Geoff,

This has been discussed before. WinQuake skips the first few samples in the event 
file so that the display starts at the top of the first second. This is only done if 
there is a fraction of a second in the nanosecond start time field.

 > A. Put the fraction 0.151 into the nanoseconds variable times one billion

This is the correct way of doing this. Since the nanosecond field is a 32 bit long 
you will need to multiply .151 times one billion and place the integer result into 
the nanosecond field.

Regards,
Larry Cochrane
Redwood City, PSN

Geoffrey wrote:
> Hello Folks;
> 
> Please bear with me here I know you do not like to hear this
> but it will help me decide how to finish my program.
> 
> I am rewriting my recording and display programs
> to use 12 bits instead of 8 to avoid saturation problems.
> In my display program I create a PSN4 file to be able to
> read times in winquake and share with the rest of you.
> I am having problems getting WINQUAKE to properly
> show the time of the very first sample since
> MY first sample can be something OTHER than
> an exact second and usually has a fraction.
> 
> There are two Items in the header that affect time readings
> And I'm wondering how to properly use them.
> 
> WINQUAKE TYPE4 HEADER DATA
> GIVEN: FIRST SAMPLE TIME 00:00:35.151
>        Hour?    == 00  byte value 0 to 23
>        Minute?  == 00  byte value 0 to 59
>        Seconds? == 35  byte value 0 to 59
>        SPACE HOLDER?   byte value 0
>        NANOseconds&    LONG signed integer value 4 bytes low endean
>        0 to 999,999,999
>        OFFSETseconds#  DOUBLE floating point value  8 bytes IEEE
> 
> 
> Find: What is the proper way to indicate
>       the fractional value which remains ?
> 
> A. Put the fraction 0.151 into the nanoseconds variable times one billion
> B. Put the fraction 0.151 into the offset area following the nanoseconds area
> C. CROP the recorded DATA to start at an exact second
> D. THERE IS NO WAY TO GET THE FIRST SAMPLE TO SHOW THE PROPER FRACTIONAL VALUE
> 
> What's the best answer here ?
> 
> I know if I crop the file to an exact cal mark second
> I get pretty good fractional times. I am guessing
> this may be the best way to go since I may be using
> WQ in ways it was not designed to be used.
> This will mean saving 4 minutes of modulo history
> instead of three. I do not want to do this
> unless absolutely necessary.
> 
> When I look at my own data I will assign
> discreet date-times UTC to each sample
> relative to a given calibration mark
> and not guess between samples.
> The cal sample happens at the transition
> between second 59 and second zero the first
> sample after this event and I call that the
> exact second and calculate all other sample times
> from that one reference point.
> WINQUAKE works different, possibly better,
> I just need to understand it better.
> The times I use are in the WINQUAKE "?"
> type relating to synchronization with WWV
> radio broadcast.
> 
> Thanks for answers,
> geoff
> 
> ps, I have found lots of errors in my older program
> leading to not so good time readings in the past.
> Every time I do this, the results seem better
> than before.
> But now I'm reaching the point of diminishing returns.
> 
> 
> __________________________________________________________
> 
> Public Seismic Network Mailing List (PSN-L)
> 
> To leave this list email PSN-L-REQUEST@.............. with 
> the body of the message (first line only): unsubscribe
> See http://www.seismicnet.com/maillist.html for more information.
> 
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)