Trayvon Martin, Vigilante Justice

[Natural Born Skeptic]

So give me one reason we should hand wave this away?

I've stated my standard several times. I want to see peer reviewed papers showing double blind tests in which experts prove they can identify voices with accuracy using these types of (substandard) recordings. You know, like the same standards we apply to psychic claims!

 

[Unabogie]

I watched the 3rd video. None of the comparisons were of a scream vs. speech.

Not a single one.

In fact, it was the same phrase repeated over and over they compared.

And the one with music had the music so low in the mix it was barely audible.

If this guy is the best expert in the field I question the entire field.

Show me where this guy can identify a voice under similar conditons as the Martin tapes. I suspect no examples will be forthcoming.

Care to answer my question? If it's proven to be Martin screaming, is that game over for Zimmerman?

 

[Alt+F4]

So give me one reason we should hand wave this away?

Here's more than one. We don't know the source of the recording this expert heard. Was it the same one released to the public on Friday or does the Orlando Sentinel have some other version/copy? Did the Orlando Sentinel pay these men for their expert opinion. We don't know.

 

[RandFan]

Are you seriously arguing that his CV proves something? Seriously?! Do you have any idea how easy it is to be a "respected expert"? I ask again, can you or anyone point us to peer reviewed double blind tests showing the accuracy of voice identification with substandard recordings such as these?

IF you can I'm entirely open to changing my opinion, but I actually have this pesky little obsession with proof.

Skepticism is a fine thing. Clearly this technology and the two separate conclusions would need to be ruled on by a judge to see if it is admissible and then challenged in court by the defense. I don't know for a fact that it demonstrates anything. It does tend to push me toward thinking it wasn't Zimmerman. We typically deal in probabilities. Knowing what I know right now I could serve on a jury and having no more information I would find Zimmerman innocent. Do I think he's guilty? After the analysis? Yeah, a little bit, but no where near the threshold of beyond a reasonable doubt.

 

[Dan O.]

How do the approaches differ Dan?

The FBI study was of experts in court cases that were using spectral analysis to compare the same spoken words between the suspect and the target. The biometric analysis used by Owen is creating a characteristic model of the vocal tract and comparing the models. The earlier process is heavily influenced by the vocal cords generating the sound. In the later case, the vocal cords are simply a noise source which is used to map the vocal cavities.

 

[crimresearch]

With the claimant. Except apparently when the claimant is Zimmerman. Then it's, "the legal concept of presumed innocence prevents us here in this critical thinking forum from critically examining his claim of self-defense".<snip>
So it's wrong to ascribe that position to you. Unlike, say, referring to all people here who aren't defending Zimmerman as woos, "witch hunters", race-baiters, and/or a fanatical "lynch mob" who are not interested in a fair trial despite being the only ones interested in any kind of trial for the first several pages of this thread.

I wasn't aware that Zimmerman had applied for the JREF $1 million.

When did he become a 'claimant'?

And when did critical thinking about a legal case include ignoring how the law actually works?

And you might want to check the definition of 'race baiting'... I don't think that term means what you think it means.

 

[Unabogie]

Here's another company that does this work. As far as I can tell, this branch of forensics is routinely used in court cases and is widely accepted.

http://www.audioforensics.com/index.htm

How about someone email them and ask these questions?

As far as I'm concerned, this sudden skepticism of this particular brand of forensics is just convenient hand waving. If the two experts had concluded the voice to be Zimmerman I have no doubt WildCat would be touting this as a major break in the case.

 

[Unabogie]

Here's more than one. We don't know the source of the recording this expert heard. Was it the same one released to the public on Friday or does the Orlando Sentinel have some other version/copy? Did the Orlando Sentinel pay these men for their expert opinion. We don't know.

Did you watch the video? His software works by dragging a wav file into one window, them other wav files into another. It takes two seconds to produce a result. I'm sure he was more than happy to demo his product and didn't require payment for his 5 minutes of work. Beyond that, why does it matter if they were paid? They could just as easily have ID'd the voice as Zimmerman. They weren't hired by Martin's family, in which case you'd have a point. They were enlisted by a news organization.

 

[TheL8Elvis]

I was answering Dancing David's question.

If you don't like the answer, please feel free to contact the Orlando Sentinel and explain to them that the forensic expert and chair emeritus for the American Board of Recorded Evidence using the exact same software that has already been used to convict murderers in court cases in the past is wrong. And don't forget the other forensic audio expert who came to the same conclusion without using the software.

I hope you have more of a reason than "because I personally don't believe it because it contradicts the opinion I've been arguing for a week in a message board thread", though.

It's damn lucky how many people in this thread are voice recognition experts technology all of a sudden. It looks to me more like people are googleing and cherry-picking, but I could be wrong.

Anyhow, about the hilited part - I need to know very little about the actual technology or ABRE to draw some conclusions.

One thing we find is an abstract regarding the prescribed methodology: http://expertpages.com/news/american_board_voice_comparison.htm
If this is accurate, the " recommended practices for the handling, preparation and analysis of recorded evidence" do not appear to have been followed in this analysis.

Second, please refer to the following link on the "Founder, CEO, and Publisher" of the ABRE, specifically the about page:
http://www.droblock.com/about.html

After you review that page, feel free to come back and post about how impressive or not impressive the ABRE sounds.

As far as the software being used for other convictions, I missed reading that. Do you have some links ?

 

[Natural Born Skeptic]

Here's another company that does this work. As far as I can tell, this branch of forensics is routinely used in court cases and is widely accepted.

http://www.audioforensics.com/index.htm

How about someone email them and ask these questions?

As far as I'm concerned, this sudden skepticism of this particular brand of forensics is just convenient hand waving. If the two experts had concluded the voice to be Zimmerman I have no doubt WildCat would be touting this as a major break in the case.

NO WE WOULDN'T. I'd have the same questions.

Some of you keep saying courts use this and it's widely accepted. So what? Courts allowed repressed memories to send people to jail. Courts on the other hand refuse to allow defense attorney's to teach people about the unreliability of eyewitness testimony and eyewitness identification even though it's unreliability has been proven rather conclusively, and we've had a bunch of people released from jail after DNA testing proved it.

So using what courts accept and don't accept as proof it's accurate is not the standard we should be looking for.

 

[Skeptic Ginger]

I've stated my standard several times. I want to see peer reviewed papers showing double blind tests in which experts prove they can identify voices with accuracy using these types of (substandard) recordings. You know, like the same standards we apply to psychic claims!

Right, because a top expert in the field who has published in peer reviewed journals can't be relied on to have tested the software he developed.

 

[Unabogie]

It's damn lucky how many people in this thread are voice recognition experts technology all of a sudden. It looks to me more like people are googleing and cherry-picking, but I could be wrong.

Anyhow, about the hilited part - I need to know very little about the actual technology or ABRE to draw some conclusions.

One thing we find is an abstract regarding the prescribed methodology: http://expertpages.com/news/american_board_voice_comparison.htm
If this is accurate, the " recommended practices for the handling, preparation and analysis of recorded evidence" do not appear to have been followed in this analysis.

Second, please refer to the following link on the "Founder, CEO, and Publisher" of the ABRE, specifically the about page: http://www.droblock.com/about.html
After you review that page, feel free to come back and post about how impressive or not impressive the ABRE sounds.

As far as the software being used for other convictions, I missed reading that. Do you have some links ?

This is the very definition of an ad hom attack. Dr. Block is suspect, therefore ABRE is a fraud, therefore Owen is a fraud, therefore audio forensics is woo, therefore we can't possibly ID the voice as not being Zimmerman.

Got it.

 

[Moss]

I'm flabbergasted people still list being in Marquis Who is Who. The only entry requirements are filling out a form and paying the fee.

 

[Skeptic Ginger]

SG has done a great job of showing how easy it is to generate "expert credentials". I'm sure the polygraph experts and chelation analysts will appreciate how easy it is to defeat her level of skepticism.

An example of Tom Owen's views on science and forensics:

Report of the meeting of the SC-03-12 Working Group on Forensic Audio of the SC-03 Subcommittee on the Preservation and Restoration of Audio Recording, held in New York, NY., US., 2009-10-11

A lengthy discussion was held regarding Daubert challenges. Chairman Owen pointed out that in the U.S. this is very important and that error rates, adequate data sets and other information must be provided for the ENF to be accepted as a scientific reliable means of authenticity. The group inquired among themselves whether it had been entered in a court of law as a means of authenticity of audio. E.B. Brixen commented that it had. The document producer C. Grigoras will be contacted regarding the admissibility law experience in Eastern Europe.

This will be discussed further at the next SC-03-12 Meeting.

New Projects

Voice Identification: issues related to Voice Identification standards were discussed at length. Owen agreed to provide a draft of the document. A discussion was held regarding the FBI inclusion of voice in their new biometric database (fingerprints, voice, eye scans, foot prints, palm prints, DNA, and other biometric features) which is scheduled to go on line shortly. The document will hopefully be on line for comment by December 2009.

Authenticity of digital audio: A paper on Digital Authenticity of audio with an error rate and methodology was discussed. We have the AES43 standard for analog authenticity that follows the Watergate Report Procedures. It would be nice to have a digital companion to the AES43 for digital files. D. Begault brought this up and it will be further discussed in Denmark [39th AES International Conference on "Audio Forensics: Practices and Challenges", Hillerod, Denmark, 17 to 19 June 2010] and future meetings.

Guy's obviously a woo pushing flake.

 

[Unabogie]

NO WE WOULDN'T. I'd have the same questions.

Some of you keep saying courts use this and it's widely accepted. So what? Courts allowed repressed memories to send people to jail. Courts on the other hand refuse to allow defense attorney's to teach people about the unreliability of eyewitness testimony and eyewitness identification even though it's unreliability has been proven rather conclusively, and we've had a bunch of people released from jail after DNS testing showed it.

So using what courts accept and don't accept as proof it's accurate is ridiculous.

Do you have any evidence at all the audio forensics is woo? Case where it's been used to wrongly ID someone? Cases where it's been proven to be snake oil? Up until now I have considered it to be more in line with fingerprinting. Sounds have markers and if you know what to look for, you can compare two sounds and calculate their similarities.

Absent any contradictory evidence, I find the dismissal of audio forensics as a field to be ridiculous.

 

[TheL8Elvis]

Ah yes, the eyewitness who gave that rather elaborate story to the press some time after the event; much like Trayvon's girlfriend.

That's dishonest. You have been following this thread long enough to know better.

He called 911 and gave the police a witness statment that night. He talked to the media the next day, and presumably told the same story.

That in no way is comparable to the girlfriend, is it ?

 

[Dan O.]

http://www.audioforensics.com/index.htm

Acoustic gunshot analysis​

There is probably enough information in the recordings to identify the location of the shooter and the locations of the phones that recorded the shot. There might be a little more information on the acoustic shadows of the shooter and victim but it would be pushing the limits.

 

[Alt+F4]

As far as the software being used for other convictions, I missed reading that. Do you have some links ?

I believe it was Mr. Owen that testified in this murder trial. I don't know if the software in question was used or some other technique. Important to know that the voice in question belonged to the murderer, who was alive.

http://www.lohud.com/article/20120211/NEWS02/302110060/Sheila-Davalloo-guilty-murder-rival-s-slaying

 

[Unabogie]

Acoustic gunshot analysis​

There is probably enough information in the recordings to identify the location of the shooter and the locations of the phones that recorded the shot. There might be a little more information on the acoustic shadows of the shooter and victim but it would be pushing the limits.

Was the underlined supposed to be a link?

 

[TheL8Elvis]

This is the very definition of an ad hom attack. Dr. Block is suspect, therefore ABRE is a fraud, therefore Owen is a fraud, therefore audio forensics is woo, therefore we can't possibly ID the voice as not being Zimmerman.

Got it.

WHAT ELSE do I base the ABRE on, but it's founder ? That's a serious question, BTW. Do you have other references to is as a professional organization ?

I point that out, because bogus professional organizations are used all the time to bolster credibility.

I'm simply being skepitcal. I'm not attacking. I'm not pretending to be an expert. I didn't call voice recognition woo. What's wrong with the approach I am taking ?