No decision was made in the hearing to determine if the state's voice experts will testify at George Zimmerman's second-degree murder trial because of a defense witness being unavailable to testify at Saturday's hearing.
Judge Debra Nelson said she will discuss the matter sometime next week and jury selection will go on as planned on Monday. Court is in recess until jury selection at 9 a.m. on Monday, Nelson said.
The third witness was slated to testify via video conference but was stuck inside a delayed plane.
Zimmerman was in court, which began promptly at 9:30 a.m. Saturday, wearing a dress shirt, gray sport coat and tie. The hearing was focused on the audio experts testifying about the screams heard in the background of neighbors' 911 calls.
The defense first called forensic speech/audio expert Dr. John Peter French from the United Kingdom, via video conference. When questioned about his experience, French said he's handled 5,000 cases and appeared in court 200 times, mostly for prosecution.
Attorney Don West then questioned French on how he analyzes a recording audio sample. As far as analyzing screaming, the features French said he looks for to identify the voice are accent, rate of speech, vowel sounds, and more.
French said it is difficult to do a voice comparison when someone is shouting and that you can't determine from someone's normal voice what they sound like when under severe attack.
"I've never come across a case in the 30 years of my career where anybody has attempted to compare screaming with normal voices," French said, adding that voice comparison with screams is "uncharted territory."
French said the 911 call in question is not suitable for comparison, saying it "wouldn't have gotten to first base."
The state has started to question French, who said he only needs a new syllables to exclude a voice source, depending on quality.
The state also asked French on how puberty could affect the voice quality and pitch.
"People who haven't undergone the voice change, generally speaking, they'll have a higher pitch voice," French said.
The second witness the defense has called on Saturday is George Doddington, voice recognition exert and electrical engineer who is being questioned by defense attorney Mark O'Mara. Doddington worked for Texas Instruments on voice recognition for tech-like door entry.
Doddington said speech is not like fingerprints, iris scans or DNA--it is an energy signal or "performance."
He also said that recognition systems are error-prone.
O'Mara also focused on having Doddington discuss the difference between speech recognition technology-- meaning what's said-- and speaker recognition-- meaning who says it.
Doddington said doing speaker recognition on one second of good speech is "absurd" and that speech over the phone degrades the performance, implying the short, isolated screams on the 911 call are too brief and degraded quality for proper evaluation.
The technique of looping a one second scream for 10 seconds, Doddington says, is a "violation of common sense." In addition, he called trying to identify the screams as a "fool's mission," saying in order to test screams, you would have to subject people to "unethical amounts of stress."
The state quickly questioned Doddington, who says he likes disagreements in his field and calls them "the sign of progress."
When asked why Doddington listened to the tapes when he doesn't evaluate forensic samples, he replied, "what got me to listen to the tapes is Dr. Reich's findings. It sounded absurd and it was."
Doddington then told O'Mara that it was "not possible" to identify the audio.
The experts the defense is questioning are Tom Owen, the expert who excluded Zimmerman as the source of the screams in the 911 calls, and Dr. Alan Reich, the expert who said he hears Trayvon Martin saying "I'm begging you" in the background of the 911 calls.
"The words at the screaming level were almost entirely those of Trayvon Martin," Reich said on Friday when asked about his analysis of the 911 call from a witness the night of the shooting.
French said that he was "disturbed" by Reich's methodologies he testified to.
Owen said on Friday he used Zimmerman's voice from his reenactment video to try to match the screams on the 911 call, raising Zimmerman's voice in the reenactment video and looping the 911 call clip with the screams in order to make a long enough sample to analyze.
French opposed to Owen's methodology, saying it was," not acceptable in the wider scientific community."
Both Owen and Reich testified on Friday at the Frye hearing, as West challenged their methodologies, findings and qualifications as audio experts.
The screams are crucial pieces of evidence since they could determine who the aggressor was in the confrontation.
The defense could also ask Nelson on Saturday to bar prosecutors from using certain words during the trial, believing words such as "profiling," or "vigilante," or even "wannabe cop," to describe Zimmerman would create an emotional response in the jury.
Nelson did rule on Friday that Zimmerman's 6 p.m. curfew could be extended until 10 p.m. at the end of Friday's hearing, but said he is still not allowed to travel out of Seminole County except to his attorney's office. The state had objected to the motion, given prior issues with bond.
Zimmerman is charged with second-degree murder for fatally shooting 17-year-old Trayvon Martin during a struggle in a gated community where he lived. He is pleading not guilty, claiming self-defense.
Martin's family claim the cries came from the teen while Zimmerman's father has testified they were those of his son.
Stay with ClickOrlando.com and watch Local 6 News for a full report.