3.58
#...
·
·
1
like
|
|---|
|
I just got my new asiair plus - and I am stunned by quality, stability and convenience! It offfers quite good stacking facilities on board, so you can stack DSO from images series and Moon from video (haven’t tried Sun yet) right on the asiair. So I was wondering, whether the quality was comparable to the classic suite of DSS and Autostackert. My findings are, that AsiAir is really good, but the calssics still do a slightly better job. See yourself in my 2 images I did for testing out - the DSS/registax versions are the final revisions of each. - Heart Nebula: https://astrob.in/ypfp27/B/ - Moon Slice: https://astrob.in/nlecli/B/ Questions: - So has anyone made similar experiences (or did I create flaws in my process)? - Does anyone know what kind of stacking is under the hood on the AsiAir? - Are there tweaks to improve AsiAir stacking? |
0.00
#...
·
|
|---|
| I was wondering the same thing. So I also want to know a bit more! |
4.26
#...
·
·
3
likes
|
|---|
|
Stacking registers the stars and can eliminate a little noise if statistical distribution mode is used. The method itself does not finish an image—stretching the hell out of the darker pixels does that. In other words, it’s the next stage, the processing, that brings out the final image. While automated systems can make the image viewable, the current state of tech benefits from human intervention. So I suspect ZWO is attempting to make stacked images for confidence and quick sharing on social media. I often live-stack in ASIair and subsequently process that image. There is usually no difference between doing that and stacking from scratch, though there are exceptions. Most of the images on my site started out as live stacks. also, counter to trend, I use Photoshop exclusively for DSO. Here is Heart and soul that starts out as a live-stack and was the processed in PS. https://www.astrobin.com/0aioxp/B/?nc=&nce= |
|
3.58
#...
·
|
|---|
|
Great insight, Dave - thank you! I agree that the post processing is where the real magic happens. I am just not confident enough yet, whether its the input stacked image or what I do with it - thats what I asked Suspicion is, that its my end that can further improve (as not every session delivers the same result on the same matrial) - so good to know the stacking outputs are similar :-) |
|
4.26
#...
·
|
|---|
|
Man, you got that right. I like to stack right after the session because I’m anxious for the reward. But I often over-do everything. A saner head the next day usually produces a better result. Heck, some images have been done 3-4 times…just so I can improve on some little thing. yeah, I suspect the stack itself is just fine. I bet the ASIAIR does not use standard deviation in its routine, however. This is useful for concealing hot pixels and can occasionally improve round stars, otherwise, I’ve gotten nice results with its “premade” stacks. good luck on your own processing efforts. It’s a particularly creative part of the hobby that ican be immensely satisfying. |
5.77
#...
·
·
2
likes
|
|---|
|
Hi Arnie, Thanks for your post - I'm very pleased to see another ASIAir owner getting great results  I've used the ASIAIR live stacking a few times, and it works very well. The main benefit is that it visualises the image capture process, so you can see your target 'emerge' from the background noise during the course of an imaging session. The main downside is that you have no control at all over the stacking process, which is entirely non-selective. When stacking using other applications such as SiriL or DSS (personally I use Affinity which is powerful and very easy to use), you can select individual frames based on image quality, making it easy to reject poor images such as those containing aircraft trails or cloud cover. Other apps also let you choose the stacking method - mean / median / sigma clipping - which helps enormously when it comes to eliminating artefacts such as satellite trails, without sacrificing what might otherwise be a good image. So for example, if cloud cover affects 50% of your images in Live Stacking, the cloudy images pollute the stack, and the quality cannot be recovered. On the other hand, if you capture individual frames and stack them later, you can inspect every frame and discard the cloudy ones, and stack only the good ones, yielding a much better result. Live stacking will never be as good as selective stacking. Of course, ASIAir lets you do both if you select 'keep all frames'. Just to clarify a few points about stacking ... stacking combines several related but separate processes :- 1. Registration Registration aligns images, usually by offsetting and/or rotating individual frames, so they all line up exactly. This helps a lot if your tracking/guiding are a bit out, for example after a meridian flip. There is no value in stacking frames that aren't well registered. 2. Compensation If you use compensation frames, this process removes hot/dead pixels, background bias signal and dust. ASIAir lets you specify compensation frames for this purpose. Master Frames, like light frames, also need to be stacked, so I find it a bit tedious shifting master frames from a server to the ASIAir. 3. Summation Essentially, summation inspects each corresponding pixel in each image in the stack, eliminates the outliers, and sums the rest. The idea here is that summation averages out noise, which is random, and reinforces signal, which is not. The result is a significant improvement in the signal/noise ratio - this is the main purpose of stacking. As a side note ... Signal improves linearly with the number of subs, but noise is reduced according the the square root of the numbers of subs. So - if you take 25 subs, the S/N is improved by a factor 5, if you take 100, by a factor of 10. There is a diminishing return on the number of subs, ~ 50 being the sweet spot, because of the emergence of read noise with larger numbers, not to mention the extra processing time. However, as Dave rightly points out - the next stage is where automation ends and creativity begins. As with many technical endeavours, the thing that usually requires the most work is BCK - between chair and keyboard |
1.20
#...
·
·
1
like
|
|---|
| Live stack does not do meridian flips nor does it selectively stack (since not all frames have been seen yet). Live stack not being selective does diminish quality since subpar frames get included. I use live stack to see the inflight progress of the stack roughly. The subframes are preprocessed in PixInsight script WBPP after the fact. Auto run mode can be used instead of live stack mode since the live stack is really not used at all. Hope this helps. |
|
5.77
#...
·
·
1
like
|
|---|
|
Ive never actually tried using the ASIAir to do a live stack with a target crossing the meridian - but I would have expected it to manage the MF in such a case - I might try it next time I do know that other apps (eg Affinity) manage this seamlessly and elegantly - after analysing all the images, they simply rotate the post-flip images to match the pre-flip ones. |
|
3.58
#...
·
|
|---|
|
Thanks for your insights, guys. As you refer to AsiAir Live stacking - is there any difference between the life stacking and the post-acquistion DSO stacking in the main menu? PS: I will definitely try out Affinity now - thanks for your recommendations! |
|
5.77
#...
·
|
|---|
|
Hi Arny, I've never used the DSO stacking feature in ASIAir, but fas as I can tell, it is exactly the same as the live stacking feature. I encourage you (and everyone on AstyroBin) to try Affinity Photo - here is a link for you. It has a robust set of Astrophotography stacking tools, as well as a great set of free macros that you can download here, designed by Affinity Experts specifically for astrophotography. It is also a very powerful image editor - so you can stack your images and process them in one app. It's more-or-less a direct replacement for Photoshop at a fraction of the price, and you own the license - no annual license fees ! It's cross-platform (Mac, iOS, Windows), and has a 30-day free trial. Happy to help if you have any questions about it |
|
3.58
#...
·
|
|---|
Gary JONES: I tried tonuse live stacking, but AsiAir refused to stack without Bias and Dark frames, but also declined my master darks and bias from day before at same settings. how do I get bias/darks into live stacking, ie where do I take or place them? |
|
5.77
#...
·
·
1
like
|
|---|
|
That's interesting ... You should be able to generate stacks without compensation frames - I've created many stacks on ASIAir without them. ASIAir can generate compensation frames - which are then stored in folders on the internal SSD or USB stick. When you do live stacking, you need to tell ASIAir where these files are. So, you can use any bias/dark/flat frames if you just place them in the corresponding folders, provided they are FITS files and the same resolution as your camera. Tell ASIAir where these files are located on your USB stick, or copy them to the internal SSD - which will probably allow the stack to be processed more quickly. I hope that helps - perhaps you can share more details of your workflow if this doesnt work for you. Cheers, Gary |
0.00
#...
·
|
|---|
Arny: I use the Asiair strictly for automation. Stacking is done in post processing, in my case I use Pixinsight Weighted Batch |
8.77
#...
·
|
|---|
| I use Air Plus for polar alignment, focusing, guiding, imaging and downloading data but prefer to do all of my pre-processing (calibration, registration, stacking, etc.) in Pixinsight and all my post-processing in PI and an app called ACD Photo Studio Ultimate. I shot subs last month of M42 and they were full of satellite streaks. Pixinsight got rid of them beautifully. PI has so many more pre-processing options if a sub is problematic in some way that I am taking the time to learn it as best I can. ASIAIR Plus is ingenious and I love it when I am out imaging, but I think for stacking it's pretty elementary. |
|
5.77
#...
·
|
|---|
|
I tend to agree with Jerry ... ASIAir is great for taking pics, but I would never use it for processing pics - other than perhaps live stacking just to validate that images are being captured correctly. I've tried PixInSight, APP, SiriL, and many other apps for stacking and post-processing. IMHO the best app for anyone learning to stack images is definitely Affinity Photo - its super-easy and has most of the controls you need. After that, I'd try SiriL. Although the user interface is a bit clunky, it is free, and it has great tools for selecting images prior to stacking. |
|
3.58
#...
·
|
|---|
Jerry Gerber: I agree, Pixinsight is stunning, yet quite complex, but beautifully logical and very well executed. Just tutoring myself through to get reliable results of a basic worflow. Any recommendations for workflows or tutorials to get better at it? |
|
3.58
#...
·
|
|---|
Gary JONES: Followed your advice to try out affinity, Gary. Really cool tool and very powerful and unlike Phtotoshop very well supporting astrophotography, especially to allow the full workflow from stacking to finalizing the image. But at the end I found it too close to photoshop by means of aiming to drive visual outcome effects. So as of now I seem to swing more towards PixInsight ... |
|
5.77
#...
·
|
|---|
|
Hi Arne, I'm very pleased to see you tried Affinity I'd encourage you to persist with it though ... I also tried PixInsight - as well as SiriL and APP and others - but found I spent much more time struggling with the software than I did creating images. I fell into the trap of trying to perfect things from a technical point of view, at the expense of spending more time being creative. The thing I like about Affinity is the nice balance it strikes between power and ease-of-use. I've stacked many images using Affinity, SiriL, PixInSight and APP, and really found very little difference between them. The only feature I'd like to see in Affinity is more control over image selection. The current method selects the best x% of images based on analysing the brightest 40 stars for definition and roundness, but that means you throw out 1-X% of images, regardless of how good they are. Id like to see a method based on FWHM - I put that to the Affinity development team and they are considering it. I get around this shortfall my using SiriL to analyse my stack first, then select images based on FWHM < 3, and I stack those. The other thing I like about Affinity is the end-to-end workflow - registration, stacking, astro-specific filters & macros, followed by image optimisation - it's all very easy - and inexpensive |
