Cepheus, 12 panel mosaic

[don4l]

Click on image for full resolution.

After spending years trying to capture a mosaic of Cygnus,  I've now managed to do a similar job on Cepheus in 6 weeks, which is very pleasing.  I've still not managed to grab all the frames first time around, because I underestimated the amount of exposure that each frame would need, so I had to do some of it twice.

I did have a quick go at this a couple of years ago, but I really didn't understand the fact that the camera would need to be rotated for each East or West movement. (I'll show the non-rotated result below the main image)

This is 12 panels.  On average, each panel has 10 x 5m exposures, through a 3nm Ha filter.  Total about 10 hours. The camera is a G3 16200 on a FSQ106 at 387mm.   The field of view is about 11 x 11 degrees.

 

There are lots of Sharpless objects in Cepheus.  The next image is a screenshot from Cartes du Ciel which shows them.

 

 

Each frame was calibrated and stacked in CCDStack.  Everything else was done in Affinity apart from the cropping and scaling of the CdC screenshot.

Affinity uses non-destructive editing. This means that I can easily go back and make adjustments.  So, if anybody wants to make suggestions for improvements, please feel free.

Thanks for looking.

Edited September 27, 2020 by don4l

[Andy R]

Great picture, I like it very much. Very easy to loose yourself in the detail when zoomed in.

10 hours well spent👍

Thanks for sharing

[Ouroboros]

Yes, that’s a very nice image.  The camera needs rotating because .... effectively the sky is on the inside of ten celestial sphere? Is that right?  
How much overlap do you need to leave between images? 

And when generating a mosaic how do you ensure that the background sky looks seamless between adjacent images? 
 

 

[don4l]

33 minutes ago, Ouroboros said:

Yes, that’s a very nice image.  The camera needs rotating because .... effectively the sky is on the inside of ten celestial sphere? Is that right?  
How much overlap do you need to leave between images? 

And when generating a mosaic how do you ensure that the background sky looks seamless between adjacent images? 
 

 

On my first attempt at this, in 2018, I just moved the camera sideways until the right side of the frame was where the left side had been for the previous frame. This was the result:-

I not a mathematician, and I spent ages trying to figure out the effects of photographing the inside of a sphere.  In the end I decided that the answer was much simpler.  The camera is on a polar mount, and the whole thing rotates as you move East or West.  As you get nearer the Pole, the effect increases.

To fix the problem, I didn't resort to any fancy maths.  I have a two stage system.  I platesolve the first image and display it in Cartes du Ciel.  (This bit takes a bit of learning).  Then,  when I am lining up the next frame, I take a test image and platesolve it.  CCDCiel (My imaging SW) can then display the frame in Cartes du Ciel:- 

The red box is CdC's "CCD finder rectangle".  It has been placed in that position be CCDCiel's platesolver.  I can see that I need to rotate the camera a bit anti-clockwise before I take another short sub to platesolve again.  When the red box is lined up exactly as I want, I am ready to carry on.  This usually takes 2-3 minutes.   

The tricky bit is learning to process images so that they can be displayed in CdC.  They need to be platesolved FITS images, scaled for 0 - 65000 (16 bit) or 0-255 (8 bit).

Here is one to play with if anybody wants to try.  Just put it into CdC's pictures directory, and choose "Setup -> Pictures -> Image Archive" and go to the Elephant's Trunk.

 

 

CepB2.fit

[don4l]

1 hour ago, Ouroboros said:

 

And when generating a mosaic how do you ensure that the background sky looks seamless between adjacent images? 
 

 

I missed that question earlier, so here goes...

That's the really hard bit.  Each frame is in it's own layer and I adjust the contrast and brightness until they almost match.  Then I use gradients on the adjustments to apply them with different strengths across the frame.  This works most of the time.

When I can still see some of the joins I then delete "splodges" of the upper layer so that the offending difference is no longer a straight line.  I've done that twice in this image.  When it isn't a straight line it becomes very difficult to see.