why 38 the value of zeta Aur ?

Hi, GMara. According to the best photometry we have, zeta aur nominally is V=3.750 with an uncertainty of 0.030. The source of this photometry is the Lausanne Photometric Database (GCPD). So for the visual chart we just rounded 3.75 to 3.8. You can see that 3.72 falls within the uncertainty of 3.75 so photometrists are justified using 3.72. I'll leave it to one of the photometrists to explain exactly why they chose to do so.:) A side tip: To find more precise values of photometry on star charts, you can use the AAVSO Variable Star Plotter. Just plot the field interested in and click the "Photometry" checkbox. On the results page, find the label of the star you are interested in. Click here to see the list of photometry for stars around epsilon Aurigae. Look for the star labeled 38. That's the one we are talking about in this thread.

Thanks Aaron for the quick reply! Yes, if we have this value (3.75) then i can understand why it is 38 in visual maps.G

Hi,In that case, does it make sense to record any visual estimates of eps Aur to two decimal places, if the given, approximate, magnitudes of the comparisons are already in error by a similar amount?David Conner.

Hello!Well from what i can understand is that we could be able to give an accuracy of 0.05. Since the accepted value for eta aur is 3.75 and in visual maps 3.8 then 0.05 is an acceptable error. We can limit our measurments to our error bars. Best wishes,Grigoris

Hello,I think there are a number of points here that I am not clear about.1) If an observer can accurately determine the apparent visual magnitude of a star to within an uncertainty of +/- 0.05 mag, then should we not write 3.80 instead of 3.8, to distinguish it from 3.75 or 3.85? 2) Can most/many/some visual observers actually make visual magnitude estimates that are accurate to +/- 0.05 mag? We might feel we can, but there are times when I have difficulty in determining which of two stars is the brighter. Sometimes one appears brighter than the other only for the opposite to appear true a few seconds later! In my own experience this means they are within 0.1 mag of each other at best. 3) Aaron's posting of June 2 2009 on the 'How do you get an accurate magnitude estimate visually?' thread ("For example, in my last observation I struggled between calling it a 3.4 or 3.5. So I submitted it as a 3.45. ") touches on this. If we cannot decide between mag 3.4 and mag 3.5 is this because we are actually dividing the magnitude range 3.2 to 3.8 into 12 accurate steps by eye, or that we cannot even divide it into 6 steps accurately? This topic might be a subject that we in the Mark 1 Eyeball Team could usefully investigate.David Conner.

These are great questions! I'll try to answer. 1) The AAVSO method of interpolating magnitudes between two comparisons, assumes that the typical error in a visual observation is 0.2v. So reporting to a precision of 0.1v seems appropriate, and for the stars the AAVSO typically collected visual data on in its first 80 years or so, this is perfectly acceptable. If you had two comp stars whose difference in magnitude were only 0.1 or 0.2, and you were a very experienced observer, I don't doubt you could tell the difference between them and a variable to 0.05. I've done it myself. The AAVSO method is not the only way to do visual observations. In fact, we may actually be in the minority when it comes to visual observations. Many of the European observers use the Pogson, or fractional method, where no magnitudes are used at all when making the observation, just letters. Then the magnitude is derived later. So at the eyepiece they record something like B-2-V-4-D. They are reporting that they could distinguish six steps between comp stars B and D and the variable was 2 steps fainter than B and four steps brighter than D. When they do the reduction they use the magnitudes of their comparisons, which may be to two decimal places, and simply the divide difference between the two comps by the number of steps, most often coming up with a fractional number. If your comps were 11.10 and 11.60, you are estimating the magnitude using steps smaller than a tenth of a magnitude, and may well be discerning steps smaller then 0.1v. Depending on who you report data to, you may actually state your observation to two decimal places.In this case, 11.27. If you think you can determine the magnitude to less than a tenth of a magnitude, (and aren't just guessing!), I don't have a problem with observers submitting 3.45 or 3.65 as an estimate. In the early stages of the eclipse of eps Aur I did this myself. Now that there is a larger gap in the sequence in its current state, I don't feel comfortable trying to do that. 2)There are some visual observers who can make incredibly accurate visual observations, approaching 0.05 or better precision at times. Sebastian Otero does this on some bright naked eye stars. In fact, he claims this is easier to do with a semi-bright urban sky background than in a dark country sky, so take heart city dwellers! He has his own methods and has practiced on a special set of stars for a long time, but it can be done. He regularly observes stars with amplitudes of only a few tenths. Most of us are not Sebastian. I'm pretty confident in my own estimates at the 0.1 level, as long as there are enough well spaced comp stars in the sequence. 3) I don't think the human eye/brain combination can reliably divide the difference between two stars into more than about 6 or 7 steps. That is why we strive to include comp stars every 0.3 magnitudes or so, if possible. I don't now if there is any empirical data to support this, but I've talked to almost all the best visual observers in the world at one time or another about this, and they all seem to agree, more or less. Mike Simonsen

I presented a poster paper at an American Astronomical Society meeting in 2007 reporting on results of a quick-and-dirty look at visual precision and accuracy. Our conclusion was that visual precision+accuracy (we didn't differentiate in this study) in large data sets (like for the star Mira, which has over 50,000 observations by amateur astronomers in our database!) was 0.2-0.3 magnitudes. We found a ton of variables that affected visual obs including phase of the Moon, distance to the Moon from the star, time of night, color of star, etc. Surprisingly, we did not find a substantial effect caused by observer experience. Remember, that was for a large database with thousands of different observers. Individual observers can often do much better. Wayne Lowder, for example, was a legend who could do 0.02 magnitudes visually (many photometric observers can't do that!). Click here for an example of his work. So it's important to draw the distinction between the accuracy of an aggregate and of an individual. Also, in the post David alluded to earlier, we can see how statistical processes can be used to greatly increase the accuracy of a large visual dataset.All of these questions are ones that we hope to address in the next year with the Mark I Eyeball team. Expect some results published by early 2011. We hope to put these questions to rest. :)

Hello! [answering to david's questions - sorry to be late on that!] 1) Yes, I agree that the correct way to present this is to put 3.80 so as to show that the accuracy goes to the last digit. 2) I totally agree to this and ... 3) I think that the comparison stars must be enough close (like what you mention) to determine a half mag estimation. Grigoris

Hi Mike,I belong to the British Astronomical Association's Variable Star Section and use the fractional method of making estimates, like your example B-2-V-4-D. Whatever the actual numerical magnitude this gives, we round the result to the nearest 0.1 mag. So 11.27 would be rounded to 11.3. If we were to report a result of 11.27 as in your example, this would imply that we could distinguish between mag 11.26, mag 11.27 and mag 11.28. Estimating magnitudes to 1/20 of a magnitude is one thing (i.e.+- 0.05 mag), but to 1/100 of a magnitude (+- 0.01 mag) is probably a bit optimistic. It seems like we are making a case here for actually quoting error bars/uncertainties with our observations, in order to remove just this sort of ambiguity. With hindsight, it seems surprising that this isn't done. (The BAAVSS report forms include a 'class' comment for our visual observations; class 1 means +- 0.1 mag, class 2 means +-0.2 mag, class 3 means 'better than no observation at all, but not very accurate!')David.