Win a Pentax Limited Lens: Raffle Now Live!

I haven't been very happy with my post processing and I wondered if someone could advise me how I can do better.

I've finally found a clear night and got my Vixen Polarie reasonably aligned to give me round stars (if not points). I processed the frames in Nebulosity and a little back end adjustment in Lightroom. I was wondering if someone could show me some better results. I posted the original stacked FIT image (EDIT: no processing) and my pp result in jpg in dropbox here: Dropbox - Orion_2016_01_08

Done with K-5II, SMC Pentax 200mm F4, 41 images at ISO 800, 10s. No flats, no bias. I used Nebulosity synthetic flatten which worked nicely. My flow was demosaic, trans and rotation align, stack, colour correct, adjust colour offset, synth flatten. Then in lightroom a little stretching and clarity.

The images are low in noise but I don't get much nebulosity to show. Note that I took these images from a city balcony with an almost full moon only 15 degrees away However, it was a relatively clear cold night at -17 deg.C. So another question... do I mostly just need darker skies?

Thanks

Hello gbeaton,


Two observations:


1. It would probably be better if you uploaded an unprocessed stack (before white balancing, clarity etc.). Any "wrong-doings" in these steps are hard to reverse even with your 79 MB FIT file.


2. Your exposure times are very low. Surely, the Vixen Polarie can do better than that with a 200mm lens. Remember, high ISO doesn't help you catch more photons and while the core on the Orion Nebula is bright, the outer parts are quite faint.


Now, I don't have (yet) any software to cope with FIT files,but I am currently playing with a demo version of StarTools and managed to get this screen dump of the resulting enhancement of the central part of your image in a few steps (including a finishing touch with white balance and curves in PhotoImpact X3):



So, you actually did capture quite some nebulosity.


I have a similar tracking platform to yours - an iOptron SkyTracker ver. 2 and I allow myself to show a very preliminary result with a Pentax Q10 on a small APO 474mm f/6.78 astrograph taken in the midst of my city under severely light polluted and turbulent skies. Individual exposures are 30s at f/6.78 and ISO1600:



For details, look at my post in: https://www.pentaxforums.com/forums/136-pentax-q/172196-lets-share-shots-q-355.html#post3877112


So, no you do not necessarily mostly need darker skies. A simple City Light Suppression (CLS) filter might help you increase contrast (I forgot to mention that I tried just that for the above picture and did find an improvement), but first of all I'd recommend you to try increase your exposure times.

Thanks for the help. I thought the fits I posted was unprocessed - just stacked. I'll have to check back later as I'm on an iPhone right now.

Yes, the file I put on Dropbox is an unprocessed original stacked image.

Although its close, I can see you have achieved a little more definition in the nebulosity compared to mine - most notably in the outer reaches. Could you let me know what edits you did?

I think the iOptron is easier to align - from a review I read. I have some 20s captures but my alignment gave a little trailing. But I'll definitely keep trying to do better alignment and longer captures.

Something that really surprised me was that even though the Polarie was North Eq aligned, I had to do rotational alignment! Could that be lens aberration issue?

Hi, I'm travelling till Thursday and just see your image processed by Stone G. only on an Ipad. But there it looks like you have mainly also a focus issue.This 2 year old M42 was made with a 70-200 F2.8 mm Sigma, ISO 1600. 123x30s at ISO 1600. It was shot at F6.3 so it should be not too far away from yours that was shot at 4.0.

I would recommend to tinker a Bahtinov mask for focus, maximize the exposure time and the number og stacked Images. When I am back I would like to give your data a try in processing.
Cheers, Pete

Pete, i am not too sure that the fat stars (like the ones in my own image too) are focusing issues as such. To me, they look as much as the result of very bad seeing. In my case, I used a Bhatinov mask and yet I got those poor star images - in fact so poor that Deep Sky Stacker stubbornly refuses to recognize them as stars.



Focusing my Pentax Q10-TSED70Q system on Rigel using a Bhatinov Mask. 30s exposure at ISO 1600.

gbeaton, I suspected a processed file because when I try to develop your image in StarTools, I get a strog green colour cast in your background sky. What were your observing conditions?

The steps that I used in StarTools were simple:

1. Develop - using home in
2. HDR - using optimize
3. Contrast - auto
4. Colour - using the sliders for colour bias suppression
5. Final noise reduction


Finally, I take note that Pete_XL's example is also based upon 30s exposures.

Hi,
your biggest problem is the full moon. There is just no way (short of narrow band filters like Ha) to capture DSOs. To highlight: years ago I took a picture of the full moon at 10 o'clock on a winter's night - 10 sec exposure, ISO1600, f/4, 32mm focal length. This is the result:

You wouldn't think it's dark night

Few weeks ago I took pictures of Orion with no moon around - 182.6 sec exposure (a cheap intervalometer's idea of 3 minutes) ISO 800, f/7.5, 600mm focal length and this is a single frame:


Personally, I don't even think of taking the telescope out when the moon is more than half full. You need more exposure time, which you can only get at a darker night. Final image of Orion then looked like this:



In just over a weeks time we are back in business and here the moon will rise after 1 a.m. Plenty of time to get some shots in.

Ckear skies!
HJ

Hey Stone G., sorry for beeing so unprecise. My comment did not refer to your Q10 image. I meant the data of gbeatons picture which shows very smooth big stars and a lack of internal structure. I would not expect this from his setup, that is why I suspected a weak focus. I own the same lens and it is quite sharp on stars though it produces a bit of red fringing.

Sorry once more
Pete

Thanks Pete, I just purchased an old SMC K 200mm (that I'm really happy with at $80) and I just turned the focus to the stops. Read I could do this. I'll try again with a proper focus.

Hi Stone, that's another issue that I thought I could ignore. I get the green cast after de-mosaic and I though I would correct it later in the stacked image. Otherwise I could use 0.5 on my green channel but I figure I would lose info. I have a Sony 16.3 Mp sensor. Should be quite standard so I use the default colour channel settings for de-mosaic.

Whoa! really?!?! that looks like a cloudy afternoon! I have to admit. I have so little good nights here that I decided to do some AP at any clear night. At least I'll get some practice aligning and such until those perfect conditions come along - probably after the earth has warmed up

Wow! this is the odd thing where some great images are done with sometimes much less exposure time. I guess its like you say, no moon, and probably clear, dark skies!

Thanks guys, I'll definitely work on better alignment for longer exposures, focus and hopefully the weather will co-operate too.

Please spare a bit of that co-operation for me too - I need it desparately!


Good Luck!

Hello gbeaton,

I played a bit with your data in Photoshop but did not have much time.
The original looked like this in Photoshop:

Testing data from gbeaton by pete_xl, auf Flickr
It shows heavy vignetting and I wonder where the green tint comes from.

I processed it a bit but lost nearly all color while neutralizing the green background. And while stretching it to the maximum content of the data a blue circle appeared in the southeast of M42 (it is more obvious in the uncropped version). I think it is a big lens flare from the moonlight - had that myself once.

for all pixels click image or look here: Testing data from gbeaton by pete_xl, auf Flickr

There definitely is some color and structure in the data, But the flare and the green color kill al lot of it and limit the processing.
Without color the structure can even be processed better because the colors are quite blown out.

for all pixels click image or look here:stk_recon_ISO800_10s SW.jpg by pete_xl, auf Flickr

You could perhaps use the b&w as a luminance layer for the color layer but as others mentioned above perhaps its worth more to go out under better conditions with an optimized setup for camera and lens.
Cheers
Pete

...
Wow! this is the odd thing where some great images are done with sometimes much less exposure time. I guess its like you say, no moon, and probably clear, dark skies!...

[/quote]



In astrophotography the general rule for exposure time is: more is better. The upper limit, ideally, should be determined by the darkness of the background. If you look at the level diagram of a single frame the peak should roughly occupy the left third of the diagram. With DSLRs you have an additional handicap: during long exposures the sensor is heating up and generates artefacts commonly called "warm pixels". There are various ways to work around them (dark frames, dithering etc) but from experience I usually stay at or below 5 minutes. I've tried 10 and even 15min before but everything becomes a mess. In comparison, people using cooled CCDs are able to take 30min frames.


I think everybody who starts with AP also starts with short exposure times, I did. Polar alignment is not as crucial and it is generally easier to get round stars. With each session my exposure time seemed to creep up until I reached the limit of my camera (K-30). I also should mention I live at a fairly dark area. Good conditions are, when you can see the dark clouds in the milky way with the naked eye

I added one of the original DNG files to the dropbox. All shots show a lot of red cast as can be seen in the DNG. I'll have to do some experiments to get the de-mosaic right. I think the vignetting must be the old lens, even though its rated 9.2.

Cool, I had no idea. I hope your right. I have a lens hood which I just repaired that may help.

I'm hoping our cold temperatures (-15 to -20 C at night) may help me here. There has to be some benefit for living with cold!

Thanks guys for helping. A lot of good stuff to go on.

As suggested, my RAW to colour conversion was way off. I don't know how to get the right settings in Nebulosity so I just exported TIFs from Lightroom first and then did the stacking. I added the FIT and some pp results to the dropbox folder - marked with "lrx". There is a considerable difference. I don't really see what looked like flare anymore but there is some non-flatness.

I thought the colour correction and background offset correction worked well in Nebulosity so I didn't try to remove them in Lightroom before exporting. What do you think? Should I do some pre-processing in LR?

I can't say anything to Nebulosity as I use the DeepSkyStacker for stacking and mainly Photoshop afterwards. Usually I equalize and normalize the white point in lightroom in each frame bevor stacking. Often the sky color changes during the the session and DeepSkyStacker sometimes fails if I do not help it. I leave all other settings in LR unchanged and set the already preset sharpening to 0 before I export for stacking. All testing with pre-processing in LR did not result in better stacking results - most were worse.

The big flare I mentioned becomes visible only with maximum stretching in the end of the processing. I can't access your Dropbox from here. If the blue circle is not present in your results you have not stretched it to the maximum extent. I think you should also see it when you apply a maximum contrast in LR by shifting the brighness ruler to the right and then the black one to the left in several consecutive steps.

You conclude in the earlier post above that the vignetting could be an effect of the old lens. All lenses have some amount of vignetting. The 200/4 is not much worse than others. But in AP when stretching the dark part of the histogram to separate background from nebulosity the vignetting effect is multiplied and becomes limiting factor for the processing. I would recommend to use flat frames to correct the image. As you use lenses and not a telescope it should be sufficient to produce and use one master flat for each lens. You can use them in every stacking process to get rid of the vignettes. You can use the "T-Shirt-Method" with two stretched white shirts over the lens that points to the cloudy sky to get the flat data. Using flats is an essential standard in AP. If there remains still some vignetting it can be removed together with other (color) gradients in the beginning of the post processing.