Audio Experts

Status
Not open for further replies.
I'm personally skeptical of this technology. Not sure how something can be determined if you have no sample upon which to compare to the recording.

You make a good point. I'm confused. Did they have a Trayvon sample or not? They're saying that they need more sensitive ('quality') headphones to determine the screamer? And then the note about discontinuing 'the preparation of an audio event chronology of specimen Q1'. What the heck? I don't have any idea what that means.
 
You make a good point. I'm confused. Did they have a Trayvon sample or not? They're saying that they need more sensitive ('quality') headphones to determine the screamer? And then the note about discontinuing 'the preparation of an audio event chronology of specimen Q1'. What the heck? I don't have any idea what that means.

I'll post a pretty in depth post in a bit, but to answer your first question quickly: even if they had, the source (911 tapes) were not of quality to do comparisons.
 
I'll post a pretty in depth post in a bit, but to answer your first question quickly: even if they had, the source (911 tapes) were not of quality to do comparisons.

Thanks. I did understand that part. It's stuff at the end, the gobbledygook that has me stymied. At first they're talking about getting 'quality' headphones and then an April addendum that I don't remotely understand. Looking forward to a deciphering.
 
Thanks. I did understand that part. It's stuff at the end, the gobbledygook that has me stymied. At first they're talking about getting 'quality' headphones and then an addendum that I don't remotely understand. Looking forward to the deciphering.

They couldn't do a comparison for the same reasons we already discussed. Voice on 911 tape is under "great emotion." Not enough separate words on 911 tape to make an accurate comparison. The fact that the voice is on the background and another person is talking to 911 during that same time.
 
They couldn't do a comparison for the same reasons we already discussed. Voice on 911 tape is under "great emotion." Not enough separate words on 911 tape to make an accurate comparison. The fact that the voice is on the background and another person is talking to 911 during that same time.

Like I said I got that part. It's clear they were having trouble assessing any bit. They recommend quality headphones. But then an agent comes in in April and withdraws an assessment using gobbledygook. That's the part I want to understand. Unless you are saying that's what they meant. I don't understand what that line I copied means.
 
Like I said I got that part. It's clear they were having trouble assessing any bit. They recommend quality headphones. But then an agent comes in in April and withdraws an assessment using gobbledygook. That's the part I want to understand. Unless you are saying that's what they meant. I don't understand what that line I copied means.

Special Agent Majeski advised telephonically on April 2nd, 2012, to discontinue the request for preparation of an audio event chronology of specimen Q1.

He basically stated they no longer needed them to do comparisons. Agent Majeski is the Tampa Division FBI agent that originally requested the comparisons. This report basically says "He asked for it, we gave it to him, he thanked us and went on with his buisness" - in very paraphrased words.
 
I'll post a pretty in depth post in a bit, but to answer your first question quickly: even if they had, the source (911 tapes) were not of quality to do comparisons.

Just wanted to follow up on that post. No tv/internet as of my last post last night and the cable company is giving me the run around, something about construction in the area. As soon as I have a way to post that doesnt include my phone, I'll post it.
 
Alright, cable company got me back in business this evening. I'll paste what I said about Mr. Owen from another thread here so it's easy to find, then I'll detail the FBI report. I still haven't done much research on Mr. Primeau.

I'm not certain if this has been mentioned, but according to Mr. Owen's website his voice comparisons are bunk.

Website: http://www.owlinvestigations.com/article1.html

Specifically:
Tom Owen's Website said:
III - THE METHOD OF VOICE IDENTIFICATION


The method by which a voice is identified is a multifaceted process requiring the use of both aural and visual senses. In the typical voice identification case the examiner is given several recordings; one or more recordings of the voice to be identified and one or more recorded voice samples of one or more suspects. It is from these recordings the examiner must make the determination about the identity of the unknown voice.


The first step is to evaluate the recording of the unknown voice, checking to make sure the recording has a sufficient amount of speech with which to work and that the quality of the recording is of sufficient clarity in the frequency range required for analysis. The volume of the recorded voice signal must be significantly higher than that of the environmental noise. The greater the number of obscuring events, such as noise, music, and other speakers, the longer the sample of speech must be. Some examiners report that they reject as many as sixty percent of the cases submitted to them with one of the main reasons for rejection being the poor quality of the recording of the unknown voice.
<snip - read all of the information, please>

According to his very first step he could not make such an analysis of these screams. There wasn't any speech, the quality was via telephone (which alone poses many hazards to higher frequencies*), under the speaking of the person on the phone/dispatcher, from an unknown distance. Given that he states specifically that the more "obscuring events" the longer the speech has to be, and the fact that there is no speech.. how can he make any determination at all? There's more, that's just the first step.


Edit below:
I just wanted to make it clear that I'm specifically talking about Mr. Owens' analysis of the screams. I know the special prosecutor probably recorded Mr. Zimmerman in the walkthrough when they had him yell a number of times. In Mr. Owens' analysis he states that he compared the screams to the voice (Mr. Zimmerman's) heard in the 911 call.

* - I tried to find info on this to link, but it seems as if the digital age has fixed this in modern telephone systems and in turn made it a bit more difficult to find info on it. I would place a wager that because the 911 telephone system has some redundancy systems in place it hasn't been updated to digital lines. If you remember "back in the good ol' days" when you were on the phone, the voice of the person you were speaking to would sound different than it does when you're speaking in person. This is because the telephone back then filtered out frequencies above a certain point.

csziggy said:
The software Mr. Owens used for his analysis, Easy Voice Biometrics, is recently introduced for sale in the US, though it has already been accepted for use in court.
Which is just a highly "dumbed down" version of CoolEdit, in my opinion - based upon the tech info for the software found here: http://www.easyvoicebiometrics.com/d...tionalinfo.pdf

Looking over that, there are 3 methods used by the software:

NOTE Edited into this comment: We know from the FBI's findings that there is less than 4 seconds worth of screams that could be used for identification process.

Spectral-Formant Method provides high reliability of identification results and has the following
advantages:
• Requires just as little as 16 seconds length of speech sample.
...

Pitch Statistics Analysis Method.
...
• Requires the minimum of 10 seconds length of speech sample (which is even less than the Spectral-Formant method).
...

NOTE Edited into this comment: The above method is what is used for comparing emotional states. If anything Mr. Owen should have used that one, according to the documentation of the software.

Gaussian Mixture Models based Method (GMM).
...
• High speed of the speaker search.
• Ideal for clear recordings with low noise level.
• Ideal for long recordings.

...

I think it's safe to say that not only does Mr. Owens' own words contradict his analysis, but the documentation from the software he uses says he shouldn't have even tried.

FBI's analysis: They said what Mr. Owen should have - in fact, I wonder if they use the same software that he does (or have a version of software based on it). I don't see how much more straight forward it could have been. They could not determine the word after "fu**ing" - which Mr. Owen (I believe) was the one that said it was "punks". They did not have the longevity to make any conclusions based on the poor quality of the audio sampled against Mr. Zimmerman's screams from the walk through that investigators did with him. This brings me back to Mr. Owen. He did not have those screams recorded after the fact, his tests were based on the 911 call audio that Mr. Zimmerman placed. Once again, I say that Mr. Owen isn't just in left field, he's so far past left field that he can't even see the stadium.
 
I haven't done too much digging into his background or where he got his certifications/etc, but one thing that has caught my attention is that when he did his interview he said he used "critical listening skills" and I wanted to find his definition of "critical listening skills." Here's what I found on one of his websites:

http://www.audioforensicexpert.com/audio-forensic-information/audio-authentication/
Audio Authentication is an art and a science that requires three separate processes. The first is critical listening skills of the audio recording to detect any abrupt changes in sound levels, background noise or other unusual sounds in the recording that may be interpreted as possible edit points, anomalies or alterations.

Couple things to point out:
He states numerous times that he only used his ears and listening to the calls. Steps two and three (not quoted above but can be found at the URL above) both require computer assistance.

Make your own opinions, I know what mine is.

Edit: This thread is making me feel like it's a sounding board for myself.. does no one else have any opinions on the audio experts? :\
 
I don't have a background in this area, but I'll offer a couple of observations based on some subject research and case documents:

- mammals have widely varying speech pitch based on stress, physical exertion, adrenalin release etc. to the point that I've seen audio 'experts' misidentify the SPECIES of mammal making the sound, much less an individual.

- various key factors like accent, quality sample, controlled word sampling, and most importantly a normal speaking voice baseline are not available in this case

- if anyone is going to identify a stressed call for help... it's a family member... and T's father, when interviewed the first time (before the case got rolling and all the media came to play), said that the recording was definitively NOT T... also G's father the same day said definitively the recording was definitively G

so, I do not believe that the base technology, when applied in this manner, is nearly accurate enough to rely on as a evidence point to convict someone for murder.

further, I would tend to believe T's father in his assertion that T was not the one crying for help.
 
- mammals have widely varying speech pitch based on stress, physical exertion, adrenalin release etc. to the point that I've seen audio 'experts' misidentify the SPECIES of mammal making the sound, much less an individual.

This made me laugh because it reminded me of some audio expert making claim that the sounds were from "bigfoot." Thanks for the laugh!
 
13 posts into this thread and no links to audio of 911 calls or to anything related to the "audio experts."
 
So does this effectively end their career. If they find themselves in another court case, can the attorney that is cross examining them bring up their lack in judgment in evaluating this recording? Its clear to me from their websites that the recording does not meet their standards for evaluation. Indicates a lack of good judgment on their part. They should have said the recording is insufficient for evaluation.

Anyone know what the papers paid them as a fee?
 
So does this effectively end their career. If they find themselves in another court case, can the attorney that is cross examining them bring up their lack in judgment in evaluating this recording? Its clear to me from their websites that the recording does not meet their standards for evaluation. Indicates a lack of good judgment on their part. They should have said the recording is insufficient for evaluation.

Anyone know what the papers paid them as a fee?

I don't think they'd go to court with their findings in the shape that they had for this case. I would hope not anyway, and if they do I would hope that cross would show that their methods were invalid for the reasons I posted above. No clue on what they were paid for their appearances, I'd be more interested to see what the software paid Mr. Owen, if I'm perfectly honest.
 
I haven't done too much digging into his background or where he got his certifications/etc, but one thing that has caught my attention is that when he did his interview he said he used "critical listening skills" and I wanted to find his definition of "critical listening skills." Here's what I found on one of his websites:

http://www.audioforensicexpert.com/audio-forensic-information/audio-authentication/


Couple things to point out:
He states numerous times that he only used his ears and listening to the calls. Steps two and three (not quoted above but can be found at the URL above) both require computer assistance.

Make your own opinions, I know what mine is.

Edit: This thread is making me feel like it's a sounding board for myself.. does no one else have any opinions on the audio experts? :\

BBM. I don't have the technical background but your research pretty much clarifies what I've thought all along, that the audio is too poor quality and there's not enough to work with to make a solid comparison. Thanks for the time you spent looking into it and sharing the info with us.
 
- if anyone is going to identify a stressed call for help... it's a family member... and T's father, when interviewed the first time (before the case got rolling and all the media came to play), said that the recording was definitively NOT T... also G's father the same day said definitively the recording was definitively G

my snip

but trayvon's mother was immediately certain it was trayvon - do you believe that the father is more likely to have it correctly than the mother?

I myself agree with your opinion re: a family member yet I think a mother would know her child in distress more likely than the father.


I know what GZ's father said, although I dont know if he could say so objectively knowing what he already knew? but what if anything did GZ's mother say? I've never heard.
 
my snip

but trayvon's mother was immediately certain it was trayvon - do you believe that the father is more likely to have it correctly than the mother?

I myself agree with your opinion re: a family member yet I think a mother would know her child in distress more likely than the father.


I know what GZ's father said, although I dont know if he could say so objectively knowing what he already knew? but what if anything did GZ's mother say? I've never heard.

It is evident that the father's first reply "no" when asked whether the voice calling for help in the audio was Trayvon's was his honest answer. When he retracted his statement after discussions with Crump, it brings doubt to his succeeding statement. The mother's statement was already biased since she already met with Crump prior to making her statement about the audio, and they all then wanted to put her son in the best possible light.
 
It is evident that the father's first reply "no" when asked whether the voice calling for help in the audio was Trayvon's was his honest answer. When he retracted his statement after discussions with Crump, it brings doubt to his succeeding statement. The mother's statement was already biased since she already met with Crump prior to making her statement about the audio, and they all then wanted to put her son in the best possible light.

2goldfish, I've gotta agree with this....

T's father was surprised with the recording and asked by LE (law enforcement) for an opinion if it was his son... he responded right away that it wasn't....

Once the TM family got their heads on straight and knew the recording questions were coming... they were much less likely to give a non biased answer. Furthermore, once they got an attorney, his job was to craft everything the family did to position for a civil award (read: defeat an SYG motion) and represent the family as an agent (trademark TM and associated terms, websites, images)...

They've done both of those things effectively so far... so I don't see how anything the TM family said after the first LE interview can hold much water...

Same goes for GZ's family... once it became evident the public outcry was going to be so great as to force the state to reverse itself, they went into protection mode.. but that being said, I haven't seen any outright suspicious or profiteering oriented....

As far as the mother knowing better, I'm certain that's a myth, it will have much more to do with how much exposure the family member had with TM, if they ever heard them under extreme stress, how perceptive they are to tonal changes and construction, etc...

Another point often brought up is the high tone of the calls for help making it impossible to be G... I dunno about the science involved, but I've heard huge 6'4" 300 lbs deep voiced guys sound like 12 yo girls when in great pain or startled etc...

So , in general.. I find it extremely hard to believe the audio evidence presented so far as being conclusive to IDing T as the voice.... in fact I find it laughable.....
 
Status
Not open for further replies.

Members online

Online statistics

Members online
168
Guests online
2,214
Total visitors
2,382

Forum statistics

Threads
589,947
Messages
17,928,048
Members
228,010
Latest member
idrainuk
Back
Top