- 25 -
1.14.2.3
1.14.2.3.1.
Inspection of the CVR tape
The full length of the tape was examined visually. Only one tape splice was found for which both ends of the tape had been cut precisely at a 45° angle. This was the original splice necessary to create the tape loop. The length of the tape was not measured though the duration of the recording was determined from the playback and indicated a length of approximately 224 ft as compared with 215 ft specified in the manual. Apart from increasing the duration of the recording, this additional length would not have affected the functioning of the CVR.
1.14.2.3.2
There were traces of mineral deposits, particularly along the edges of the tape, that were consistent with salt water immersion. Significant physical damage at the beginning of the tape matched that of the inside diameter of the CVR tape stack. The tape damage was consistent with that on the tape transport and its armoured container. A point mark and a line across the tape, 2.92 inches and 1.77 inches from the end of the tape, respectively, corresponded to the distance between the erase and record heads. The distance of these marks from the end of the tape and playback of the audio information, indicated that the tape loop was cut before the erase head when it was being removed from the container in the USSR in 1983. This was the normal procedure to ensure that the cut was made a short distance into the oldest information.
1.14.2.4
CVR tape playback
1.14.2.4.1 The CVR tape was played back at the BEA cockpit voice recorder laboratory. The frequencies of the power supply interference on one of the channels were monitored by spectral analysis and the tape speed synchronized before the copy tapes were started in the record mode. A working copy was used to make the written transcription of communications and initial noise analyses.
1.14.2.4.2
The task of producing a transcript of the CVR involved translation from Korean to English. In addition to Korean Air personnel, the assistance of a Korean translator was arranged by the French authorities. The elapsed time of the copy tape to the nearest second was displayed and noted for
each audio record.
1.14.2.4.3 For the first nine minutes, voice recordings were confined to casual conversation on the flight deck and public address announcements. Apart from the voice recordings, a number of background noises were examined. Intermittently through the first seven minutes forty-five seconds of recordings some unusual sounds occurred which were identified as a keyed continuous wave semi-automatic Morse code tone. Due to the poor quality of the recorded signal and the limited activity no read-out or user identification was possible.
1.14.2.4.4 From 18:15:42 to 18:20:10 hours radio telegraphy signals of varying amplitudes with a carrier frequency of 500 Hz were evident. Analysis confirmed that the signal was a keyed continuous wave semi-automatic Morse code sequence of numbers transmitted at forty words per minute. A transcript was made but no user identification was possible. The reception of such signals on HF by aircraft was a common occurrence.
1.14.2.4.5 At 18:22:56 hours KE 007 reported reaching FL 350. Just over three minutes later, at 18:26:02 hours, a rapid series of loud noises was heard on the cockpit area microphone (CAM) track. This was the moment of missile detonation.