TPoint for the visually inclined
(Imagers should read this too)
Last time I talked about how to do a quick polar alignment with one of the Paramounts. Really, not too different from any other German equatorial mount. Serious observers (visual or photographic) know that a good polar alignment is only the beginning of the story however.
TPoint as I see it has three purposes. Purpose #1 is to build a pointing model (and this is the basis for the other two major features of TPoint, which we’ll talk about later). A pointing model is a set of mathematical equations (and coefficients if you want to put a fine point on it) that predicts your mount and scope's pointing behavior. For example, if every time you pointed to an object in Ursa Major you were two degrees too high, TPoint would figure this out, and when you point to an object in Ursa Major, it would automatically point two degrees lower so that you’d point right to where you really wanted. This is a simplified example, but it does make the… um…. point.
Now, as I said last time if you have a good polar alignment, you have your 'scope nice and orthogonal to your mount's axes, and you have a sufficiently wide field of view, you could be done. Visually it’s not that tricky to pick out your object, and center it up in a wide field eyepiece with the hand controller. Then perhaps you switch to a higher-powered eyepiece. If however you would like to skip that step and get better accuracy on the first attempt, or perhaps your using a camera with a smaller field of view, you’d like to get better results on the first slew.
Where some methods of modeling your 'scopes pointing behavior can use only a couple of stars or samples, TPoint can use as many samples as you want to give it. In fact, the more the better, and it does an amazing job of figuring out all the different little misalignments in your system. With just six points (the minimum for TPoint’s system model), TPoint can outperform anyone’s two or three start alignment models and has a pretty good idea of just how 'crooked' your stuff is.
Assuming your not wildly off the pole, you can build a model with as few as six samples that will dramatically improve your pointing all over the sky. Quite possibly centering your objects in a moderately high-powered eyepiece every time. Furthermore, the more samples you add, the better it gets.
Let’s walk through step by step how to initialize TPoint and add “pointing samples” to that mathematical model I was talking about.
Step 1: Sync your mount
I mentioned this last time; it bears repeating, but only in summary. Slew to a nice bright star and center it in your eyepiece or in your camera’s view. Tips on how to best do this are in the last article. Sync the Paramount on this bright star using the command on the telescope tab. Stay on this star… you’ll need it again in a sec.
Step 2: Add your first Pointing Sample
Since we are already on a star and have done all this work to get centered, this is an excellent first star to use as a TPoint sample. Select “TPoint Add On” from the Telescope drop down on the main menu. Select the “Setup” tab, and then on the drop down labeled “TPoint Add On Settings”, select “New”. It will ask you if your sure you want to clear out your old pointing data, and since we are talking about portable use here, the answer is always YES.
Now, dismiss the TPoint window, and go back to the Telescope Tab. Click the “Add Pointing Sample” icon. Now you’ve started your TPoint model, and we want at least five more samples.
Step 3: The next five samples
The next five points can potentially be the most challenging to acquire for a freshly setup scope and mount. To get the next five points, select a nearby star and slew to it. It works very well if you just work your way around the bright stars in a constellation. Your first six points do not have to be all over the sky, and in fact, they can come quicker if you choose them all near each other. Not too near, but let’s say work your way around the bright stars in the big dipper for example (or select another constellation, now you have an idea now of how far to make the spacing).
When you slew to the first new star, it might not even be in your field of view if your initial polar alignment was not that good. Center the star using the hand controller or the jog commands from TheSkyX. There is also the “Star Search” command under the telescope tools menu we mentioned last time if your having trouble finding it.
Center the star, and add the sample. You may get the dreaded, “Pointing Sample Appears to be in Error” message. Assuming your not really “in error”, what this means is that your initial polar alignment is not terribly close, or some other parts of your system are not that well aligned. It actually means your mechanical uncorrected pointing is already at least 30 arcminutes off at this location. Not to worry, TPoint will take this into account, go ahead and tell it that you’re sure you are on the right star. It’s not that unusual to get this on at least two or three of the first six stars unless you have an exceptional polar alignment right off the bat.
An Aside: When I was first learning to use a Paramount… this is embarrassing… but I was using a live video camera to center the stars. I had a terrible habit of clicking on the sky chart to make TheSkyX the top application again, but in doing so, I changed the currently selected object. I’d then add this point as a sample, ignore the warning message (it’s there for a reason!), and well… chaos would ensue. Have no fear of adding a point that “appears” to be in error, but make SURE it only “appears” to be, and your not about to add a wildly inaccurate sample.
Step 4: Rinse, Repeat
Even if you're getting the dreaded “appears to be in error” message on every single star, you will suddenly get an amazing improvement in pointing accuracy after the sixth star when the TPoint model actually starts kicking in. Visual observers (or maybe really wide field photographers) might even quit here. If your initial alignment wasn’t very good, you will still get substantially improved pointing with just this small a model.
If you have a long focal length, or your other errors are still substantial, you might want to improve your model further. The next best advice I have is to have your first six samples on one side of the meridian, but at least one or two on the other side as well. In fact, with 12 points (six on either side) TPoint pointing can be quite amazing already. I can’t quantify this exactly because it varies with focal length, field of view, and all manner of mechanical considerations in your scope and setup.
Oh, but wait, we skipped something very important! After the first six points, bring up the TPoint dialog again, and select the “Model” Tab. You will see a display that tells you about how close you can expect to be to your target when you perform a slew. There is a button on this dialog called “Super Model”. Click it. Super Model does a more in depth analysis of your pointing errors and can make a further dramatic improvement in your pointing accuracy. Twelve points, well distributed across the sky dome followed by a Super Model will virtually guarantee superb pointing performance for even the most demanding visual observer.
So, now all night long you could conceivably add sample after sample all over the sky as you casually do your observing from target to target, each time further improving your pointing model and performance as you go (don’t forget to rerun Super Model from time to time). But as the man on TV says, “But wait, there’s more!” I never got back to the other two purposes of TPoint. Once you have a pointing model established, TPoint can TELL you what’s wrong with your alignment both qualitatively and quantitatively. You can use this information to adjust your mount, or your equipment to improve your polar alignment even more, which for imagers is no small boon. Further, at the end of the rainbow is the proverbial pot of gold… ProTrack. More on these two topics, and how to automate large pointing runs next time...
09-15-2012 12:38 PM