Mosaic Math help please

Not quite sure what you mean with "have a look at the math". Using an iOpton SkyGuider myself I'm aware that it can be quite tricky to accurately align at the target - in particular for the declination axis where you have only manual adjustments. Beeing off by 8 arcmins in the declination axis is not unusual for a tracker. While I don't do mosaics yet, I perform imaging over various sessions and <= 7 arcmin in dec is so far my accuracy in reproducing the position between different sessions. Your RA discrepancy of 1m is not far away from my tracker experience of <= 40s.


I'm actually puzzled about your target coordinates. For the Vest Veil nebula - which you missed - pointing at 52 Cyg - which is just in the upper left edge of your image - shows RA 20h 46m 36s / +30°47'58". For a Mosaic with a tracker I would point at a DEC of 31° for both images and then just move along the ra axis for the 2nd image. Using 52 Cyg as a pointer star may help.RA 20h 53m 38"    DEC +29° 50' 22" just look like wront target coordinates to me.

Clear skies
Wolfgang

Mike Butorac:

Not quite sure what you mean with "have a look at the math".

Thanks Wolfgang - by checking the math, I'm wondering if by looking at the size, radius, pixel scale, and telescope details, you can determine how much error you can permit as you move from panel to panel.

I forgot to mention that the camera was supposed to be rotated 60° but it looks like near 66°. The target was the entire veil area so the center of the mosaic was in between the 2 major veil areas. Telescopius allowed for a 10% overlap. 

I made another attempt but kept each panel in line so RA was supposed to be equal. When I checked before the 2nd panel was shot, RA was a bit more than 1 arc minute off.  I reset it before I started shooting it.

Hi Mike,

Astrometry data tells you most of the information you need. Each of your images covers 3.27x2.23 degrees. In case you're orienting the images along the RA and DEC axis this means your images cover 3.27 degrees along the declination axis and 2.23 degrees along the right ascension axis. With such an orientation the west and east veil fit into a box of approx 1.4x2.4 degrees (DECxRA), so you could image them in one shot by aligning the long side of your image with the RA axis and you would have about 0.4 degrees of wiggle room on each side. But I fully understand that this is a test. By using 2 shots with an overlap of 10% as recommended by Andrea you get an image field of 3.27x4.24 degrees (DECxRA). So you can afford about 0.9 degrees discrepancy both in DEC and RA on each side. 

Please note: This calculation changes for different rotation angles but at least it serves as an orientation.

So as a conclusion in your first attempt you committed 3 "mistakes":
1) By centering the east veil nebula in the first image you're losing tolerance on the upper side of the 2nd image. Arrange the east veil nebula a bit lower in the image.
2) This effect is enhanced by choosing an excessive overlap of about 30%
3) Wrong coordinates both in RA and DEC for the second image increased the issue.
Due to your choice of 60 degrees rotation your "wiggle room" was somewhat smaller than in my calculation above. This is not a mistake but just personal preference. As a consequence of all four effects the west veil was cut off.

1 arcmin is 1/60th of a degree, so don't worry too much about such deviations. Using a tracker without an automated goto it's within the tolerances you have to live with. If this was due to drift you may still try to improve your polar alignment skills over time.

Have fun and clear skies
Wolfgang

You should have had the long side parallel to the RA great circle and shifted in Dec.

@Mike Butorac Plate solving really helped me with my mosaics: I use Stellarium and enter my sensor and scope details, frame up how I want the mosaic manually, record the coordinates, and then use ASTAP to do plate solving to make sure everything is lined up accurately when in the field. I've been using the same tracker as you (star adventurer) and with repeat manual adjustments and plate solving, I can usually get within a few minutes of the intended target. Using this technique, I didn't need to keep track of which edge of the sensor is parallel to which axis, as ASTAP would also tell me the rotation angle, so I could get it to match with Stellarium exactly (taking advantage of the mosaic rotator that's stock with the WO RedCat 51).

This was the technique I used to make a ~356 Megapixel "work in progress" mosaic seen here (EasyZoom.com).

This is what Andrea means:


The vertical lines are the RA circles - along each line the RA value is constant. The horicontal lines are the DEC circles with constant DEC values along the line. You can then take two images shifted along the DEC axis.

I actually suggested the opposite: adjust the long side along the DEC circles and move in RA. I find this preferable with a tracker which is motorized only in RA. But maybe there is a reason to do it otherwise.

 
Here the long side of the  image sensor is aligned with the DEC axis and you can shift along the RA axis.

Clear skies
Wolfgang

Mike Butorac:
Thanks Wolfgang! But how does this translate to my physical camera alignment?

Given your images above you have to rotate your camera by 90 degrees to get aligned with the declination lines. 
Alternatively you keep your camera rotation as ist is and follow Andrea's advice. 

Clear skies
Wolfgang