have been provided. But again, there was no suppression of information because the defense was already on notice that there was additional work and a change of opinion. So, while it [Rudin’s February 15 text message to Imes] still should have been provided, there was no suppression and in the court’s view, there was no prejudice. Therefore the motion for mistrial is denied. The motion to recuse Mr. Imes or any other prosecutors is denied. The motion to strike Mr Liscio’s testimony is denied.”
McGee protested Judge Smith’s ruling, contending the defense team had not been given clear notice of the evidence against Merritt consisting of Rudin’s findings early on. Consequently, McGee said,  when the evidence against his client subsequently changed in his favor, the defense was again deprived of adequate notice. This has compromised Merritt’s right to a fair trial, McGee asserted, since the defense team was not armed with the disclosure that Rudin’s opinion had changed before Liscio testified, preventing the defense from using Rudin’s latest findings in the cross examination of Liscio. McGee said the defense had not been given an adequate opportunity to solicit defense experts and review evidence and information prior to testimony. McGee said the district attorney’s office was engaging in a “shell game” that amounted to prosecutorial misconduct.
Countering McGee were Supervising Deputy District Attorney Sean Daugherty and Supervising Deputy District Attorney Mark Vos, who asked Judge Smith for a finding clearing Imes of any accusation of prosecutorial misconduct. Smith refused that request, leaving open the potential of a future sanction against Imes, but said he would not hold any such hearing relating to that now and wanted to get Merritt’s trial proceedings un-sidetracked so that testimony could proceed and the case, now nearing the five-month mark, could go to the jury.
With the jury once again in the courtroom, further testimony was heard from Dr. Rudin.
Ruden testified that “In all model matching methods, whether it is image processing, metrology or anything like physics, you have data on one end and a model on the other end. So, what you try to do is bring as much data into the model, because if you have too little data, anything can match. Your theory can confirm anything. So, you have to use all of it.”
Rudin appealed once more to have the prosecution and the defense support him in his quest to carry out the live reprojection validation experiment he has been requesting all along, saying it would be the surest way to validate once and for all whether Merritt’s truck matches the vehicle seen in the Mitchley video.  He said that based upon the data he now has, he has to reject Merritt’s truck as being a match to the vehicle in the security video.
During his cross examination by Imes, Dr. Rudin was shown one of his submitted exhibits, which included an image of a three dimensional model of Merritt’s truck with the four selected feature points shown and designated as A, B,C, and D, along with another exhibit he had prepared, a height preserving projection reprojection of the points A, B,, C, and D onto the vehicle in one of the frame’s taken from the Mitchley video.
Imes questioned Rudin with regard to the three dimensional model exhibit of Merritt’s truck, proceeding as if it represented the vehicle seen in the video, and asserted that the measurements of the vehicle corresponded exactly to the measurements of Merritt’s truck, either not understanding that the exhibit was intended to show what the  measurements should be if the Merritt truck is the vehicle seen in the Mitchley video, or purposefully blurring the distinction between it and vehicle in the Mitchley vehicle.  This line of questioning created some degree of confusion by suggesting Merritt’s truck and the video vehicle were identically configured and of the same dimensions.
In his redirect examination of Rudin, McGee presented all of the image pairs Rudin had prepared as exhibits, having him explain what each was, in an attempt to clarify the matter.
In cross examining Rudin, Imes elicited from him an acknowledgment that he was functioning at the extreme end of the forensic analysis envelope in applying his live reprojection validation experiment and height preserving projection with rigid constraints methodologies to the ongoing case relating to Merritt’s truck and its comparison to what is on the Mitchley video.
“According to you, those methods have never been used with evidence such as this to do this process, correct?” Imes asked.
“Not exactly like this, yeah,” Rudin said. “But it’s the same method mathematically. It’s the same thing.”
“And your measurement experiments you did since you testified in February, they’ve never been done either, correct?” Imes asked.
“Not exactly like this,” Rudin said.
“Because when the judge asked you a question at the break on the record, you said, quote, ‘It’s a new method I dreamed up,’”  Imes said. “Is that what you said?”
“That’s the way mathematicians talk about it,” Rudin responded.
“So, there is no way to validate or test to ensure that your measurements are accurate, correct?” Imes asked.
“Yes, there is, because point-per-point is exactly the same method that people use for human height, and that was validated,” Rudin came back.
“However, you have no external validation,” Imes said. “No one can recreate your work, right?”
“Sure they can,” Rudin said. “Anybody who has a projection engine  – that’s what I call it, a projection engine – so anybody who has a single view metrology [a tool used for computer-based image modeling and rendering] can now use this method.”
“But no one has reviewed your work in this case to judge its accuracy,” Imes said.
“So, let me explain to you what’s new,” Rudin began.  “I don’t think I’m getting through. What’s new here is that after I do the individual positioning -”
“I object as non-responsive,” Imes interrupted Rudin and appealed to Judge Smith. “I asked him whether there was review of his work.”
“Overruled,” Judge Smith said. “He can explain his answer.”
“There was a review in the sense of accuracy for each point,” Rudin said. “But then I’m saying something that nobody else said before, that wait a minute, it should be a rigid constraint, because if it’s the right vehicle, then the relationship between this point should be preserved. Yeah, that part is new, but it’s obvious, like daylight, right? Vehicles don’t shape shift. We don’t have vehicles like that yet. Maybe we will in the future, but we don’t have that. They’re all made of rigid metal, and so the relationship should be preserved. Nobody observed that before, but if the points are properly located point-per-point, the computing distances between them is a trivial formula. I can write it for you on the board if you want me to, right now. That’s Geometry 101. So, from that point of view, it’s not new, but the realization of rigid constraint is completely new. So, it’s up to you to call it validated or not validated. In my opinion it’s completely validated. I just spoke to a number of people in the forensic community and they think it’s beautiful and it’s right.”
“None of those people reviewed your work to make sure it’s accurate, did they?” Imes pressed.
“Well, I’m starting to send this thing all over the place, including the American Academy of Forensic Science, where it will be introducing extension of this work and so on,” Rudin said.
“Because your goal that you testified to yesterday was if you could make this work, you could retire, correct?” Imes said.
“Well, you see, sir, what you are doing is you are impugning my character and my motives instead of arguing my methods and mathematics,” Rudin said. “That is called ad hominem. It is a very dirty trick, used way back in Roman history.”
“Isn’t that what you said?” Imes asked, ignoring Rudin’s protest. “Did you not testify yesterday, quote, ‘If I could make this work, I can retire?’”
“I may have said something like this, but it’s not the reason I am developing methods,” Rudin said.
The last witness the defense brought before the jury was Sergeant Edward Bachman, who was recalled after previously testifying as a prosecution witness and then being extensively questioned by the defense during the presentation of its case.
McGee asked about Bachman having written a report about his review of the Mitchley video in which he made notation of three important features.
“I found three distinctive things in the video,” Bachman said. The first, he testified, was an electrical plug box on the front bumper. He said he believed that to be a significant element of the video because it was consistent with an electrical box on Merritt’s truck’s front bumper. The other noteworthy items were the vehicle’s rear marker light and its exhaust pipe on the passenger side. Bachman said he used the three features to exclude the vehicle on the video as a match for the McStay family’s Isuzu Trooper, as all three were not present on or in a position consistent with the Trooper.
“You said that is not present on the McStay’s Trooper, correct?” McGee asked.
“Correct,” Bachman said.
Having established that Bachman was fully aware of the rear marker light and where it was positioned on the vehicle seen in the Mitchley video, McGee then angled to determine why Bachman had not then for the sake of logic and consistency applied the lack of a rear marker light in that position on Merritt’s truck to exclude Merritt’s truck as matching the vehicle in the video.
“Why did you not highlight that that light is missing from this side?” McGee asked, referencing the rear passenger side of Merritt’s truck.
“Objection; assumes facts not in evidence, that the light is missing,” said Imes.
“Appears to be missing,” Judge Smith said, amending the question.
“Why did I not note that it wasn’t there?” Bachman asked
“Right,” said McGee.
“I was noting consistencies in there,” Bachman said, a tacit acknowledgment he had ignored, either inadvertently or deliberately, an indication that the vehicle in the Mitchley video was not Merritt’s truck.
Judge Smith did not at first allow McGee to question Bachman about an award the department had been provided by the International Association of Police Chiefs for the department’s work in “solving” the McStay family murder case by narrowing its investigative focus to Merritt and then arresting him in 2014.
After a sidebar conference with McGee and the other attorneys, Judge Smith allowed the questioning to proceed far enough to determine that Sheriff John McMahon, then-Lieutenant Chris Fisher and Bachman had received the self-nominated award on behalf of the department.
The prosecution called Sergeant Ryan Smith, who has testified a multitude of times previously in the trial, this time as a rebuttal witness. Sergeant Smith verified that on February 10, 2010, phone calls from Dan Kavanaugh’s cell phone were placed to the San Diego County Sheriff’s station in Fallbrook at 12:49 pm, 1:09 pm and 1:57 pm. From the same phone on the same day, calls were placed to the San Diego Police Department at 1:03 pm and 1:08 pm, Smith testified.
Smith’s testimony was meant to suggest that relatively early after the McStay family disappeared, Kavanaugh had expressed concern about them.
Thursday, it appeared for a time that Merritt was on the verge of taking the stand to testify on his own behalf. Before he did so, however, his defense team asked Judge Smith, outside the presence of the jury, to consider preventing the prosecution from questioning its client about his criminal record if Merritt did testify. Judge Smith then reviewed Merritt’s arrest, criminal conviction, incarceration and probation record. This included a February 1977 conviction for burglary, a July 1977 conviction on a petty theft charge, an October 1978 criminal trespass conviction, a November 1978 burglary conviction, an April 1985 conviction for receiving stolen property, a February 1987 conviction for receiving stolen property, a May 1988 conviction on receiving stolen property, a June 1988 conviction on a parole violation, and an April 2001 conviction for burglary and grand theft. At the time of the McStay family murders, Merritt was wanted on a probation violation warrant.
Smith, in discussing those convictions, said that if Merritt’s convictions in the 1970s and 1980s when he was a relatively young man “twenty or thirty years prior to 2010 and 2019… were the only convictions and he had a clean record from 1979 or even 1989 onward, maybe those would not be as relevant. But when there is a continuing pattern of theft-related matters which included two state prison commitments and several grants of probation and the conduct continues, I think that is relevant in judging credibility. So, I would allow those to be used for impeachment purposes. My inclination would be that as to allowing facts of those instances in to show common plan, scheme, motive would not apply in this case. The defendant in this case is charged with murder, not the theft of money or property from the McStays. So, a string of similar thefts to show he must be the person who did the theft in this case, assuming it was sufficiently similar,  to do so doesn’t really have much probative value on the issue of the homicides. Obviously, it adds significant prejudicial effect, so my inclination would be to not allow evidence as to the facts of any of those or other underlying instances.”
Thereafter, in a private conference with McGee, Maline and the defendant’s third co-counsel, Jacob Guerard, faced with the prospect of being aggressively cross examined by the prosecution, Merritt elected not to testify. Shortly after making that announcement to the court, the defense rested.