Taking flats with a rotator


With help from chasmiller46 and Ken, I finally have my rotator up and running. But I’m not sure how to get the rotator into the position I need for flats. I’d like a set of flats with the rotator at the same angle as the light frames I used it with, but I’m not sure how to get the rotator there. I can’t just set the angle in the target settings dialog box, because that requires a platesolve. And I’m hesitant to mess with the rotator’s position settings in the control panel, both because the number shown there is way off, and I’m afraid that messing with it might lead to a cordwrap. Suggestions?




Admittedly, our flats workflow is not ideal for matching angle in a different session. I will give this a little thought and try to figure out how to present a workflow that will do this (hopefully with no user decisions or options).

In the meantime, you can turn off the auto rotate functions of the target, inspect the FITS header of the light image you are trying to match and find “POSANGLE”. This is not a sky angle and asking your rotator to move to it will create the same orientation in the imaging train.


I’m afraid I gave up on using SGP with a pyxis rotator and OAG. The primary angle that drives imaging and flats for me is the OAG angle - which may be different from the image angle.

When you start SGP it appears to use a given angle - or something - for the “set angle” value - but after a plate solve everything shown is the image angle - I think.

The actual rotator angle is in the fits header - I think - but I had trouble having both sgp and pyxis having access to the rotator at the same time. I need to set the actual angle of the oag using the actual angle of the rotator - and I can’t do that with sgp. It only lets you set the image angle.

So currently I just use the pyxis controller to set the oag angle and it does not go in the fits header. I have to write it down in a notebook - and then make sure flats match that angle. There didn’t appear to be a way for me to take images based on a measured image angle in sgp - and then later tell sgp to go back to that same angle for flats just by specifying the pyxis angle.

Long ago I requested an option just to use the actual angle of the rotator for setting the angle for an image - and ignoring the plate solve angle - but no changes were made. So I use a notebook.



Ken, to move the rotator to the POSANGLE shown in the FITS header, can I use the Set Position box in SGP’s control panel? Or do I have to use the Optec software? I’d rather not use the Optec software unless necessary, because I’m afraid that may change the “zero” position of the rotator, and lead to cordwrap.

Frank, as near as I can tell, the setup is working fine. I don’t change the angle of my OAG with respect to the main cam FOV, so that may explain it.


Yes, this is preferred. The angle entry field in the control panel is indicative of positional angle and not sky angle so using it this way should get you what you need,


There will always be some difference between the angle the rotator is at and the sky angle of the image - and I believe SGP keeps track of that difference - but I’m not sure if it is stored in the equipment profile or not.

If you install a rotator and use SGP to tell it to go to 10 degrees for the first time - it will go to that value and display ‘10’. But later if you plate solve it will read the sky angle of the image - and calculate the delta - and display that new angle, say 20 degrees, in the rotator position location. So it changes even though nothing moved.

If you then shut down everything and take flats the next day in daylight, and then load the same profile and sequence, should you set the angle to 10 or 20 to take flats? The fits header will have both angles in it - but which value does SGP want if you just start up in daylight and can’t do a plate solve?

I don’t know - but I didn’t like the fact that the raw angle is not displayed - and the angle shown will change. And if you do use the raw angle somehow - you have to open the fits and read the header to know what it is.

I guess if you had the camera angle set exactly to match the sky angle it wouldn’t matter - but I am often changing things in my setup and the camera angle isn’t critical to me - whereas the oag angle is. I don’t want to have to align the camera angle exactly. The pyxis has a feature built in to set its zero any way you want - and that allows me to set the zero of the oag angle without physically turning it exactly right.


Is "side of pier" recorded in FITS headers?

another vote for enhancing the flats flow - i finally built a altniak compatible panel so now at least i can get the whole auto-exposure thing going properly. i’m enjoying that.

if i try to do the flats in the evening (right after starting the rig) then i have to use mechanical angles; if in the morning i will be stuck taking flats at whatever the last rotator angle was, or if i happen to be awake, i have to use sky angles.

and i can’t set the angle in an evening flat target because that seems to trigger a solve and sync, which… doesn’t work when pointed at a flat panel :slight_smile:, so in the evening i pretty much have to set a (mechanical) angle by hand and then run my flats sequence.

finally, i noticed that if there’s no non-flat frame at the end of a flat target, then SGP leaves the panel turned on. so i’ve had to put a bogus bias frame at the end of my flat sequence to make sure the panel gets turned off. in other words, turning the panel off does not seem to be an end-of-sequence task like (optionally) warming up or parking, etc., and the panel only turns off if there is a subsequent non-flat frame to run. i guess in a situation where you run flats and then run another target, it resolves itself. but if flats are the last thing that runs in the morning, the panel would be left on… i think.




Another vote for attention to flats.

My work around has been to disconnect the rotator from the sequencer immediately after the center and rotate feature has been used. Not ideal, but workable per night.

Incidentally, Rob opened a request for me, but nothing ever came of it last year. :frowning:

Also, SGP Rocks, so I never get too bent out of shape with the bugs. :slight_smile:


So it’s not only me having this anomaly! I have the same issue with the flat panel not turning on/off as expected. The sequence I used is lights/darks/flats/bias. That way the panel is off when it needs to be and on when it needs to be.



@hxpi I have the exact issue. I do the same as you, I take my bias after my flats because if the flats are last the panel does not turn off. I do what @AstroMatt does as well with disconnecting the rotator just to make sure it does not rotate any before flats are taken. Also, if I image the same target as the previous night, I disable the rotator in SGPro and I do not power on my rotator. If I power on my rotator is will rotate to home position at power up and now everything will be off a little when SGPro does it start of sequence centering.

Not a big issue for me and that’s why I’ve never brought it up before because I know Jared and Ken have other things that I’m sure is more pressing to fix and my workaround seems to work for me.

However, It would be nice if during a request to center, prior to a start of a sequence, Centering would also take into account the rotator without needing to wait for the actual sequence to start. I typically like to center on my target prior to kicking off the sequence to get everything running and tracking such as PHD2 and to test out my exposure time to see if I need to increase or decrease my times. However, it appears when I go and ask SGPRo to center, Centering does not take into account the rotator.

Again, not a big issue I just do my workaround. I’m still very happy with being able to get a good nights sleep while my rig is busy taking lots of images :grinning:



Ok, sorry to necrobump but +1 to improving the rotator flow. I use a flipflat. Here’s what I do, and I don’t like it:

Use the framing wizard to create a target(s)
Start the sequence
SGP points the scope, rotates the rotator, and plate solves. My nitecrawler is within 0.3 degrees every rotation… It’s awesome!
The first image is a 30 second throw away, because if you have a first photo as a flat the flip flat will close and the OAG will fail to settle, killing the sequence.
My next 4-7 frame types in the target will be flats. This takes a considerable amount of “science time” because I’m taking flats not light frames.
Once the flats are finished, I take light frames, then move on to the next object.

My location is bad in that I have a very small view of the sky. I have to image an object over several successive nights. There are objects in the south that I see just at the horizon for only an hour. Orion, for example, I get two hours a night if I’m lucky. So I have to image it every night. Technically I should be taking flats before every set of light frames, but that would chew up a dumb amount of time every night.

What I’d like to see is a workflow that is something like this:

A checkbox that says “Create flat frame object on successful plate solve”

After you plate solve successfully on an object it will create a separate flat frame object with the exact rotation angle, that can be taken after the evening’s imaging.

Sequence then completes, or clouds force the sequence to fail (I run unattended)

SGP closes the flip flat and parks the scope (PLEASE ADD A FEATURE TO PARK THE ROTATOR- MY filter wheel is insanely big (SX-MAX 11 position 1.25" filter wheel) and has hit my pier multiple times. Also the ultrastar OAG camera has a terribly designed connector on the end of it, and that can hit my pier as well. I’d like the scope to always park the rotator to a particular angle before it slews to park.

SGP takes flats (and biases or darks?) after rotating the camera to the correct angle, which is saved in the flats object after the original plate solve.

There is a checkbox: “Take flats every run” that will not require the user to reset the progress on every flats object every day. So flats will always run after the sequence ends/fails.

If the sun actually comes up during flat frame imaging, the sequence is aborted, but the remaining flats can be taken the next evening before you start imaging because you have the angle stored in each flats object.

Post sequence procedure then runs, and disconnects (in my case) all of my equipment.

That’s how I’d like flats to work.

The fact that SGP allows me to run my scope unattended is basically magic, IMHO. <3 this software and thank you MSS for making such an amazing product.


The other downside to the way I do things is there is some goofy thing about how SGP and the flip flat interact that can cause the flip flat to lose it’s mind. It literally has to be unplugged from the USB hub (remotely turning the hub on/off won’t fix it) in order to restore functionality. Next time this happens I’ll grab the logs and post, perhaps there’s something that can be done on the SGP side, or I can post a bug with Alnitak if that’s necessary.

My long term goal is fully automated remote control, and I’m soooooooo close.


Hello I was considering a Nitecrawler but the above seems to imply that there is no good procedure for using a rotator and taking flats when you run multiple targets in a night (except to waste imaging time by taking flats in sequence). Are there any planned changes to this issue ?
Thank you


AIUI if all the optical components that could affect flats are between the CCD and the rotator then rotation won’t affect flats. Those will be the CCD cover glass, the filters and any additional optics such as a focal reducer. Usually all these will rotate with the camera so rotation will not have any effect.