Peter Tickner wrote to Eric Frappa cc. Tim Haymes about our observation in 2028 September:

I ran the second laptop displaying NTP time seconds using a London NTP 
server.  I could see in the videos taken of both laptops in action 
that the seconds displayed from my GPS TimeBox agreed with the NTP 
seconds displayed.  I was then able to check with TANGRA the start and 
end times on two of the three SER videos I took and compare them with 
the displayed GPS and computer clock times that I had videoed as I 
started and finished the SER recordings.  They are all in the same 
second as the SER recording with the laptop as usual slightly ahead by 
a few thousandths of a second from GPS in three of the four timings. 
It is slightly behind in the other.

This is not a detailed scientific test but it does strongly suggest 
and support my view that it is unlikely my SER files have a timestamp 
that is at variance from GPS of more than the built in time between 
each video frame being recorded at the frame rate used.  Also any 
variation is highly likely to have been less than a second and 
unlikely to explain our time differences on this event.

I hope that is helpful to know!

Kind regards,

Peter

PS I will try to video as many of my events as I can in future to 
check laptop and GPS time to confirm the SER file time stamps.