Object centering 101


#1

So I’m just trying to understand how to get SGP to do what I want.

I start by setting up a sequence for a new object. I use a hyperlink from Astrobin to get the coordinates. I put that in for my object and the coordinates populate into the fields. Now I’m ready to start the sequence.

When it starts, the scope centers on my object, but its not quite the center I would like it to be.

So I tried taking out the astrobin link and I centered the object myself using the keypad in EQMOD. I then took the RA and DEC coordinates from EQMOD and entered those manually for the object in SGP.

So I start the sequence again and it centers the object again, but its still not what I wanted. Am I doing this wrong?

Thinking about it now, the one thing I did not do, which maybe I should have, was to perform a sync in Carte Du Ciel so it knows the true location of that object.

Thanks in advance for helping a rookie learn the ropes. If this all works out, I’m gonna buy a license when my trial expires.


#2

The first question to ask is what epoch are you using? SGP expects J2000 if you are entering JNow coordinates it won’t be right.


#3

I’m set for J2000.


#4

I’m trying again tonight and still noticing that it moves the scope a little bit when the sequence starts. I’m half-tempted to just turn off centering, but my primary goal is to get meridian flips to work. I want to be able to auto-flip and have the scope pointed at the same place after the flip when the sequence resumes. I guess my question is how does everyone else accomplish this? How to keep the same centering after a flip?

Thanks in advance for the help!


#5

Hi Don,

You can use the “Center Here” function. After the image downloads in SGP, right-click on the image where you want the center to be, then select Center Here and SGP will center it there.

There’s a better way though: use the Framing and Mosaic Wizard. I think it is by far the easiest way to setup a target in SGP. You’ll find it on the menu under Tools => Framing & Mosaic Wizard. It is a paid add-on, but it has a free trial period for you to use it while you decide if you like it (you will!!) The instruction page is here: http://mainsequencesoftware.com/Content/SGPHelp/FramingandMosaicWizard.html

You use the FMW to setup your image ahead of time with the target centered and rotated just how you want it to be in your final image. The FMW saves the location and rotation information to the target in your sequence so it is ready to go when you start your sequence. There’s no messing around at night during imaging time. Frame in the daytime at your convenience, then just run the sequence when it’s time for imaging.

Andy


#6

I guess I misunderstood your earlier question. Are you saying that the centering is not precise and repeatable, or that the centering is precise but not at the requested coordinates, or are you saying that you are looking for a way to frame an image? If you are having trouble with centering accuracy, you’ll probably need to post your SGP log file and give us the link to your reference image.

Andy


#7

Your first post looks interesting. Clouds just moved in so I’m done testing for the night, but I will look into “center here” when I get a chance.

So the main goal for me is to be able to do auto-meridian flips and have the target be exactly where it was before the flip. I THOUGHT it would be necessary to center the target at the beginning of my sequence so the coordinates of the target and my actual view in the camera would match. That way, when the flip is done, I know it will slew to the coordinates of where I want the object to be centered. Maybe that’s a poor approach.

So I started doing some testing by choosing a star near the meridian by maybe 30 minutes before a flip would need to occur. I copied down the coordinates reported by EQMOD and used that for the coordinates in my sequence. Then checked the box “center here when sequence starts”. Then click Run Sequence but when I do that, it always seems to move off from the position where I started.


#8

Here is my log where the object moved slightly.

sg_logfile_20151109182507.txt (557.9 KB)


#9

Are you using or loading a pointing model, ie loading up sync points in EQMod, that can affect where you think it should be, and where EQMod thinks it should be. I normally don’t load any sync points, and make sure the syncs are issued in Dialog mode where they don’t get added to the point model on each sync.

Lynn


#10

I have another thread going where I was asking a basic question about the meridian flip time and Ken pointed me in the direction of 2 threads regarding issues with EQMOD. Apparently I’ve got a lot of changes I need to make if I want this to work.

But to answer you, yes I’m using EQMOD with CDC to point to several stars and sync them for mount alignment. Then I launch SGP and go from there.

In the thread that Ken referenced:

One of the guys named Scott is saying its better to use plate solving to do your alignment, but I have no idea how this is accomplished. I sent Scott a PM asking for details, rather than post in the thread and hijack the conversation which is far more important to the community than solving my little problems. If anyone happens to know where is a tutorial on how to use plate solving to do alignment as Scott suggested in point “A” of his post, I would be grateful.

Also, I have never used “dialog mode” and not even sure where that is or how it works.

Thanks to all for your replies!


#11

With plate solving, I don’t worry with point models with my EQ-6 mount. While my mount is permanently mounted, if I need to re-align it because of changes, I simply slew to a bright star, release the clutches to manually line it up, the tighten them back. After a slew to the target, the first plate solve almost always places the object of my field of view.

The key thing about “dialog” based alignment, is a sync doesn’t create a new sync point in the point database, it just adds an offset.

There are a whole slew (sorry) of videos on Youtube explaining different aspects of EqMod. This is the one that discusses Dialog based alignment.

Good luck,
Lynn


#12

Don,

I see 4 auto-centering events in that log, all far below your tolerance setting of 100 pixels.

[11/9/2015 6:43:49 PM] [DEBUG] [Center Scope Thread] Automatic Auto Center Success - Total Error <= Allowable error: 7.4 <= 100.0 [11/9/2015 6:46:11 PM] [DEBUG] [Center Scope Thread] Automatic Auto Center Success - Total Error <= Allowable error: 13.3 <= 100.0 [11/9/2015 6:49:17 PM] [DEBUG] [Center Scope Thread] Automatic Auto Center Success - Total Error <= Allowable error: 2.1 <= 100.0 [11/9/2015 6:51:15 PM] [DEBUG] [Center Scope Thread] Automatic Auto Center Success - Total Error <= Allowable error: 6.7 <= 100.0

Could you be more specific about “moved slightly”? The log says that the worst case centering had an error of 13 pixels which is well below the 100 pixel tolerance you configured. If 13 pixels is unacceptable, you could decrease your tolerance to whatever precision you need (perhaps increasing the number of attempts if you find that SGP cannot reach the desired precision in the 5 attempts you have configured.)

Andy


#13

Well to be specific, there was one test I did where I was centered on a fairly bright star which was not dead center but maybe a little left of center. I input the coordinates from EQMOD into the object data for SGP and started my sequence letting it center before taking exposures. After centering it moved the bright star down to the bottom of my camera’s view. My frame size is 36’ by 46’ so I would guess it moved the star down by about 17-18’ and maybe to the right by 10’.

I guess I will repeat the test on the next clear night and try to be more studious about taking notes and screenshots.


www.mainsequencesoftware.com