We are moving all the time.
Our observatories on the earth surface are moving too fast and the earth is moving too. If your observatory it’s close to the equator the speed is more than 1600 km/h on the earth’s surface.
We are using a long focal distance in our telescopes, if you try to shoot a very narrow place on the sky and if the target is not fixed like the stars, if it’s a satellite, asteroid or any other moving object, the task to reach them is question of precise positioning and precise timing!
At my remote observatory I am using several remote clock servers over internet with a typical delay of 50ms. If you are using a software to compensate these changes you could work with a relative accuracy of a few milliseconds. Few milliseconds could be 1 arc second error on the position for “fixed” objects, but for “moving” objects the error is bigger due you need to catch them on the right time at the right position. This is a problem specially on really fast objects like ISS (27.600 km/h).
I was build something called Stratum1 time server using a GPS signal with a precise PPS pulse.
A Stratum 1 server accuracy is below a millisecond measure.
In my implementation the StdDev of the clock offset is about 8.3µs on two days logs. Just for who doesn’t know about µs (microseconds), in seconds is 0.0000083s or 0.0083ms accuracy 😀
It’s possible to look on last 24h statistics (by hour) on this site: http://astro.carballada.com/allsky/day/
The other elements on my local network are syncing with this Stratum 1 server with a very low latency. The corrections are also under the microsecond order.
First computer has an Offset of min -852µs and max 754µs.
Second computer has an Offset of min -401µs and max of 301µs.
This figures are from W10 machines with the NTP client from Meimberg, available here.
The NTP configuration file looks like:
# Use specific NTP servers server str1clock.local minpoll 5 maxpoll 5 iburst prefer
I think that kind of corrections are really related to the microprocessors temperature compensations (the internal clock speed change in function of the temperature).
I hope to operate more easily my observatory and obtain more accurate results on my future observations and…. if not, I am really happy to dive on all this topics and prepare all the necessary hardware 😀