Auto centering on camera with overscan

I have a QHY-12 with overscan enabled to calibrate out bias drift. This means that the picture area has a column about 200 pixels wide on the right side that is not useable for the image; this is the overscan area. The problem is that the auto-center routine does not know this, and centers the scope to point at what it thinks is the center of the image when in fact it needs to be offset by about 100 pixels to the left. This becomes a problem after a meridian flip and the scope centers again, this time offset to the opposite side. The net effect is when the images are registered, I end up losing about 200 pixels of the frame because of this offset.

Is there a way to tell SGP’s autocenter routine to use and offset?

@gcardona SGPro cannot infer overscan areas. We could allow an option to allow a user to specify this area, but capture software, ultimately should not be used to account for hardware idiosyncrasies. The driver should take care of this for you.

I don’t need SGP to infer the overscan area, I would just like to be able to specify where the center of the image is when I enable overscan. The driver is being set up to show the overscan area so I can use this data later in calibration.

gcardona, were you able to find workaround for this?

I have the same request. I have QHY695A camera and I also need to leave overscan area in my frames to calibrate bias drift. I need to be able to specify actual center of useful image area on my frame so I will not loose signal when I do meridian flips. basically exactly same reason as described above by gcardona.

please add this feature - it would help to have an option in framing & mosaic wizard to specify center of the frame, which then should be used to center scope on the object.

to illustrate problem, this is how my frame looks like:

black area is overscan area, green area is useful signal area

then after meridian flip and stacking/aligning images - I end up with something like this

parts of data on the edges is unusable because it exists only on certain set of frames.

my full frame with overscan area is 3072x2240 , actual image is 2752x2160
so after I align and stack I loose 320px horizontally and 80px vertically.
so instead of my full camera resolution of 2752x2160; after I align and stack I end up with images of around 2400x2080. it would be really great to get a fix for that

i believe this is relatively easy to implement, one way I can think of

  1. add optional parameters to define effective imaging area coordinates relative to full frame in the form : top x left, bottom x right. think this should be on camera tab of equipment profile/sequence control panel.
  2. then with simple math it should be possible to determine actual center of effective imaging area in relation to full frame and use that point to center image
  3. then it would be nice if when you define your frame in “framing and mosaic wizard” the rectangles would represent effective imaging area of sensor if thats defined on camera tab, or just entire sensor size if effective imaging area settings left blank.

also this is not some kind of device specific request - I believe many cameras have overscan regions and quite many of them experience bias drift, so I believe this will be valued addition to SGP


I have the exact same problem when using my newly purchased QHY16200.

I believe the overscan feature is useful to correct bias shifts between bias, darks, flats and lights.

I am quite annoyed to loose 20% of my image when SGP do an auto meridian flip. Furthermore, I am quite surprised that SGP does not support CCD cameras with an overscan region (this seems to be a long last standing issue between SGP and its interested customers). As axlns states above, it does not seem that difficult to implement the definition of the imaging area within the acquired frame to make sure SGP centers the plate-solved imaged area and not simply the CCD frame…

I am a SGP happy user for 3 years now but was a little disppointed when Main Sequence Software decided to move toward a $55/year licence business strategy last Fall. If we are not able to obtain this support with this new yearly fee, I can predict I will look into the direction of NINA within the next few months…

An frustrated SGP user who just invested $30k in a new observatory.