We are less than a day away from the start of the George Zimmerman trial.

A key hearing over evidence in the case ended after a witness was unavailable to testify Saturday.

However, two audio experts, Dr. Peter French and Dr. George Doddington, did testify for the defense.

Court is scheduled to continue 9 a.m. Monday and can be seen live on News 13 and Bay News 9.

Zimmerman faces second-degree murder charges for the shooting death of teen Trayvon Martin.

———————————————

Latest updates from Saturday's hearing

  • 2:51 p.m. - The next witness is aboard a plane and cannot testify. The hearing is continued. Court is in recess until Monday at 9 a.m.
  • 2:29 p.m. - While waiting for the witness to get to a place to video conference, Don West asks the judge regarding George Zimmerman’s requirement to check in while he is in court. He is required to phone in during certain times of the day. Court is in recess until the witness is ready.
  • 2:26 p.m. - O'Mara finishes his questioning of Dr. George Doddington. The defense will call their next witness.
  • 2:22 p.m. - O’Mara goes back to Mantei’s question about Dr. Doddington’s initial decision to not listen to the sample. “There’s no value in trying to make and identification because it’s not possible,” Dr. Doddington says.
  • 2:20 p.m. – “Forensic speech community. They don’t participate in these kinds of evaluations,” Dr. Doddington says. O’Mara asks about the forensic science community being a subset of the field and if they have excised themselves from the general community. “How do you share knowledge without doing evaluations? That’s the heart of it,” Dr. Doddington says as part of his answer.
  • 2:10 p.m. - Mantei asks Dr. Doddington about NIST and its involvement in setting standards. The state finishes. O’Mara asks questions to clarify NIST’s role and purpose.
  • 2:00 p.m. – Mantei questions Dr. Doddington about the “forensic sample.” Dr. Doddington calls Dr. Reich’s analysis “absurd.”
  • 1:50 p.m. - The hearing has resumed with State attorney Richard Mantei questiong defense expert Dr. George Doddington.
  • 1:12 p.m. - Court is in recess until 1:45 p.m. for lunch.
  • 1:10 p.m. - O’Mara finishes his questioning asking about Doddington’s work with the NSA and his security clearance.
  • 1:00 p.m. - O’Mara asks about Listener Bias concept, the ability of a listen hearing what he’s preconditioned to hear. Dr. Doddington says it is apparently very effective, according to Mr. Reich.
  • 12:50 p.m. - Dr. Doddington is asked about Mr. Owens’ testimony about the length of the sample. He says looping the data leads to “statistical dependence.”
  • 12:32 p.m. - “I have for a long time tried to induce the forensic science community to get calibrated,” Dr. Doddington said. He goes on to describe something called Human Aided Speaker Recognition, HASR. “In 2010 NIST evaluation, we crafted a special Human Aided Speaker Recognition part that involved 15 trials that tried to induce the forensic community to participate.”
  • 12:25 p.m. – “It’s all ridiculous,” Dr. Doddington says as O’Mara describes audio with screaming or shouting. The idea that you could, well, I’ll differ to Dr. French on this. He said it quite eloquently,” he says about the audio.
  • 12:00 p.m. - In reference to a question about modalities, Dr. Doddington says he’s interested in evaluation. “People will do what they will,” he says.  He says lay human listeners had the best performance in a 1985 analysis.  "Human listerners are pretty much the same now as they were then," he says. "What has changed is in the computer area. Computers are thousands fo time more powerful."
  • 11:55 a.m. – Dr. Doddington talks about speaker recognition and how the data is analyzed.
  • 11:45 a.m. - Dr. Doddington describes projects he has worked on and discusses the Gaussian scale, described by O’Mara as an algorithm.
  • 11:20 a.m. - Defense witness George Doddington takes the stand. He says he is an electrical engineer and describes his work and education history.
  • 11:18 a.m. - Hearing has resumed.
  • 10:56 a.m. - Questioning finishes with Dr. French. The court takes a brief recess.
  • 10:55 a.m. - Defense attorney West asks Dr. French to explain when a voice changes for men. Dr. French says answering about physical changes might be beyond his expertise.
  • 10:50 a.m. – Mantei goes back to the topic of vocal track measurement. In general terms, would you say younger persons will have a high pitched voice, he asks. Dr. French answers yes.
  • 10:40 a.m. – West finishes questioning. State attorney Richard Mantei begins questioning.
  • 10:30 a.m. - Dr. French describes some things he finds “disturbing;” the resonance with the sound of someone saying “stop.” “I can’t see any basis to that claim,” Dr. French says about the claim that that was Martin. The sound is not age related at all in his opinion, Dr. French answers.
  • 10:25 a.m. - West asks Dr. French about aural spectrograph approach. Dr. French compares it to voice printer approach, and says he wouldn’t condone it. Dr. French says state expert Dr. Alan Reich’s approach of voice print analysis has never been used in the U.K. and says there is no such thing as a voice print.
  • 10:20 a.m. – Dr. French says biometrics, the method used by state’s expert, Mr. Owens, is not a good method for testing the 911 call. He mentions reduplication and adjusting the pitch, as well.
  • 10:15 a.m. –“Are you saying there’ s no reliable research and data to allow a reliable voice comparison based on screams,” West asked. Dr. French answers yes. “My view in this case is that recording isn’t even remotely suitable for speaker comparison,” Dr. French says, adding he would’ve refused to analyze it.
  • 10:10 a.m. - West asks about ways to analyze voice characteristics. Dr. French adds they look for individual abnormalities that set the voice apart from the norm. West asks about speech length for finding those identifiers. Dr. French says it’s critical that it is speech, not scream, but says the screaming in this case – several minutes – could not find a result. “It wouldn’t be the least helpful…no matter how much it was,” Dr. French says.
  • 10:00 a.m. - West asks if it is generally accepted that voice samples are not acceptable for voice comparison. Dr. French says yes. In further questioning, Dr. French describes research that analyzes properties of recording of stressed cries and normal voice samples. The results state you can’t compare them for voice identification, Dr. French explained because pitch tends to go up and that can’t be predicted, along with resonance. The timber, or the voice quality, can’t be analyzed from screaming, Dr. French said.
  • 9:55 a.m. - West asks about voice comparison procedures for Dr. French’s lab. While describing evaluation process, Dr. French sums up saying examination is complex. West asks about analyzing shouting. Dr. French says 10-15 percent of cases submitted to his lab are unsyncable for analysis.
  • 9:45 a.m. - West asks Dr. French about his experience in scholarly and forensic work.
  • 9:35 a.m. - The hearing gets underway with the swearing in of a witness, Dr. Peter French, via video conference and questioning by Don West. Audio briefly not available.
  • 9:26 a.m. - George Zimmerman's attorneys arrive for the hearing at the Seminole County Courthouse scheduled for 9:30 a.m.

———————————————

Recap of Friday's hearing

The hearing is centered on the 911 call where you can overhear screams for help just before Zimmerman fatally shot Trayvon Martin.

Zimmerman’s defense attorneys are trying to convince the judge not to allow the jury to listen to state audio expert analysis of that call.

Two state experts testified Friday, with one saying it was not Zimmerman yelling.

The other testified that it was the 17-year-old’s voice that could be screaming for help.

The 911 call was made by a woman who lives near where the shooting happened.

“My testimony is that because of the limitations of the number of words, I can reach a conclusion, I just can’t say positive yes or no,” said Tom Owen, a state audio expert.