At 04:18 PM 7/2/00 , you wrote: >All, > >I made a few changes to the format (see >http://www.seismicnet.com/psn4tbl.html) based on the comments from Edward. >All floats have been changed to doubles and I changed some bytes and shorts >to longs. The variable length section can now have data fields larger then >255 bytes. > > Also, it would be useful to include a clock > >correction in seconds as a double (type/length) that could be added to the > >nominal time that the samples were originally time-stamped with when >recorded. > >Particularly in the case of "unlocked" data, that estimate of true time >might be > >fuzzy and controversial. > >It seems to me that the start time should already include this information. >Basically the start time of the first sample is based on the best >information the data logger has at the time it saves the data to disk. Hi Larry, I think the problem is that if the time originally recorded for the first sample is not correct, then it's better to use a correction field than to simply change the number that was originally stored with the file. In that way, one always knows what the original time stamp was and the correction tells how far the clock was off. If you later figure out a better clock correction, then a new correction can be used and you never loose track of what you're doing, or have done. John John C. Lahr 1925 Foothills Road Golden, CO 80402 (303) 215-9913 john@........ http://lahr.org/john-jan/science.html __________________________________________________________ Public Seismic Network Mailing List (PSN-L)
Larry Cochrane <cochrane@..............>