All times UT unless otherwise noted

How to report "Full Precision" timings

Derek C Breit - N & S America Lunar Occultation Coordinator

There is a Video Tutorial available that runs through the steps involved in reporting a Total Lunar Occultation.
It's at:  https://youtu.be/eCU0e4URZz4

Reporting a Graze?  Try: https://youtu.be/JlhLvF8i_5k

And don't forget to finish the process by reporting the light curves.  These can be done while working on the report
itself or can be done at a later time.  Light Curve "how-to" videos are listed on the YouTube page referenced below.

At the bottom of the IOTA YouTube page are several tutorials you may find helpful.  They may not provide every detail 
needed (if you have something special or unusual to report), but they will give you a very good start.

http://www.asteroidoccultation.com/observations/YouTubeVideos.htm

Occultationists using a stopwatch, WWV, or similar, should find the proper procedures in the IOTA Observers Manual

" What Precision and Accuracy do I report for my observations when using video and an IOTA-VTI or a KIWI OSD ? "

Reporting event times to a fraction of the frame has some difficult issues associated with 
it. Most methods of determining a sub-frame time are dependant upon an _assumption_ that 
the exposure on the video CCD occurs over the entire frame or field interval, that there 
are no effects caused by stellar diameter, that the star is not a previously unknown close 
double star, and that Fresnel diffraction is adequately dealt with given the sampling 
resolution of the frames or fields. The problem is - the reported uncertainty needs to 
allow for the assumptions that are being made, and many of those assumptions are not easy 
to recognize. Thus while the assumption about exposures _might_ be reasonable for an 
asteroid occultation, the assumption is not valid in a lunar occultation where you have 
the bright moon affecting the camera. And there is no basis to make assumptions about 
close double stars. We have no trouble accepting data with a precision associated with the 
frame or field interval - as that is the sampling interval that is associated with the 
time stamps. But asserting a timing uncertainty significantly less than the frame or field 
interval is only appropriate if you can demonstrate the validity of your conclusion on the 
basis of your entire system. Running software over the output signal does not address 
issues that are inherent in the video camera - so that the purported accuracy may be quite 
illusory.

It needs to be noted that the extra precision is of no practical value with our current 
analysis. In practical terms, it is a bit like making an item of furniture, and measuring 
the length of a piece of wood you are going to use to the nearest micron. Yes it can be 
done - but the extra precision is not used. In this regard, it should be noted that 
essentially all lunar occultation reports based on video, report uncertainties of +/- half 
the frame interval. This is a valid uncertainty measure which is fully justified. I would 
encourage all observers to report at this accuracy - unless they can demonstrate under 
controlled conditions that their entire system under observational conditions in fact has 
the asserted uncertainty.

Virtual Dub and LiMovie need to be downloaded and installed and the user familiar with them, then..

> Use LiMovie to measure the signal of your AVI to the Frame or Field level, as appropriate. (Fields shown unless noted)

Click the 'GRAPH' button at the lower right (as depicted above)

Select a dot near the midpoint of the lightcurve (as depicted below)

You also find an 'OFFSET TIME' and an 'ERROR'. These are to be ignored for the reasons stated by Dave Herald at the top of this webpage. Frame '486.0' is all you need. You will note that '486' is displayed at left side of the main window under the word 'Current'.

***See farther down this page for "How do I know I picked the right frame or field?"

Close the Diffraction Window, and the plotted graph, and click the 'Field Show' button (lower right). Scroll the window with all the numbers at the upper right to display the target field, in this case '486.0', as depicted below.

> In the window with all the numbers, it says field 486.0 ends at 4h 29m 11.167s. Below that it says that this time is the end of field 1 and that there is a field 2. Looking at the main window, field 1 is therefore field 31179, not 31180.

FIELDS

Center time of "Field 486.0" = 4h 29m 11.158s +/- 0.008s and is the 25% light level and is what should be reported.

This is "Full Precision Timing".

FRAMES

If you had done all of the above without ever clicking the "Field Measure" button, you could have directly read the time as 4h 29m 11.167s (the end time of field 1). Therefore the time to report is thus.

4h 29m 11.167s +/- 0.017s

Both of these Accuracies should be rounded up.

+/- 0.017s = 0.02s for Frames, +/- 0.008s = 0.01s for fields

If you need any assistance with this, let me know.