George Black Jr. of the National Transportation Safety Board gave no indication of what caused the "airframe rattling noise."
From takeoff to the end of the tape lasts less than 2 minutes, 24 seconds, Black said at a news conference.
The first portion of the flight to the Dominican Republic appeared normal, with the co-pilot at the controls. But 107 seconds after the plane had started its takeoff roll, a rattling was heard; 14 seconds later, a second rattle was audible, Black said.
Twenty-three seconds later -- after "several comments suggesting loss of control" -- the cockpit voice recording ends, he said.
The plane's second black box, the flight data recorder, was recovered Tuesday after a 24-hour hunt through a Queens neighborhood staggered by a double dose of tragedy. At least 262 people were killed when the plane crashed.
A source close to the investigation said that the NTSB was looking at whether the plane's engines failed after sucking in birds, a phenomenon that has caused severe damage to airliners in the past.
But Black said that an initial inspection of the engines found no evidence of a collision with a bird. He said a more detailed analysis still needs to be done.
REST OF ARTICLE: http://www.sfgate.com/cgi-bin/article.cgi?f=/news/archive/2001/11/13/national1704EST0773.DTL
In this day and age there is no excuse for these black box devices to just stop working before the cock pit comes apart or explodes. batterys should be located within inches of these devices, and they should conitinue to function even if the rest of the plane is in complete power failure. Also with the cost of digital memory and tiny cameras there is no reason why these black box devices should not contain video and audio from several areas of the plane, both inside and outside. These simple steps would most always ensure that whatever happened to make a plane crash would be very verifiable.