James Keith McElveen — Direct/Cross/Redirect
216 linesJUDGE WHEELER: State may call its next witness.
MS. CAPPLEMAN: State calls Keith McElveen.
JUDGE WHEELER: Keith McElveen, please.
JUDGE WHEELER: All right, good afternoon.
JUDGE WHEELER: Before you have a seat, we're going to swear you in.
JUDGE WHEELER: Please raise your right hand and respond to the clerk's oath.
COURT CLERK: Please solemnly swear or affirm that the testimony you shall give shall be the truth, the whole truth, and nothing but the truth.
JAMES KEITH MCELVEEN: I do.
JUDGE WHEELER: Please have a seat.
MS. CAPPLEMAN: Sir, please introduce yourself and spell your name.
JAMES KEITH MCELVEEN: James Keith McElveen. The last name is spelled M-C-E-L-V-E-E-N.
MS. CAPPLEMAN: How are you employed, sir?
JAMES KEITH MCELVEEN: I'm a forensic engineer.
MS. CAPPLEMAN: How long have you been a forensic engineer?
JAMES KEITH MCELVEEN: About 35 years now.
MS. CAPPLEMAN: What is a forensic engineer?
JAMES KEITH MCELVEEN: We take audio, video, computer data and try to restore it or enhance it as needed.
MS. CAPPLEMAN: What training and experience do you have in this area?
JAMES KEITH MCELVEEN: I have given or received over 20 courses in speech enhancement, in audio restoration, and forensic audio.
MS. CAPPLEMAN: All right. And during the course of your employment, have you worked for the federal government?
JAMES KEITH MCELVEEN: I have.
MS. CAPPLEMAN: And you currently are self-employed, is that right?
JAMES KEITH MCELVEEN: I work for Wave Sciences, a company that I founded.
MS. CAPPLEMAN: Were you asked to clarify the audio from a restaurant recording in the case that we're here about today?
JAMES KEITH MCELVEEN: I was.
MS. CAPPLEMAN: What were you provided with in this case?
JAMES KEITH MCELVEEN: Thumb drives with audio and video evidence.
MS. CAPPLEMAN: Okay, and were they from two separate devices from a restaurant recording?
MS. CAPPLEMAN: And what device — was there one device that was particularly better audio-wise?
JAMES KEITH MCELVEEN: Yes, there were. One device had clear-looking video in the file; the other one had a mesh cover to the video camera. The one that had the mesh cover was the much better audio capture.
MS. CAPPLEMAN: So you're with better video and one with better audio, is that right?
JAMES KEITH MCELVEEN: Exactly.
MS. CAPPLEMAN: So your focus is the audio, correct?
JAMES KEITH MCELVEEN: Correct.
MS. CAPPLEMAN: So you didn't do anything with the video as far as clarifying the video or doing anything with that?
JAMES KEITH MCELVEEN: That's correct.
MS. CAPPLEMAN: All right. So let's talk about the audio. So you focused on the second device that had the better audio recording. What types of — I mean, was it one file or multiple files that you were provided with?
JAMES KEITH MCELVEEN: We were provided with multiple files. Originally was one huge file which was too much, too big for our tool to handle without choking, so we then asked for it to be broken up into three pieces.
JAMES KEITH MCELVEEN: And the pieces we asked to be overlapped so there would be no chance that something would be missed and word left out here or there.
MS. CAPPLEMAN: All right. So this is approximately a 45-minute recording to begin with, is that right?
JAMES KEITH MCELVEEN: That's correct.
MS. CAPPLEMAN: All right. And we've separated it into what we're calling beginning, middle, and end segments?
MS. CAPPLEMAN: Okay. And there's some overlap between the segments to make sure nothing gets lost?
JAMES KEITH MCELVEEN: Exactly. I think it was about two minutes per segment of overlap.
MS. CAPPLEMAN: What can you tell us about the quality, or the issue of the quality, on these recordings — or this recording?
JAMES KEITH MCELVEEN: The audio quality was very poor due to distance, due to noise, and due to the configuration of the equipment. The location changed somewhat. The place that second recording device was positioned was on a bench, and I believe it was leaning against the operator of the equipment, and therefore it would rub against his leg or was propped against his leg, from the best I could tell from looking at the video and listening to the audio. So that created pointing issues as well as rubbing noise issues. There was music going on, there was — the way that the device was pointed, it wanted to pick up noises from the kitchen.
JAMES KEITH MCELVEEN: If and when the video from device 2 is viewed, you'll see, behind the mail that it's pointed generally toward, that there's a swinging kitchen door, which is very thin, and the dishwashing sink was right beside the door. So you get a lot of dish clatter. There was also like coffee grinder and blender noises and all kind of things going on.
MS. CAPPLEMAN: All right. Um, could you explain for the jury the processes that you employed to try to clarify or improve the audio of the voices on the recording?
JAMES KEITH MCELVEEN: Initially we used the standard techniques that have been practiced in forensic audio for the last 30 years. They yielded minimal improvement — some improvement, but not significant. So then we used our own tool that I mentioned earlier, and it works in a substantially different way. Instead of listening in a direction or trying to learn the noise, it tries to learn the person who's talking, and it works to focus in all the sound on the recording on that person, that person's location.
JAMES KEITH MCELVEEN: So the easiest way to think about it is like with a camera, and you have a camera lens.
JAMES KEITH MCELVEEN: So if I focus the camera on your face, things close and far away are going to be all blurry, but your face is going to be sharp. And that's what our tool does — it focuses in on that person and then blurs out and reduces the rest.
MS. CAPPLEMAN: And did you attempt to focus in on the person of both Charlie Adelson, as well as — well, I should say the male speaker as well as the female speaker on the recording?
JAMES KEITH MCELVEEN: Exactly. In each of the three segments, we had the set the tool up independently for each segment. And then in some segments there just wasn't much voice to deal with. In segment one, the beginning, there's a lot of noise. There's a lot of motion. The operator has not moved very close to the male and the female. And there's not much voice there to get a focus lock on.
JAMES KEITH MCELVEEN: Starting in about seven and a half minutes into the first segment, the positioning changes and it gets a lot more clear, and we could begin getting a lock. And by about — I think it's 17 and a half minutes in — they finally move and a much better lock is capable of being achieved, at least on the male. The female did not get a very good lock at all in the first segment. In the second segment we were able to get a lock on the female for the first time, and the lock on the male was still good.
MS. CAPPLEMAN: All right. So was there a significant difference in the quality or ability to clarify the male voice versus the female voice?
JAMES KEITH MCELVEEN: Yes, there was. The male voice was stronger, partly because of being masculine and a stronger voice box, partly because the male was talking toward — in the general direction toward — the device that's capturing the audio. The female has a softer voice, um, doesn't talk nearly as much, so there's fewer opportunities to glimpse her voice, and then was talking in the general direction away from the recording device. And I would like to just continue one point. Um, you can — once you've listened to the different segments, the male repeatedly, eight or more times, says "What was that?" And he was only sat two and a half feet maybe, when their heads are close together. So he was having difficulty hearing her with her talking in his direction. So the recording device was in a very disadvantaged location.
MS. CAPPLEMAN: All right. Even though the recording devices were at the next table.
JAMES KEITH MCELVEEN: Yeah. It just kind of worked out that way.
MS. CAPPLEMAN: Exactly. And did you make some efforts to actually go down to this location and take some measurements of the way sound operates within the room where these recordings were made, to try to assist your clarification efforts?
JAMES KEITH MCELVEEN: I did.
MS. CAPPLEMAN: Could you tell us about that?
JAMES KEITH MCELVEEN: We went there, to the restaurant location, and made a good number of recordings over the period of two days, trying to get enough of kind of post-event forensic recordings to see if they could be used. But we were incapable of getting more than three or four additional words that were just "I'm going to the bathroom" type of utterance — well, actually not that — but due to the fact that many things have changed, both in the environment and in the recording equipment in the intervening years.
MS. CAPPLEMAN: I want to show you what I've marked.
JAMES KEITH MCELVEEN: I do. It has my marks on it.
MS. CAPPLEMAN: All right. And what does the exhibit contain?
JAMES KEITH MCELVEEN: This is the recordings from the devices that have been enhanced, and that I returned to your office.
MS. CAPPLEMAN: All right. So does this exhibit contain your best efforts in reference to the voices then in the restaurant?
JAMES KEITH MCELVEEN: That it does.
MS. CAPPLEMAN: And is there any part of your processing that actually interferes or changes the voices themselves or the words that are being uttered?
JAMES KEITH MCELVEEN: Absolutely not. The way that our tool works is you configure it, you input the audio, you hit process, and then you go away. There's no ability for it to add, to delete, or to otherwise edit any of the content. The operator has no ability to make any changes in real time while it's doing its processing. You just get what comes out the end. And then our forensics procedures that we practice in our laboratory require us to follow a certain workflow, so we have a procedure that's gone through certain steps that are executed the same way for each recording.
MS. CAPPLEMAN: All right. And are all of those steps and processes the standard in your field?
JAMES KEITH MCELVEEN: Yes, they are.
MS. CAPPLEMAN: Judge at this time I'd ask to introduce State's Exhibit 115.
JUDGE WHEELER: All right. 115 will be admitted into evidence subject to any previous objections.
MS. CAPPLEMAN: No further questions.
JUDGE WHEELER: Cross-examination?
MR. DECOSTE: Sure, thank you.
MR. DECOSTE: How you doing? I hope you had a good flight down.
JAMES KEITH MCELVEEN: Yes, thank you.
MR. DECOSTE: Just a couple of questions; you'll be heading back to the airport. So Ms. Cappleman asked you about your best efforts. Let's go through the timeline. This recording is from 2016, correct?
JAMES KEITH MCELVEEN: Correct.
MR. DECOSTE: The government — they just came to you a couple of months ago, correct?
JAMES KEITH MCELVEEN: That would be the second time, and when we received the entire recording set, yes.
MR. DECOSTE: All right. So into 2022, you get — you get handed to you in a — and again at a gas station parking lot halfway between where you're from in Florida, right? And you get handed something that — the raw data.
JAMES KEITH MCELVEEN: The raw data for device 2, yes.
MR. DECOSTE: And you're given a very short time frame to work on these items, correct?
JAMES KEITH MCELVEEN: Um, I think that perhaps you're confusing something else in the deposition. The first delivery that we did was using the standard techniques, and that was what we were in a rush to provide before the deposition. We had plenty of time to do this second one where we refocus the audio.
MR. DECOSTE: Okay, so let me ask that differently and we're on the same page. You just provided that recently, correct? Not that long ago — I don't remember the date exactly. We're not talking years; we're talking like weeks and months, right?
JAMES KEITH MCELVEEN: Correct.
MR. DECOSTE: And there was a pressure to get this done within a certain time frame, correct?
JAMES KEITH MCELVEEN: I can't say that I felt pressured to meet any specific date.
MR. DECOSTE: Fair enough.
MR. DECOSTE: So you were asked questions about the first segment, right?
JAMES KEITH MCELVEEN: Yes. There's three total segments.
MR. DECOSTE: In the first segment, which would be the introduction, the beginning, you can't make anything up?
JAMES KEITH MCELVEEN: We gave it — we provided our best cleanup.
JAMES KEITH MCELVEEN: I would say that if you listen carefully you can recover words, but it's not as good as the situation that existed by the time that the operation had progressed into what is captured in segments — three, two and three.
MR. DECOSTE: Now, with respect to the equipment that was used to make the recording, you're aware of that, correct?
JAMES KEITH MCELVEEN: I don't understand the question.
MR. DECOSTE: The microphone that was used was a directional microphone — it has that capability. And your understanding of the one that was used here and how it was used is that it was pointed at the male, not at the female?
JAMES KEITH MCELVEEN: It is my assessment as a forensics examiner based upon looking at the video and listening to the audio. But I should point out that the video was a fisheye video, so it's distorted — sort of like when you're looking at somebody outside of your door, that type of view.
MR. DECOSTE: Yes. Okay. And that makes it very difficult, unless you have, um, special training in doing the undistortion of that type of, um, camera image, to be able to say exactly where it was pointed. All right. But based on the audio that you've reviewed, is it your belief that it was focused on the male and not on the female?
JAMES KEITH MCELVEEN: I would not use the word "focus." I would say that the device was turned toward that table, but the way that the audio pickup — and because of the acoustics I explained earlier, with the male talking more in the direction of the recording device and the female talking voice is propagating away in the general direction — the net effect was that he ends up being the prominent voice.
MR. DECOSTE: Okay, we can agree on that — that what is more prominent is the male voice, correct? All right, now let's talk about Dolce Vita. You recently went down there and you took a look at the restaurant, right?
MR. DECOSTE: So it's not a big restaurant, is it?
MR. DECOSTE: Is it bigger or smaller than — well, so from the bar there, from the judge all the way back to these tables to the jury, um...
JAMES KEITH MCELVEEN: By volume, it's a little bit bigger.
JAMES KEITH MCELVEEN: It's longer.
MR. DECOSTE: But not much bigger?
JAMES KEITH MCELVEEN: Not significantly bigger. From you to where?
JAMES KEITH MCELVEEN: Are you talking about from the front door to the very back at the kitchen door, or what dimension are you asking about?
MR. DECOSTE: Wherever a patron can go.
JAMES KEITH MCELVEEN: A patron is not going to walk into the kitchen.
JAMES KEITH MCELVEEN: If I'm positioned roughly at the front of the establishment and the first table — as they were configured when I went, which is not necessarily how they were configured during the original operation — then I would say that the distance to the kitchen door was probably about the third to fourth row of bench seats in the rear.
MR. DECOSTE: Okay, so you're at the entrance, patrons, the third and the fourth row right there — that's going to be the back of the restaurant.
JAMES KEITH MCELVEEN: This is just my rough estimate, yes.
MR. DECOSTE: In the width of it, it's not terribly wide.
JAMES KEITH MCELVEEN: For the patrons' seating area, yes.
MR. DECOSTE: So if somebody were to be sitting in the back — I mean, it's not like there's a back that's like all the way in the back of a restaurant, in a large restaurant, where somebody could be hiding and talking, right?
JAMES KEITH MCELVEEN: Correct. It's a small, intimate place.
MR. DECOSTE: Now, you had said something on direct, and I just want to understand it correctly, where you said you knew the distance between the female, Ms. Magbanua, and Mr. Adelson — that you knew their distance, but you weren't there when the recording was made, obviously, right?
JAMES KEITH MCELVEEN: I was not there when the recording was made, but on the clear video then, you can see the male leaning forward and back in his chair.
JAMES KEITH MCELVEEN: Although the chairs were obviously different chairs than were there during the original recording, the tables looked the same.
MR. DECOSTE: But you don't know. I mean, it could have been a table that was this wide or a little bit wider.
JAMES KEITH MCELVEEN: Yes, and I gave an approximate distance.
JAMES KEITH MCELVEEN: I was not trying to say that it was exactly two and a half feet.
MR. DECOSTE: All right. And to help the jury, for what basically you're engaging in — you've got Mr. Adelson sitting on this side of the table, and then on the other side of the table, however wide it is, you have Ms. Magbanua, when they're having a conversation.
MR. DECOSTE: Can we agree on that?
MR. DECOSTE: I think I'm done. Give me one second.
MR. DECOSTE: Thank you so much. Good seeing you.
JUDGE WHEELER: All right. Redirect.
MS. CAPPLEMAN: So the thumb drive that I showed you, that you etched your initials into — this is your final audio product, correct?
MS. CAPPLEMAN: Okay. And does this contain video as well, or no?
JAMES KEITH MCELVEEN: I would have to go back and look at it again, but I believe — I returned a thumb drive, I believe, for the final process, did not contain video. The first delivery did contain video, where we used the standard enhancement techniques.
MS. CAPPLEMAN: Okay. I do not believe that's — that's not the thumb drive you have in your hand?
JAMES KEITH MCELVEEN: The thumb drive I have in my hand is the final product.
MS. CAPPLEMAN: Right, that's correct. Okay. And that's the final best effort on the audio?
JAMES KEITH MCELVEEN: It is.
MS. CAPPLEMAN: Okay. But you have also had an opportunity to review the exhibit that my office prepared with your best effort audio, as well as the video from the equipment that was in the restaurant that we talked about earlier — that had the better video?
JAMES KEITH MCELVEEN: I have.
MS. CAPPLEMAN: Okay. So those two things were put together. And is that a fair and accurate exhibit of the video that you were initially provided with, along with your best effort audio?
JAMES KEITH MCELVEEN: Yes. I reviewed the video with our audio put in it, and I checked two things: one, that the audio seemed to be the same audio, and it was; and two, that the timing of where the audio was inserted relative to the movements in the video were synchronous — in other words, when somebody's hand made a motion to emphasize an utterance, then the word comes up; when the mouth moves, the word comes up.
MS. CAPPLEMAN: Okay. And did you initial, or do anything to initial, that particular item of evidence that has both on it — the demonstrative?
JAMES KEITH MCELVEEN: No, I did not.
MS. CAPPLEMAN: Okay. I'm going to ask for a brief recess at this time so the witness may review the exhibit and tell us whether it is, in fact, his audio on the video as discussed.
JUDGE WHEELER: And is this on Exhibit 115 or 160? Is it on 115?
MS. CAPPLEMAN: No, sir, it's 114.
MS. CAPPLEMAN: I'm going to need a short recess, Judge.
JUDGE WHEELER: All right. Sorry. Okay. We're going to take a short recess then, and please take the jury out, Deputy.
JUDGE WHEELER: All right, jury's out of the courtroom. The door is closed. Please be seated.
MS. CAPPLEMAN: So I think what I've done, Your Honor, is introduced the final product, which I intended to introduce with this witness, along with his final product, which is 115. That's correct.
JUDGE WHEELER: Is 115 just the audio?
MS. CAPPLEMAN: Yes.
JUDGE WHEELER: Okay.
MS. CAPPLEMAN: That's the one thing I am confident of. 114 should have been marked 116, which I think this witness is about to tell us is —
JAMES KEITH MCELVEEN: I need an adapter. I need to borrow a different laptop; mine will not handle this one.
JUDGE WHEELER: 114 was admitted in through the special agent.
MS. CAPPLEMAN: Yes. I admitted the wrong item.
MS. CAPPLEMAN: And that's the real one where I did show him the exhibit outside the courtroom and actually play it for him, so I apologize for the confusion. But if this witness is able to say "yes, it is my audio on the video," it should be 116 — but this is the thing I previously showed Bronstein and called 114.
JUDGE WHEELER: So this is what you're calling what on this exhibit list? Dolce Vita composite?
MS. CAPPLEMAN: Yes, sir.
MS. CAPPLEMAN: I basically just didn't show him 114. I showed him — once I showed Bronstein 116.
JUDGE WHEELER: All right. 114 is original surveillance, raw data, correct. And I do have that exhibit, which we can introduce now by stipulation, or I can recall Bronstein and do it again.
MS. CAPPLEMAN: Basically just had two exhibits swapped, Judge. Since they're all just thumb drives...
JUDGE WHEELER: So 114, the surveillance — you showed him 116. You'd need to bring him back to show him 114, which would then be admitted subject to any other objections. We can have that stipulation then, or, if the State's going to recall the witness.
MR. DECOSTE: Your Honor, we're discussing it here. We fear that there's, you know, the off chance there's a mistake — something goes to the jury — that they go through whatever steps they need to go through to ensure that it is the exact item.
MS. CAPPLEMAN: This is 114, what should have been introduced as 114, and what I would introduce as 114 with Bronstein is currently playing right now, so the defense can check that out. It's the same item that was previously entered in the previous trial through Bronstein. Got it out of the clerk's office, remarking.
JUDGE WHEELER: Okay, so basically what we're doing right now is — 114, which was shown to Bronstein, which 116 was shown to him — 114 is withdrawn at this point unless it's stipulated to by the defense, correct?
MR. DECOSTE: Your Honor, I normally would stipulate because we've wanted to move smoothly, but we do not want to take the responsibility for any mistake that may have happened.
JUDGE WHEELER: Well, we're looking at 114 right now. Do you want to look at what 114 is, or we'll go ahead and recall — I mean, I don't know what's normal and why this is abnormal, but the State will just recall the witness. Do you want to make sure that he's remained outside?
MS. CAPPLEMAN: He's still here.
MS. CAPPLEMAN: This disc is the same as the initials in the last trial. The initials are still on it, so I just showed him the wrong thing.
JUDGE WHEELER: All right.
MR. DECOSTE: Your Honor, just so Your Honor knows, the issue is that we've had so many enhancements in this, I can't just look at it and say "that's the right thing."
JUDGE WHEELER: This is the raw data. This is not the enhancement.
MR. DECOSTE: I understand, and I can't just look at it and know that that's the raw data.
JUDGE WHEELER: She can show it to you.
MR. DECOSTE: But there's no way for us to know if that's the raw data or one that was minimally enhanced. There's so many different versions of it.
JUDGE WHEELER: How — okay, then he'll have to identify it then. It's fine.
MR. DECOSTE: And then with respect to the disc — I mean, of course I'm not saying that the government did this, but could there have been a situation that the disc is different than the envelope that has a signature?
JUDGE WHEELER: Now you're making all kinds of things up, Mr. —, because there's been no allegations of that.
MR. DECOSTE: And I'm not making it — and I'm not making it, but obviously they had some things that were disorganized, and we don't want to take — we don't want —
JUDGE WHEELER: The one mistake that was made was the wrong exhibit was shown to the wrong witness — and so it was a correct witness but it was the wrong exhibit, so the State will recall. Let's go ahead and — what do you need to make sure? What do you need to make sure now with this witness?
MS. CAPPLEMAN: I need this witness to agree that 116 is his audio coupled with the video that he was originally provided. All right.
JUDGE WHEELER: All right, so as you do — all right, you — why don't you go ahead and ask some questions about that.
MS. CAPPLEMAN: Proffer him?
JUDGE WHEELER: Please.
MS. CAPPLEMAN: All right. Um, Mr. McElveen, have you had an opportunity to review State's 116?
JAMES KEITH MCELVEEN: I have. What State's 116 contains, to the best of my ability in the current situation, is the original video that I was sent by your office, coupled with the audio that we returned for Device 2.
MS. CAPPLEMAN: That would be the proffer.
JUDGE WHEELER: All right. Do you have any questions?
MR. DECOSTE: No.
JUDGE WHEELER: All right. Okay.
JUDGE WHEELER: So we can bring the jury back in. You'll ask for it to be admitted through this witness, I'm gathering, correct? And then when you're done with this witness, then we will bring in the other witness.
JUDGE WHEELER: Be seated. Thank you for your patience. We're ready to continue with the testimony. And Ms. Cappleman, we're ready to proceed with your continuation of the redirect.
MS. CAPPLEMAN: Thank you, Your Honor. Mr. McElveen, have you had an opportunity to review State's 116?
JAMES KEITH MCELVEEN: 116, I have.
MS. CAPPLEMAN: And does 116 fairly and accurately contain a copy of your best efforts on the audio, as well as that video that we talked about coming from the other device?
JAMES KEITH MCELVEEN: Are you calling this one 116? Okay.
JAMES KEITH MCELVEEN: Then yes, I have. Sorry, one of us was confused.
JUDGE WHEELER: Any objections other than those previously made on the record?
MR. DECOSTE: No, Your Honor.
JUDGE WHEELER: All right, it will be admitted as 116.
MS. CAPPLEMAN: All right, just to clarify: 115 your final product just audio, and 116 your final product plus the video.
JAMES KEITH MCELVEEN: Correct.
MS. CAPPLEMAN: Thank you.
JUDGE WHEELER: All right, we can excuse the witness?
MS. CAPPLEMAN: Yes, sir.
JUDGE WHEELER: All right, thank you, sir. You're free to go, have a good day.