Dealing with elongated stars at edges of field...

[ram]

Hi all, I was wondering if there was an easy magical process for dealing with elongated stars at the corners of the image. I have great stars in the middle but at the very edges, they get more elongated and misshapen. I can of course crop it out but I'd like to be able to capture the whole field in the target I'm doing right now and I was wondering if there was a way to make them round/nicely shaped throughout via processing. Thanks!

--Ram

[JayTee]

I see you already have an FF. That is what is needed to get round stars all the way to the edge. Maybe you need to invest in another FF. I have and use 2 and swap them out depending on which scope is imaging. Both of my FF's are only field flatteners with no reduction component built-in.

Cheers,
JT

[ram]

Thanks JT. I am already using a reducer/flattener combination for this scope - are you saying this is the problem? So I'd need a dedicated flattener on top of this combination? (Otherwise I have reducer/flattener combinations for all my scopes.)

--Ram

[Hankmeister3]

Ram, a field flattner is just that, a field flattner. Technically, a field flattner corrects for the Petzval field curvature of an optical train, especially as it relates to the main objective. But an FF doesn't address the issue of coma - a least most of the time FF doesn't. Field flattners, focal reducers and coma correctors are all different beasts. Baader makes an excellent series of coma correctors which, in my case, corrected about 90% of the "elongation" or "seagulling" which occurs as stars near the edge of the frame. I don't believe it's astigmatism which is causing elongation or distortions of stars as you near the edge of the frame either. Astigmatism would be an optical defect ground into your main objective but that would also affect the stars at the center or near center of your images.

It sounds and appears to be a classic case of coma. I ended up buying two Baader coma correctors (MPCC Mark III), both are designed for an f/4 fast Newtonian like the SkyWatcher Quattro and Orion series. However, because no two optical corrective devices are ground exactly the same, I found one f/4 coma corrector works best with my own f/4 Quattro 8 and the other f/4 coma corrector works the best with my Guan Sheng 8 f/5 and a smaller Orion 6-inch f/5 which I have but rarely use. There are more expensive coma correctors out there, some designed specifically for short focal-length refractors, but I've always felt the Baader CCs are the best buy dollar for dollar plus they work really well with the "working distance" of my two Canons. There is an optimal working distance between the rear face of the coma corrector glass and a camera's focal plane which should be observed, too, to really fine tune things.

The Cassies I own are fully corrected for coma and also offer a pretty flat field, so I found I didn't need to use anything other than a focal reducer from time to time if I didn't mind the light fall off toward the edge due to vignetting.

[DMouse]

Hi RAM, you didnt say which telescope your having the issue with? I do see on your equipment list that you have the TAK reducer. I have an FSQ106dx4 with the TAK reducer and have pinpoint stars to the edge with an ATIK383L+ camera. The key is spacing from the reducer to the camera chip. I bought a custom spacer from Precise Parts in Miami. They arent cheap but it saves messing around with combinations of adapters. So as long as you have the right reducer for the right scope then it is all down to correct spacing.
Hope you figure it out!!

Clear Skies

Paul

[ram]

Hank, you may be on to something, but just to be sure, here's a recent image (S2) of the Heart I've just started: http://ram.org/images/space/downloads/i ... v0_tmp.jpg - if you still think the corner stars are a coma issue I can invest in a coma corrector and try it out. I thought refractors and coma issues were minimal to nonexistent.

Paul, it's the SV70T but in general I have it with all my scopes to different degrees and cameras. Obviously the bigger the camera sensor and smaller the focal length, the more pronounced the effect is (I don't see it at all on the FS128 and the TOA/FS 130-R combo). Though I think this issue is worst with the Stellarvue reducer/flattener than with the Tak reducer/flatteners which as you note are more sensitive to spacing but the problem is also less pronounced with them and in the case of the FC100 and FC35RD I didn't have a microfocusser until recently which helped me alleviate this further.

You're right, it is fundamentally about focus and spacing I think unless it's a problem with coma as Hank notes, and I want to point out that the image above has a spacing of 60mm instead of the standard 55mm since my spacers were stuck and I couldn't unscrew them. What I've heard from Vic at Stellarvue and others also is that each reducer can be slightly different so I once was able to get pin point stars with something like 53-54 mm spacing rather than 55 mm spacing in my case but I've not been able to replicate it and furthermore that could be just be due to just reduction of the image frame. The camera itself is sensitive to +/- 0.5 mm of spacing so between that and if you add that to reducer tolerance you can get variations of +/- 1mm or more at least with the Stellarvue and QHY combinations.

The Tak reducers (I have two of them) have different spacing specs, so I think that's like 76mm but even there I had some issues so I bought one of those VariLock spacers and that helped me mitigate most of it and now with the microfocusser on the FC100DF (which was present in the FS128) it's not the pressing issue.

The VariLock I have goes from 20-29mm and I need about 16-17mm of spacing so I've ordered two 8mm spacers from Agena but it won't be here in time for tonight's clear skies (forecasted). The Stellarvue reducer/flattener actually requires 66mm but they provide a 11mm spacer along with the reducer which is nicely designed and robust (it's t-thread but like a 2" size so flexture is a non-issue with this) but I do have t-thread spacers going to 66mm exactly so I will try it out tonight and hope it works by removing their 11mm spacer. To get 66mm I had to use a 23+3mm spacer which gives me an effective spacing of 28mm (due to the length of the internal female thread which is only 1.4mm whereas the reducer's male thread is seemingly about 3.4mm) and hopefully it will be stable.

--Ram

PS: Whoever moved the thread, thanks a lot! Sorry for the posting in the wrong part of the forum - I actually wanted it in Image Processing but this is okay.

PPS: I'm still wondering if there are any software solutions to creating more rounded stars for the corners in like the image above. Since clear nights are few and precious, I can always spend time fixing and reshooting What this means if we want to avoid this issue or crop the image is throwing away that data which I would prefer not to do. Of course if one corrects the issue and shoots more frames correctly integrates with PI then hopefully PI will throw out the outliers which may happen anyway.

[ARock]

StarTools has a repair module that you can try.
https://forum.startools.org/viewtopic.php?f=12&t=1170

You might have to try it on each corner separately....

[ram]

Thanks ARock! I will check it out... I woulld prefer not to throw away my old frames but it may be unavoidable for this and another reason I'll post in the Image Processing thread (but perhaps that's fixable as well, or at least I've done something close to it before).

I did fix the problem or at least figured out what it was: definitely the spacing. Today as I mentioned, I got it down to the correct 66mm (actually) needed between the reducer and the sensor and the issue is far better than before, it's almost gone in fact. I still have about 10% (compared to before) but I must be off by 1mm or less one way or the other but I can live with it. (I could actually live with the old one too especially on most targets it can be cropped out but in this case this is a repeat of one of my very first targets which I did via live stacking and I've come a little further than that now and the Heart fills up the whole frame and I want to do at least my usual work on this: it would've all gone well if I hadn't had tried new combinations and caused the spacers to get stuck.)

Clouds did cut about half my usual 3.5 hour run I can get on this target so I am getting half that but hopefully I'll also get the comet later.

--Rma

[ram]

In case people are wondering how this turned out, I believe I have made decent progress, if not come close to solving it, with the following:

Problem: Changing mind and/or distances between sensor and reducer, resulting in framing change with partial data captured in one frame and partial data in the other. Don't want to throw out data.

1. Register the mismatched images with star alignment, register/union separate and frame adaptation checked which will generate two images of the frames.

2. Use gradient merge mosaic with "overlay" and this is what I got:

Now, there are still some pinched areas but they may be fixable by GMM by altering the parametres (I did this before for the Veil) or in this case since the two frames are so close to each other, use pixel math cut and paste the regions. But still I think this is reason for hope/optimism especially since in this case I'm combining S2 data with O3 data (I have only O3 and Ha in the new framing, whereas the S2 data is the one shifted to the right). Especially since the goal is to crop out the left part, if I can get the pinching to come on the left side (has to with order of the images) then it may work. And yes, that's the right thing, since the S2 data is on the bottom here and you're seeing the O3 data but really I'd be combining S2 with S2 or S2 is what I want to see which at 3.5 hours isn't as impressive as the O3 so I'll leave this image but it's basically the same thing - I want to crop at the tip of the ray shooting out of the heart on the top left.

The O3 and S2 data for the Heart are relatively low compared to Ha but in this case, it is 6.4 hours of O3 and 3.5h of S2. I have only two hours of Ha and it is stunning and I plan to collect more.

--Ram