[RASMB] Time-stamp checks for SV data

[Sherwood at bbri.org]

Reagrding the timestamp discussion:

In Windows, each file has at least three distinct timestamps: create 
time, write time (last modified), and last access time (last read or written).

It is likely that the Beckmann sw does not create a file at the 
instant a scan is started, or the instant it is completed. More 
likely, there are two or more buffers into which scan data is stored 
as it is acquired. At some later time, the buffers will be written to 
disk. This is the common technique of "double buffering", used in 
real-time systems so disk operations don't keep the system from acquiring data.

The file creation timestamp, then, is probably later than the scan 
start time, but not necessarily the same.

To complicate matters further, sw could create some or all of the 
scan files before the corresponding scans begin. As well, Windows 
allows sw to change the file times after files have been created.

Therefore, I don't think you can place any significance on the file's 
timestamps. The definitive experiment would be to sit at the 
centrifuge and record the clock time when each scan begins and ends, 
as determined by mechanical means. Then those times could be compared 
to the headers and file timestamps.

Peter Sherwood