wimvb

Regarding the baader cc and its distance to the sensor: When using the cc with my pentax dslr (cc + t2 + dslr), the focus tube went quite far into the light path. When I replaced the pentax with an ASI174 and spacer, the cc distance was initially 1.5 - 2 mm too far out. This affected focus remarkably, and the focus tube wasn't in the light path anymore. Because of the much smaller chip, I didn't have problems with coma. But when I corrected the distance, focus moved back in again. Numberwise: Cc + dslr, focus with tube at about 10 -12 mm from all the way in, no visible coma on aps-c Cc + ASI174 + too long spacer, focus almost all the way out (about 80 %), no visible coma on small chip Cc + ASI174 + better spacer, focus about midway of its travel, no visible coma on small chip. Since the baader cc is supposed to have a best distance of 55 mm ± 1 mm, and focus seems to change substantially with distance, you could just try adding a 1 mm spacer before taking out the hack saw.

Any unevenness like this suggests that there is tilt between the sensor and the focal plane. The SW crayford focuser can actually be collimated. There should be three pairs of screws around the focuser base. These can be used to align the focuser. Here's a video showing how to centre and align the focuser of a SW newtonian (about 25 min in) https://www.youtube.com/watch?v=3LbR1nIx-jw Personally I would do such a job either during the summer, between seasons, or during a cloudy weekend. Btw, my simulations were only for on axis/symmetrical obstructions. I will have a go at simulating off axis effects.

It may partly be a focus issue. When I did the simulations, I noticed that the vertical (unsplit) diffraction spike gets wider when defocused. Eventually it will split in two parallel spikes. The vertical spike looks a bit wider in the first image. So my conclusion is that it's slightly defocused as compared to the other images. Did you refocus between exposures? If not, focus seems to shift across the image plane. If it's not that, I really have no other clue.

I wonder, is it symmetrical around the centre? Are the spikes split to the other side, on the other side of the centre of the image? Just checked your original image. The right hand side looks normal. It's only in the left half of the image.

I've run out of ideas.

That would be the way to do it. Be sure to keep the scope collimated. That's more important than diffraction spikes.

Here's the simulation I mentioned before. An aperture with an obstruction that is slightly shifted upwards and 4 spidervanes. The vertical vanes are precisely vertical, while the "horizontal" vanes connect the 3 (9) o'clock position of the aperture to the 3 (9) o'clock position of the obstruction. The aperture: (Btw, the size of the obstruction shouldn't be a critical parameter here.) The diffraction pattern it produces at focus: And at a very sligth defocus (50 microns according to Maskulator, within the critical focus) (the horizontal spikes are slightly wider, and the vertical X starts to ever so slightly shift upward)

I have the same issue as you, but much less. I've been doing some light diffraction simulations, but can't replicate the proper diffraction pattern. Mainly because any diffraction pattern I get is symmetrical. If you look very closely at the brightest star in your image, you'll notice that opposite of the split spike, the other spike is also split, but much weaker. The closest I get to this is when I simulate a situation where the vanes are not at right angles, i.e. where the secondary mirror is not exactly centered. This gives a diffraction pattern similar to that of a Bahtinov mask. Defocused, the position where the spikes cross, moves away from the central bright spot. It may be worthwhile to test this. I hope this makes sense.

I found that the black point setting (background removal) is very critical. You may need to play around with various settings. Just a thought

At occasions like these I'm glad I use PixInsight. Seriously though, great write up. And thanks again for the PI script.

That's getting really good. To get a significant improvement, you basically need to double the total exposure time ("-ish"). But adding Ha is an entirely new dataset that can lift your image with only a couple of extra hours. And if you want to add Ha in the end any way, why not start now? I would probably go for the modded cam. You have hints of Ha as it is.

And when you get the 2" coma corrector, you replace that last adaptor with the cc. The cc also takes a light pollution filter in its front, should you need one. Btw, make sure you tighten the screws in your focuser when you use that adaptor. You don't want your camera to fall off.

Now, that's a depressing bed time story. Sorry it didn't work out for you. I hope you have better luck tonight. Monday night is supposed to be a clear night out here, according to the Clear Outside app. Probably the same gap in the clouds that's passing your neck of the woods this weekend. But at the same time it predicts 7% chance of rain on a cloudless night, as oppsed to 0% chance on a 100% clouded night. I wonder who programned that app.

Great write up. Glad I have PixInsight; only a few sliders and no layers. Thanks for the script, Mark.

To test this, try the following. Wrap something around the suspected vane and image a star. Repeat with each vane until you find the culprit. Focus on a bright star (like Deneb in Cygnus), and gradually defocus. Eventually you will see the shadows of the vanes. At that point you can hold a finger near each vane at a time and identify them in the view. Either of these methods should give you the position of the twisted vane. Then loosen the screw on the outside of the tube, twist the vane until the diffraction spike looks good, and tighten the screw again. If you tighten as before, this shouldn't throw collimation off. But check collimation of the secondary mirror afterwards anyway. Easiest way to check collimation is to focus on a bright star, then rock focus in and out. The defocused star should be symmetrical and look the same on inside and outside of focus.

Dion from astronomyshed has a youtube video showing how to align a sw focuser. Also, the standard focuser has three pairs of collimation screws on its baseplate. They (should) work the same as those for collimating the primary.

That's a great start. The vignetting is best taken care of by using flats, even if gradient removal tools also can do their share. I think that from here, the next logical step would be to add bias and flat frames. Dark frames with a dslr are debateable. Dithering (= moving the fov slightly between exposures) is more effective.

Great image with superb processing. And, as Olly said, unusual in nb. But it clearly works.

The first is my favourite too. You managed to bring out more of the faint Ha background. (Btw, the Ha background causes stars behind it to appear orange.) My first guess was that this is the image with more subs, since more subs = less noise, which means you can stretch more. Otoh, as the moon brightens the sky it also introduces more noise. So, in this case the better image may well be from the smaller stack.

You be the judge to that: I've never had such distinct star spikes before. Considering a 1.6 "/pixel image scale, I'm very satisfied. The geared motor does have a lot of backlash, which my arduino code could compensate for. But the moonlite driver has no backlash compensation. Still, the autofocus routine in Ekos handles this fine.

It works good too. I autofocus with ekos/kstars from my livingroom. No need to go outside to refocus.

Nice to have you back here, Gina.

That's a very well thought out design for your obsy. I like the idea of the main roof sliding over the smaller roof. Foundations are the most important to get right, so take your time for this. Once the floor is in place, you'll find that the next phase will go a lot faster. Good luck with the build.

You folks in the tropics.

Nice! I think that polar alignment should only be out a tiny bit. Too much will mean the guiding will have to work harder. When I used Lin_guider, I let the mount drift for a while until I saw in which direction the DEC would guide. Then I switched on guiding in that direction only. This would quiet down the dec guiding, almost independent of how good PA was. Sometimes I find Lin_guider to be more Push Here Dummy than PHD. (Also note the 'Accumulate frames' settings which let you average corrections over a number of frames before applying.)