Newsgroups: *rec.music.classical.guitar* From: *William D Clinger You obviously are loosing control here Clinger. The industry standard > for the manufacture of the FDR must meet is 0.05 seconds. As you and > the other loonies, are suggesting that there is absolutely no way to > know the trend of a 100 ton object traveling 535 mph one second later, > half a second is an eternity. While Rob acknowledges this 0.05 lag, > for the benefit of the doubt, he adds another 0.05 into his > calculations, and rounds it off at 1.00 seconds quite generous of > him! With one sample per second, as in the F-2100 [1], the worst-case recording latency cannot be less than one second. That is just a mathematical fact. I was aware that Rob Balsamo has been touting a latency of 0.5 (not 0.05) seconds. That is just one of many clues to his expertise. To the one-second minimum latency implied by the sampling rate, we must add the latency due to buffering, which I explained in a previous post. Page 3 of the NTSB report does not explain whether the F-2100 buffers at the subframe level or at the frame level, but frame buffering is the most likely reason for defining frames at all [1]. If the F-2100 buffers at the subframe level, then the worst-case latency must be at least 2 seconds: one second (in the worst case) to wait for the next sample to begin, one full second to collect that sample's data in the subframe data, and then some additional time (which might be Balsamo's 0.5 seconds) to write the subframe buffer into flash memory. If the F-2100 buffers at the frame level, then the total worst-case recording latency must be at least 5 seconds: up to 1 second to wait for the next sample, up to 4 seconds to fill the frame buffer, and then some additional time to write the frame buffer into flash memory. Completely independently of the above, John Farmer concluded that "The FDR file positional data ends 6+-2 seconds prior to the reported impact location." Other investigators have reached similar conclusions. You do the math. Will [1] NTSB. Specialist's Factual Report of Investigation Digital Flight Data Recorder. NTSB Number CDA01MA064, 31 January 2002. Online at http://flight77.info/docs/AAL77_fdr.pdf [2] John Farmer. Final Analysis of American Airlines Flight 77 Flight Data Recorder NTSB FOAI Files. Online at http://aal77.com/ntsb/Final%20Analysis%20of%20NTSB%20Fight%20Data%20Recorder%20Freedom%20of%20Inform.pdf