In Progress
Plain text format indicates text directly from the NTSB's CVR Transcript.
Underlined text indicates U.S.Read additions/commentary
Bold text indicates CVR sounds noted on the NTSB's CVR waveform but absent from the CVR transcript
09:14:30.40 [Sten Molin:] positive rate, gear up please.
09:14:31.50 [Ed States:] gear up.
09:14:33.10 [CAM] [sound of thump and two clicks]
09:14:38.46 [Sten Molin:] heading select.
09:14:41.87 [Ed States:] clear left.
09:14:42.55 [TWR] American five eight seven heavy, turn left.
fly the bridge climb. contact New York
departure. good morning.
09:14:48.33 [RDO-1] American five eighty seven heavy, so long.
09:14:51.38 [Ed States:] gear's up.
09:14:52.45 [Sten Molin:] check speed, level change.
09:14:54.84 [Sten Molin:] flaps up.
09:14:56.52 [Sten Molin:] climb power.
09:14:57.14 [CAM] [sound of click]
09:15:00.02 [RDO-1] ah New York, American five eighty seven
heavy, thirteen hundred feet, we're
climbing to five thousand.
09:15:04.67 [DEP] American five eight seven heavy, New York
departure. radar contact. climb
maintain one three thousand.
09:15:10.18 [RDO-1] one three thousand feet, American five eighty seven heavy.
09:15:14.64 [Sten Molin:] one three I see, slats retract.
09:15:16.46 [Ed States:] slats.
09:15:17.21 [CAM] [sound of several clicks]
09:15:28.47 [Ed States:] clean machine.
continue hereYou may have noticed as a passenger, when the slats and flaps on an aircraft are deployed, there is more noise coming from the wing area. Now that the slats are retracted, the cockpit noise reduces a bit as evidenced by the
09:15:29.34 [Sten Molin:] [sound similar to yawn] thank you.
09:15:35. continue here in re wake before? movements. include jpg from NTSB fdr plots and perhaps some shots from anim vid
09:15:36.43 [DEP] American five eighty seven heavy, turn left, proceed direct WAVEY.
09:15:37.80 [CAM:] [sound of brief squeak and a rattle] (corrected time as per NTSB's waveform)
09:15:41.50 [RDO-1] uh, we'll turn direct WAVEY, American five eighty seven heavy. (corrected time as per NTSB's waveform)
09:15:44.2 [an object appears on the Newark radar data, 182 feet off Flight 587's right side; altitude of possible object or debris unknown] (U.S.Read addition from FAA Newark Radar Data) INSERT GRAPHIC HERE
09:15:44.80 [Sten Molin:] left turn direct WAVEY... (corrected time as per NTSB's waveform)
09:15:45.10 [Ed States:] little wake turbulence, huh? (corrected time as per NTSB's waveform)
09:15:46.25 [Sten Molin:] ...yeah. (corrected time as per NTSB's waveform)
09:15:47.31 [Sten Molin:] [sound similar to five sets of stabilizer trim switch clicks]
09:15:48.00 [an object appears on the Newark radar data, 1300 feet off Flight 587's left side; altitude of possible object or debris unknown] (U.S.Read addition from FAA Newark Radar Data) NTSB did not release this data; U.S.Read had to secure this from the FAA via a FOIA request. INSERT GRAPHIC HERE
09:15:48.85 [Sten Molin:] two fifty thank you. (corrected time as per NTSB's waveform)
09:15:49.80 [slight increase in cockpit noise; introduction of noise in the 30 - 40 Hz range which had been silent; increase in noise energy in the 60 - 100 Hz range. Aircraft not in contact with any potential wake vortices at this point and there are no control inputs being made by the Pilot.] (U.S.Read addition from NTSB's Sound Spectrum Low Pass Filter Study) INSERT GRAPHIC HERE
09:15:51.20 [0.10 vertical g "unload" from .94g to .84g, within the last second; the DFDR does not indicate a lateral acceleration of 0.1g as the NTSB continues to maintain -- it's just not on the FDR.] (U.S.Read addition from DFDR data)
09:15:51.20 [FIRST OFFICER STEN MOLIN BEGINS CONTROL INPUTS turning the control wheel to the right a split second before before his first rudder input] (U.S.Read addition interpolated from DFDR data and shown in NTSB animation video) INSERT GRAPHIC HERE
09:15:51.40 ["Try Escape" spoken by Sten Molin on Departure Control FAA frequency but not found on CVR] (U.S.Read addition from FAA tapes) INSERT GRAPHICS HERE
09:15:51.40 ["Try .." spoken by Sten Molin on Local Control FAA frequency; this is a portion of the same transmission on the other FAA frequency. Commentary on these transmissions provided later on in this timeline ... at 9:16:01.90] (U.S.Read addition from FAA tapes).
09:15:51.50 [control wheel is at 64 degrees right having moved there from 6 degrees left within the last 0.30 seconds] (U.S.Read addition from DFDR data)
09:15:51.50 [first of five rudder pedal movements begins] (U.S.Read addition from DFDR data)
09:15:51.60 [sharp, sudden increase in cockpit noise; noises introduced in 15 - 30Hz range which had been silent; noise stays at this higher level for the next 3.9 seconds before increasing again] (U.S.Read addition from NTSB's Sound Spectrum Low Pass Filter Study) INSERT GRAPHIC HERE
09:15:51.90 [Rudder pedal position is at 8.8 degrees right having moved there from neutral within the last 0.4 seconds; the first of five rudder movements] (U.S.Read addition from DFDR data)
09:15:52.10 [Filtered Rudder position is 5.5 degrees right] (U.S.Read addition from DFDR data)
09:15:52.15 [CAM] [sound of a thump] (corrected time as per NTSB's waveform)
09:15:52.65 [CAM] [sound of click] (corrected time as per NTSB's waveform)
09:15:53.00 [CAM] [sound of two thumps] (corrected time as per NTSB's waveform)
09:15:54.25 [though the aircraft is level again with a zero degree sideslip at this point the noise energy levels in the cockpit remain at the higher levels reached 2.65 seconds earlier] (U.S.Read addition from NTSB animation video and Sound Spectrum Low Pass Filter Study INSERT GRAPHIC HERE
09:15:54.41 [possible "event" noted next to Flight 587 on the "tollbooth video"; NTSB analyzed a 2nd or 3rd generation copy VHS tape; they do not have the original tape nor have they ever analyzed the original. The NTSB also did not form a study group with the parties to the investigation. The parties were never invited to analyze the tape or discuss ways to better analyze it. The NTSB told the parties what was there and what wasn't there, 8 months after the crash, and the discussion was over. This is a bit unusual given the formation of investigative workgroups for other areas of the investigation. The FBI maintains custody of the original tape(s) and countless efforts by U.S.Read to locate the department or Agents at the FBI who maintain custody have proven fruitless. U.S.Read does have a pending FOIA request to make a "digital uncompressed" copy that will be examined by a renowned expert. This video forensics specialist who teaches video forensics to the U.S. Government told U.S.Read that the original video must be analyzed to determine if the events noted on the tape are digital effects due to the "compression" loss and/or quality loss from multiple VHS copies. This specialist stated these "events" could be events associated with the aircraft and that any statements made by the NTSB about what is and isn't on the tape can be immediately discredited since they worked from a copy] (U.S.Read has obtained an MPEG compressed CD version of the tollbooth video from the NTSB). The NTSB makes this CD available to the Public via their Public Inquiries Branch.
Witnesses John and Jackie Power here and Irwin (#238)
09:15:54.50 [Sten Molin:] max power. [spoken in strained voice] (corrected time as per NTSB's waveform)
09:15:55.25 [Ed States:] you all right? (corrected time as per NTSB's waveform). U.S.Read commentary on this comment from Captain States available here.
09:15:55.50 [sharp, sudden rise in cockpit noise; noises introduced in 1 - 15Hz range which had been silent; sharp rise in noise energy in the 50 - 100Hz range. The increase in noise energy cannot be attributed to the crew going to max power as they will not start that maneuver for another 3.9 seconds (at 9:15:59.4)] (U.S.Read addition from NTSB's Sound Spectrum Low Pass Filter Study) INSERT GRAPHIC HERE
09:15:55.54 [possible event noted next to Flight 587 on the "tollbooth video"]
09:15:55.70 [Sten Molin:] yeah fine. (corrected time as per NTSB's waveform)
09:15:56.30 [sharp increase in cockpit noise energy levels in the 30 - 60 Hz range] (U.S.Read addition based on NTSB's Sound Spectrum Low Pass Filter Study) INSERT GRAPHIC HERE
09:15:56.60 [Ed States:] hang onto it. hang onto it. (corrected time as per NTSB's waveform)
09:15:56.90 [CAM] [sound of snap] (corrected time as per NTSB's waveform)
09:15:56.90 DFDR records 4th rudder pedal movement (to the left) which exceeds the boundaries of the limiter by at least 30% (Limiter was applying a max. 10 degree deflection yet rudder pedal deflected to at least 13 degrees). This is concurrent with the sound of a "snap" in the cockpit INSERT GRAPHIC HERE FROM THE WAVEFORM AND MAYBE THE LOW PASS FILTER. Did the additional pedal travel beyond the limiter also lead the rudder itself beyond the limiter? The NTSB's "inverse filter" of the DFDR Rudder position indicates the rudder panel may have exceeded the limiter by at least 20% -- yet the NTSB was willing to surrender this point to Airbus and agree with them that the rudder panel itself did not exceed the limiter.
09:15:57.75 [Sten Molin:] let's go for power please. (corrected time as per NTSB's waveform)
09:15:57.75 [CAM] [sound of thump] (on NTSB's waveform but not CVR transcript)
09:15:58.00 [CAM] [sound of loud thump] (corrected time as per NTSB's waveform)
09:15:58.35 [CAM] [sound of thump] (on NTSB's waveform but not CVR transcript)
09:15:58.50 [NTSB states this is the moment of tail separation; however, the DFDR data does not support this as there is one more reliable rudder position reading after this point and the "Loud Bang" (which the NTSB has hinted is the sound of tail separation but denies in their CVR report) does not occur for another 0.30 seconds] (U.S.Read addition from DFDR data and CVR waveform) INSERT GRAPHIC HERE)
09:15:58.60 [DFDR records the rudder position as a reliable reading of 6.33 degrees right; this is the last reliable rudder position reading] (U.S.Read addition from DFDR data)
09:15:58.80 [CAM] [sound of loud bang] (corrected time as per NTSB's waveform). (This is the loudest cockpit noise during the pilot's control movements; the source of this "bang" has not been identified by the NTSB. INSERT GRAPHIC HERE)
09:15:58.95 [Signal interruption on Sten Molin's CVR channel for 0.2 secs] (on NTSB's waveform but not CVR transcript)
09:15:58.95 ["Noise" on Departure Control FAA frequency also for 0.2 seconds; this "noise" is coincident with the Loud Bang noted on the CVR] (U.S.Read addition from FAA tapes and CVR waveform)
09:15:58.95 [DFDR registers a mike keying for less than 1 second coincident with the "noise" transmission] (U.S.Read addition from FAA tapes and CVR waveform)
09:15:58:99 [White Plains Radar records a secondary (normal) radar hit for Flight 587. NTSB/Airbus theory is that 587's tail departed 0.49 secs earlier. Yet, the White Plains, NY Radar center, the first radar to scan 587's vicinity after the NTSB's assumed tail separation point, apparently shows no debris in the vicinity of 587. The NTSB has not released radar data that might support their theory since they did not publish "primary" radar data for the White Plains radar prior to 09:16:03.59. They have not fulfilled U.S.Read's 6 month old FOIA request for any such data, and the FAA says they don't know how to get the White Plains data]. (U.S.Read additions based on radar data).
09:15:59.10 [First unreliable rudder reading (position at zero degrees); the rudder readings remain mostly at zero for the remainder of the DFDR data (2 more seconds)] (U.S.Read addition from DFDR).
09:15:59.40 [Thrust levers being moved most probably in effort to go to max power] (U.S.Read addition from DFDR data)
09:15:59.80 [Horizontal Stabilizer position also jumps to zero; this is the first position reading for the horizontal stabilizer after the Loud Bang] (U.S.Read addition from DFDR data)
09:15:59.80 [At the same time, Spoilers 2, 3, and 5 show a FAULT in the DFDR. this is the first position reading for these spoilers after the Loud Bang] (U.S.Read addition from DFDR data)
09:16:00.00 [Sten Molin:] [sound similar to human grunt]
09:16:00.40 [Thrust lever position, and subsequent engine readings, confirm that crew has gone to max power] (U.S.Read addition from DFDR data)
09:16:00.40 [CAM] [roaring noise starts and increases in amplitude] (corrected time as per NTSB's waveform)
09:16:00.80 [Spoilers 2, 3, and 5 show a FAULT in the DFDR again.] (U.S.Read addition from DFDR data)
09:16:00.80 [Horizontal Stabilizer at a basically neutral position of -0.09] (U.S.Read addition from DFDR data)
09:16:01.20 [CAM] [sound similar to single ECAM chime] (corrected time as per NTSB's waveform). POSSIBLE LEVEL 2 ALERT (NOT LEVEL 3). PERHAPS ENHANCE THIS AREA ALONG WITH OTHER ALERT SOUNDS
09:16:01.20 [All DFDR data ends yet the DFDR is powered for several seconds beyond this point; both engines were attached and running at end of all DFDR data] (U.S.Read addition from DFDR data)
09:16:01.21 [Sten Molin:] holy # (corrected time as per NTSB's waveform)
09:16:01.36 [definite "event" noted near Flight 587 by video forensics specialist on the "tollbooth video"; not a digital effect] Further analysis requires original tape or a digital uncompressed copy.
09:16:01.82 [Only one piece of debris believed to be from Flight 587 appears on the JFK Radar. But there is only one piece of debris. U.S.Read Commentary - The NTSB/Airbus theory is that the rudder shredded into a dozen or more pieces after the tail separated from the airplane. The radar data should show more than one piece of debris at this point (since the NTSB maintains the tail separated 3.5 seconds earlier). This lack of debris on radar raises the possibility the tail may still be attached, as other circumstantial evidence will show, but that something else had departed the plane. Also, the NTSB has hinted that the DFDR data ended because the engines departed at 09:16:01.20. If that also is the case, there could be at least 16 pieces of debris in the air at this time. Yet, only one piece shows up on radar. (U.S.Read additions based on FAA Radar Data)
09:16:01.90 [Stressed, garbled voice of Sten Molin saying something akin to "Losing Control" is picked up on the Departure Control FAA frequency; this transmission is 1.65 seconds in length and has at least 3 separate "keying" moments. The CVR waveform shows the exact same waveform pattern as the FAA transmission--the recording occurs at the same time as the FAA transmission, is exactly as long, and has the same number of keying moments (apparent clicks of the yoke mike button). The CVR transcript does not acknowledge what the CVR waveform clearly shows and the NTSB has never acknowledged that this transmission was indeed transmitted from Flight 587. U.S.Read first identified this transmission as such on February 25th 2002.] (U.S.Read addition from FAA tapes and CVR waveform) INSERT GRAPHICS HERE
09:16:01.90 [the "losing control" transmission on the Departure Control frequency bleeds over to the Local Control FAA frequency although at a poorer quality than the Departure Control transmission. The "try escape" transmission from 09:15:51.63 also exhibited this same bleed over behavior to the Local Control frequency. This cannot be a coincidence. This "losing control" transmission is without a doubt ... from Flight 587 given the CVR waveform signature. And it proves that "try escape" came from Flight 587 because the two transmissions behaved similarly. This is indicative of something seriously wrong with Flight 587 no later than 09:15:51.63, long before the tail departed. Why wasn't "try escape" on the CVR waveform when Sten's voice was so clearly picked up on the FAA tape? The most likely possibility is that there was a temporary malfunction of the CVR due to some other problem with the aircraft that affected the electrical system(s).] (U.S.Read addition from FAA tapes and CVR waveform) INSERT GRAPHICS HERE
09:16:02.20 [CAM] [sound similar to single ECAM chime] (corrected time as per NTSB's waveform)
09:16:03.59 [587's transponder sends it's last response to interrogating radars--this last response goes to the White Plains radar)] (U.S.Read addition based on FAA Radar data).
09:16:04.50 [CAM] [sound similar to stall warning repetitive chime for 1.9 seconds] (corrected time as per NTSB's waveform)
09:16:06.20 "FLASH" ON TOLL BOOTH VIDEO
09:16:06.20 [Flash on tollbooth video followed by a long large streak of smoke and or fire. Also noteworthy, Flight 587 has lost no altitude yet ... 7 to 8 seconds after assumed tail separation; only now does 587 begin to lose altitude. (U.S.Read addition based on tollbooth video and NTSB's Video Study)
Jim Conrad and Tom Lynch here and John Power and Jo-Ann Catansese and Fire Chief Hayden and Patrick Twohig and Deligiannis
09:16:06.35 [Multiple pieces of debris from 587 do not begin showing up on radar until AFTER this moment, according to all four radar facilities. This fact comments very powerfully on when 587 started to shed most of it's parts including the tail and 13 or more pieces of rudder, engine debris, etc.; this fits with the tollbooth video, the eyewitnesses, the CVR noises, and the debris pattern in Jamaica Bay according to the men who retrieved the parts as they fell into the Bay] (U.S.Read additions based on radar data and eyewitness and recovery workers statements).
09:16:06.40 [CAM] [roaring noise decreases and ends] (corrected time as per NTSB's waveform)
09:16:06.40 [U.S.Read Commentary - As the roaring noise first began as the crew had just gone to max power 6 seconds earlier .. it is possible that a contributor to the roaring noise was indeed the engines at max power. If so, then the decrease and cessation of the roaring noise could be a good indicator of when the engines departed (or seized .. before departing)].
09:16:06.40 [CAM] [stall warning stops] (U.S.Read addition)
09:16:06.40 [U.S.Read Commentary - both engines depart anytime between now and 09:16:12.84 when Captain States says, "get out of it, get out of it". Engine departures (or seizures) now seems likely as the cockpit noises have decreased dramatically at this moment based on the CVR waveform and Low Pass filter; the location of the engines on land--very close proximity to each other and the main crash site, and all the eyewitness statements, indicate the engines departed when the plane was lower in the sky than usual, and very close to the Rockaway Peninsula] INSERT GRAPHICS HERE
09:16:07.55 [CAM] [sound similar to continuous repetitive chimes for one second] (corrected time as per NTSB's waveform)
09:16:08.00 [Sten Molin:] what the hell are we into *. we're stuck in it. (corrected time as per NTSB's waveform)
09:16:08.55 [CAM] [continuous repetitive chime ends] (U.S.Read addition)
09:16:09.60 [CAM] [sound similar to continuous repetitive chimes for three seconds]
09:16:10.30 [CVR Waveform and FAA Departure Control frequency confirm a 0.2 second transmission from Sten Molin's mike which records one of the continuous repetitive chimes in the cockpit] (U.S.Read addition from FAA Tapes and CVR Waveform) INSERT GRAPHICS HERE
09:16:12:60 [continuous repetitive chimes end] (U.S.Read addition from CVR waveform)
09:16:12.84 [Ed States:] get out of it, get out of it.
09:16:14.78 [END of RECORDING]
notes to self -
add more RADAR - interpolated position & speed of 587 after
xponder dies.; use my new spreadsheets
note time xponder s/h replied to radar but did not - and why
not given they were on xponder 1; use alaska air as proof that
out of control does not translate into no xponder function
4th rudder movement beyond limiter - w/snap -
5th rudder movement beyond limiter
level 2 and 3 warnings - Steve comment re the cockpit being full
of them - comment from Airbus source "both RED visual for
all of mentioned will be accompanied with aural but with all that
was happening during this envelope ECAM and aural warnings would
have filled the cockpit."