Timing
Last update 2019 March 20
then 2026 July 25
Introduction
Accurate timing of observations is an absolute necessity. In FITS format the time from the PC clock is automatically inserted in to the FITS header.
For video observations of occultations or some other critical time-dependant phenomena, the time is inserted onto the frame from a GPS Video Time Inserter. Good PC clock time can also be achieved using planetary CCD cameras in SER recording mode using GPS disciplined NTP. These methods are accurate to 0.03s or better.
Since 2019, CMOS cameras have been the detectors of choice. USB3 frame rates can be >100 (0.01s) with no inter-frame delay. The addition of GPS receivers into cameras (e.g. QHY174mGPS or DVTI+CAM 430) have revolutionised the accuracy and simplicity of timing.
There are a number of ways of obtaining accurate time signals in order to set the PC clock;
– radio controlled clocks
– the internet (NTP)
– Global Positioning System receiver
Observatory set-up
Richard Miles took some time to come up with an arrangement that worked. See his diagram. Bear in mind that technology has now moved on. Parallel port CCD cameras are a thing of the past, and all are USB2 and data can be saved to large disks or memory card. Apart from components, the configuration is more-or-less unchanged for FITS imaging.
With the addition of a video camera for occultation work, the GPS unit now inserts UT into the video stream before it is recorded. Then by using an analogue-to-digital converter on a USB port, the analogue video with frame time stamps (to 0.001s) is recorded to a drive with software such as Virtual Dub, without using NTP
The purposes of the various pieces of equipment are;
Skysensor 2000 Alignment and go-to (using RA and Dec coordinates of object to be observed/imaged)
GPS receiver Provides correct time to PC clock.
CD-R drive Storage of images for later processing
Imaging and tracking is performed using Astroart and Megastar software. All equipment is mains powered – batteries tend to die quickly especially in cold weather.
For asteroid imaging orbital elements of the object(s) to be imaged are input in to Megastar and the tracks generated and displayed for the time of observation. The track is compared with the MPC ephemeris obtained at the same time as the elements and the CCD image compared with the displayed Megastar field. This should ensure that I am in the right place at the right time – so far it has worked well.
A similar process is adopted for asteroidal occultations. In this case the field of view is being monitored in real-time. For asteroids and stars of similar brightness the star and asteroid will appear as a close “double” reducing in separation. The occultation coordinator uses SkyMap pro and EQASCOM with a SkyWatcher EQ6 Any ASCOM compatible software could be used. Examples include CdC and C2A. It is advantageous to check the software web page to see if the asteroid orbit files are obtainable.
PC time keeping
PC clock
I initially relied on the PC clock to provide correct time (automatically inserted in to the FITS header). However this tended to loose about 10 secs or more per image so, during an imaging session, became somewhat inaccurate. Please be aware that a number of CCD cameras do ‘freeze’ the PC clock during image download so continual correction of this clock is necessary for accurate astrometry.
The most recent operating systems (e.g W7 to 10) don’t suffer from this sort of time freezing, and USB3 cameras are becoming the norm. Never-the-less the PC clock still needs to be regulated to achieve UT accuracy in the 10ms range that may be needed for some observations. W10 is least effected by system interrupts.
Global Positioning System
A GPS controlled PC clock is now common to many users (since 2019) with accuracy of 0.03s or better. More adept observers are using RasPi kits to control the clock (referred to as “disciplining”) through the network port. There is also a commercial GPS solution via USB. This is the TimeBox designed by Dr. Cesar Valencia Gallardo.
http://beyond-neptune.lesia.obspm.fr/sites/beyond-neptune/IMG/pdf/timebox_midavaine.pdf
Internet timing
A number of packages are listed below. The best accuracy one can expect from such software is probably of the order of +/- 0.03 secs, and +/- 0.01 secs with care.
TIP: Enable Seconds in the computer system clock.
The writer finds that hh:mm is of limited use. This can be changed to hh:mm:ss using a registry hack in Win7 (and 10) but is available to W11 in the clock settings. Worth Googling!
Dimension4
Dimension4 is a freeware program. It synchs to just one time server. You select a time server by clicking on the Server however you appear to get no indication whether or not the server is off-line if the server is offline. Actual usage, for astrometric timing, shows Dimension4 to be accurate to, or slightly better than, 0.1 secs. If you look in the history file the corrections will indicate if they are applied or not.
The D4 time updates can be abrupt, and not suited to asteroid occultations which require a smooth time service. Meinberg and NMEATime2 are preferably. The later can use a U-Blox7 GPS dongle which is easy to set up.
NMEATime2 and GPS Optical Flash (2026)
This is a hybrid control that can also be used to sync the computer clock, and find operating system recording delay.
In summary we need a GPS/GLONASS U-blox7 usb dongle at about £10. We also need the software: NMEATime2 from VisualGPS https://www.visualgps.net/NMEATime2/MoreInfo.html
This is free to use (as of 2026) and might need a licence for extended use. For the time being it appears to work on W10 without a licence installed. The GPS communication is through a USB port, which is found automatically from the setting’s menu. Time sync occurs within a few minutes, and the Loop Stats show the error is 5 milli- seconds after 10 minutes of operation, and 1-2 ms after 10min in operation.. There will be an operating system delay (OS_delay). Once measured, is applied to correct the observed times.
Michael Camilleri (Australia) produced a document describing how Optical Flash is used: https://occultations.org.nz/meetings/TTSO18/Camilleri%20-%20Flash%20Timing.pdf
This is also in JOA: https://www.iota-es.de/JOA/JOA2024_3.pdf
And Pierre Le Cam (Société Astronomique de France) describes his method for GPS FLASH clock correction <here>
The Occultation Coordinator has a summary document on GPS FLASH, in progress….
How have Timing and Recording methods changed ?
Method 2000* 2010* 2020* 2022-2026**
Analogue and digital video (AVI, SER) 17 363 590 11560
Sequential images (FITS) /CCD 98 150 500 2163
Digital SLR-camera/Drift Scan 0 3 7 15
Visual 235 218 12 9
Photometric 2 1 1 0
(*) From Euraster data base https://euraster.ericfrappa.com/results/index.html
(**) SODIS The founding manager of euraster.net, Eric Frappa, retired in 2022. The Stellar Occultation Data Input System was developed by IOTA-ES to collect and review observations. The current GB reviewers are: T Haymes, S Kidd, A Pratt.
Reporting Time:
International Occultation Timing Association (IOTA), and SODIS database (UK and Europe).
All occultation timings are be reported to SODIS via the BAA which will keep copies.
IOTA accepts all observations. Visual, video, CCD and driftscan are the common methods. IOTA and the BAA-ARPS can recommend how the timings are derived and reported. Asteroid occultation timings are published on web pages and used for professional study.
For UK and Europe, see: Eric Frappa Archive
Or Mike Kretlow’s collaboration: CORA
and Tim Haymes’ UK reports summary
| The British Astronomical Association supports amateur astronomers around the UK and the rest of the world. Find out more about the BAA or join us. |