Calibration and guide rates
Pascal Mora
I encountered some little problems with the calibration process, PHD2 says me that guide rates are varying in unexpected manner, after looking at the logs, I saw that the guiding rate in RA was 22 arcsec/sec, the DEC rate is correct (guide rate in x1 and I use ST4 guiding), how such value is possible since, normally, the RA motor is just stopped and the resulting rate should be near to 15 arcsec/sec? I verified all my camera parameters and all are ok.
I you have an explanation or if you can help to correct this, I will be happy.
I have also two minor questions about the calibration review window, it seems that you use only the first and the last star position for the rates and camera angle calculations, this may give inaccurate results when seeing is not good, why not to use linear regression to minimize this? And finally a very minor question : why RA is red in the calibration review graph?
Thanks you for your patience.
Pascal
bw_msgboard
Hi Pascal. There seems to be a lot going on here. The first thing I noticed is that you sometimes used overly short exposure times for the calibrations. With an image scale well below 1 a-s/px (tiny pixels), you should probably stick with longer exposure times (e.g. 2 sec) as you did in the middle sequences. With very short guide exposures, you are getting a lot of deflections caused by seeing – this is especially evident in the later sequences, and it can complicate the calibration process. Part of this might be because you were calibrating at a pretty low altitude – not much above 30 degrees, it appears, which is the two-air-mass level. If you want to take a hard look at calibration, try working much closer to the meridian and close to Dec = 0. It also looks like you may have changed guide speeds from 0.5X or 0.6X to 1X in the middle of the session – is that true? You might want to make a few runs with the Guiding Assistant to see if periodic error or RA drift is having an effect on the calibration.
As far as the calibration rates go, high precision isn’t really important because of the randomness of the whole guiding process. Your RA rate does seem somewhat high but that’s not likely to cause you much trouble It would generally cause you to under-correct a bit but that’s better than over-correcting. That’s one of many reasons why these alerts are just advisory – you need to decide for yourself if what we’re seeing is important or not, and it’s nearly impossible for us to set limits on these things. In your case, I think the seeing conditions are the source of most of the variability. With the ASCOM mount, many of us get a calibration under good conditions and then re-use it for long periods of time, often months.
Your are correct that the review graph is using only the first and last star positions for these calculations – but that matches what the calibration process itself does. J We have talked about using a fitting algorithm here but it’s not clear the benefit warrants the extra code. In nearly all cases, the current calibration algorithm produces results that are quite accurate enough. But maybe we would re-consider this since you brought it up with real-world data.
As for the color, I guess that reflects my personal frustration with the default colors. <lol> As soon as PHD1 provided color options on the graph, I adopted the simple rule that Red = Right ascension and I never had to think about it again. So I frequently get into trouble with the LogViewer tool where I can’t change the colors. In my defense, the calibration review graph does show a clear legend for the color choices, so I figured that was good enough. I dunno, maybe I’ll have to throw in the towel and reverse the colors in the review dialog… J
Have fun.
Bruce
--
You received this message because you are subscribed to the Google Groups
"Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to open-phd-guidi...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Pascal Mora
thank you for this very detailed answer.
Concerning the altitude at with the calibration is done, I thought that PHD corrected the guide rates accordingly, but ok I will take care of this and also of the exposure time.
Yes I changed the guide rates for testing, because I observed some oscillations in RA correction, but may be I would have done better to simply reduce the agressivity. I read somewhere that calibration must be redone when telescope was moved, but may be it is not absolutely necessary. For the colours, no problem ;-) ....
I asked these questions because, as many of us I think, I would like to have the smaller and the rounder possible stars, and paradoxically, I observed that it is much more difficult to obtain this result when the seeing is good. By example when the DEC error rms fall under 0.3-0.2 arcsec, RA error is often significantly greater. In such nights, an 0.1 arcsec difference in RA and DEC errors may lead to have oblong stars. My mount is an 'home made' mount (I think seriously made ;-)) I measured a low EP (less than 2sec pic to pic) with some low drift (verified and corrected as well as possible at the beginning of all image session, as the mount is again in test phase), so the 'additional' errors in RA may be caused by the mechanic or by a bad use of the guiding software. I have spend a lot of time to improve the first, now I try to making the most of PHD2, but it seems that calibration was not the source of my problems ... Just an idea, may be it would be interesting to know the ratio RA/DEC rms error (a kind of expected roundness index ...) on a certain period of time (related to the used exposure for example) in order to estimate the guiding quality, the graph already present in PHD is useful but not always readable and usable to estimate the final result. But I can be totally wrong on this point.
Thanks.
Pascal
bw_msgboard
Hi Pascal, maybe I can clear up a couple of points. First, if you are using an ASCOM interface to the mount (‘mount’ or ‘aux’), you *don’t* need to re-calibrate when the scope is moved. PHD2 will automatically adjust for side-of-pier and pointing declination so you don’t need to bother with that anymore. That’s why it makes sense to get a good calibration that you’re happy with and then re-use it. The only time you need to recalibrate is if the guide camera is rotated by a significant amount or you are imaging very near the pole. PHD2 does adjust the calibration for declination but not for altitude. There’s really no need to do the latter because the change in altitude, from a guiding point of view, only changes the atmospheric refraction. This is a slow, uni-directional change on each side of the meridian, so it is easily handled by normal guiding operations. My point about calibrating at a higher altitude was simply that you will be looking through less atmosphere and will be less affected by seeing.
One question I should ask: is the 1121mm focal length in the log really the focal length of the guide scope? Or the imaging scope? It needs to be the guide scope for our purposes (with an OAG they are the same, of course). The guiding RMS numbers you show look rather low unless you are working under very good conditions, so I thought I’d ask. The issue with round stars is what we all wrestle with. The real goal, I think, is to have round stars whose size is limited only by the optics and seeing. So as you said, with really good seeing, it becomes a bigger challenge. I think the RA RMS numbers are nearly always a bit higher if only because the RA motor is always running. And of course, star elongation can be caused by things other than guiding. Tube currents in the summer and mis-collimation are two common sources, issues I encounter every year. It can be helpful to take a sequence of 6-10 sec exposures with the main camera and look closely at the stars. If they are somewhat elongated or the elongation varies strongly across the field, you know the problems are optical rather than guiding.
I have given some thought to a metric that compares the RMS values on the two axes, but I haven’t found anything that is meaningful. The problem is that we don’t know the seeing or, for that matter, the star sizes in the main image. Just to take an example from my last imaging session, I ran for about 4 hours with a total guiding RMS of about 0.5 a-s. But the RA RMS was larger than Dec by maybe 30%. Does that mean the stars showed a 30% elongation – no, not at all. The reason is that I was imaging stars that were 2 a-s FWHM in size, and the measured aspect ratios were single-digit. So working with the data that are available to us, it’s hard to see how we could make any estimates about the aspect ratios in the final image. The closest thing I’ve seen is a rule of thumb that says you want the guiding RMS to be <= 25% of the stellar FWHM (see: http://www.innovationsforesight.com/education/how-much-guiding-error-is-too-much/)
Have fun.
Bruce
From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Pascal Mora
Sent: Sunday, October 04, 2015
10:23 AM
To: Open PHD Guiding
--
Pascal Mora
I use an ascom interface, so the first point is ok, I have just to get a good calibration and I will keep it until the end of the cession.
I use an OAG, so the guider focal length is the same as the mean camera. I don't have optic problem, my stars are perfectly round when I use short exposure (this was another history but now it is ok ;-)) . I think the seeing in my site is in the average (between 2 and 3 as) near to 0.5-0.6as in term of total error RMS, but sometime better at the middle and the end of the night.
I have already read the article you mentioned, very interesting, but difficult to link to the data given by PHD, how to estimate the tracking error alone? RA error rms minus DEC error RMS ? you cannot use the total error because if RA and DEC error are the same stars, will never be elongated (just bigger).
A think I have not really taken in account is the image sampling, mine may be is a bit too low.
Thanks.
Pascal