The determination of the exact time of observation has always been a fundamental requirement in variable star research. The accuracy of professional visual observations (80-100 years ago) was around 10-30s, while the accuracy of time of photographic or photoelectric observations has always been kept around 1s. Of course, the same accuracy in time can be achieved by CCD observations. If the accuracy in time of the observations is 1s, it does not mean that the time of certain light curve parameters (light maxima, minima or any given phase) can be derived with an accuracy of 1s. If it were true, then this accuracy of times of light maximum or minimum could have also been achieved by photographic or photoelectric observers. In fact, the accuracy of determination of the times of light maximum or minimum depends very much on the accuracy of the observations themselves. The errors are typically larger than 0.005 mag for small telescope CCD or photoelectric observations (depending on telescope size, stellar brightness, quality of CCD chip, etc.). The 5-15s exposition times further complicate the situation (during this time small changes in the colour dependence of air mass, guiding errors of tiny refraction changes, intrinsic changes in the brightness of the variable, etc. may occur). Possible asymmetries of the minima (or maxima), the correctness of the fitting formula, etc. also influence the accuracy of the determination of the time of light minimum or maximum. Taking into account the heliocentric correction may also be a severe problem. The barycentric correction is usually ignored. As the difference between the heliocentric and the barycentric light time may reach 6x10^-5 day it can influence the 4th decimal of the JD! If somebody claims that his/her light minimum or maximum data are accurate to 10^-5 day (=0.864s) he/she should give adequate information about the procedure of the determination of time of light minimum or maximum and the error estimate.

Technical issue on the value of the standard deviations

Standard deviation may serve as a measure of uncertainty. The reported standard deviation should give the precision of those measurements. In practice, one often assumes that the data are from an approximately normally distributed population. If that assumption is justified, then about 68% of the values are within 1 standard deviation of the mean, about 95% of the values are within two standard deviations and about 99.7% lie within 3 standard deviations. This is known as the "68-95-99.7 rule", or "the empirical rule". This also implies that in a measured numerical value only that digit is meaningful which has at least the same order as the standard deviation. To give further lower digits are meaningless because they do not have any additional information on the measured value.

Times of minima observed in different colours

The times of minima of eclipsing binaries observed simultaneously in different colours should be averaged. Different values of the same minimum obtained in different colours can be separately published only in cases when the time difference has a justified physical reason.

The Editors