This "fragmentation" that you're seeing is known as posterization and is normally caused by stretching the histogram too far.
Do you know if DSS is exporting an 8-bit or a 16-bit image after stacking? If it's only 8-bit, try exporting as 16-bit, and it might solve the issue.
Also, give it a go without the bias frames... it'll be interesting to see if this resolves the issue.
Its all 16 bit Tif file output, Mike. And I am stacking without either dark or bias frames...the whole point of the experiment! Remember, I am comparing one stacking without darks and bias, and one stacking with both .
But stretching too far is likely to be the problem indeed but the reason I am stretching it a bit far is because when DSS stacks the uncalibrated frames, the data is shifted extremely to the right as per my screen shots in the above posts.
How should I stretch such a histogram without fragmenting it if it weight heavily to the right of the histogram? When stacking with the dark frames (and bias frames, yet to trial a stacking without these), the histogram is nicely weighted on the left requiring little levelling out.
I'll have another go with the level stretching, I'll do multiple small increments until I see fragmentation but the whole thing is just too bright if I dont bring down the near completely washed out blacks...I'll keep you updated
So just to remind you, I get histograms such as this when stacking with no calibration frames, heavily weighted to the right:
But with calibration frames, I get a nice initial histogram like this:
And this is the case for all the settings I have tried. All output as I said, are 16 bit tif.
Its puzzling. I have worked through this on two different pcs too.
Phew this sure was a brainbuster. I discovered a setting in the RAW/FITS DDP setting, a checkbox called Auto white balance. it wasnt checked. Now logic tells me I should have that checked for all my stackings in the first place.
SO for the umpteenth time I restack the uncalibrated frames...yuck the same washed out stack! However, I levelled out the RGB level IN DSS. Then I exported to Photoshop. There is a slight fragmentation but not as bad as before, so proceeded to do small increments of levels and curves until signs of further fragmenting occured (or posturisation, thanks Mike) where upon I would of course undo the last levelling.
So here is the same Andormeda image at ISO 6400 20mins total frame exposure stacked and processed WITHOUT any calibration frames...and its bloody better than my original image!
So Mike, you sure have a great case. We dont need the bloody calibration frames!...well at least as long as the original captures are reasonably clean which in this case I think they are and remember this was taken at dark skies not in light polluted environment. Perhaps now I should do a test on my London deep space images and compare.
Anyway, here is the uncalibrated stacked and processed version. Woo. Click for higher res version. I didnt upload this one to Flikr as I dont want to swamp it with similar images But I uploaded it to another useful astro specific gallery site which I think is brilliant because you can search for images based on equipment used etc. Its called Astrobin.
Edit: after clicking for higher res version, you can then click on the top right corner and view the even highest resolution file which will then give you an idea of the noise level.
I'll post up a zoomed up section so we can compare the noise level of the calibrated and uncalibrated images next.
...and its bloody better than my original image!
Yep... agree with that... better definition in the dust lanes and less noise. The case for the prosecution rests m'lud π
Perhaps now I should do a test on my London deep space images and compare
Might be a good idea. As you have to stretch images in light-polluted skies further, darks will come into play more. But it would be interesting to see how much effect they have.
Mike you are our hero. Whenever I go to dark skies in future, I am not going to waste those precious dark skies on flipping dark frames. Man, if I knew I would have spent more time on Pleiades as I really wanted to get some nebulosity out of that.
Sumitra, Test done π For dark skies imaging, Mike's hypothesis is a hands down winner. Mind you, a tad more work is need in the post processing but that's a really small price to pay when gaining more quality time with the clear skies.
Will commence the London light polluted images tests tomorrow π
However, there are two inexpensive and imho essential items that I suggest you purchase to bring out the best in your telescope for both visual and imaging of planets and deep space.
The first is a bahtinov mask for achieving ultimate focus...
Again, thanks Tej! I'm going to look up Bahtinov masks tonight - another gadget I'd not heard of! I like the fact that I can use it for both visual and imaging purposes. I do like a multi-purpose tool! I can definitely relate to the feeling of uncertainty of being in focus when there's lots of turbulence. You find your target, try to focus, question your eyesight, earth rotates, repeat (I don't have tracking ability).
That's it! Adding to my shopping list, a motor drive for my EQ2 mount!!! And now a whole wealth of questions regarding motors has opened up in my mind but that's another subject matterβ¦
The second item, I think you will need is a normal t-adaptor, as the one you have is combined with your barlow but like you said, you would want a wider field of view for some of your moon (or solar) imaging. But thatβs not actually the main reason you will need it. You will need it for deep space imaging. Never use a barlow for deep space because the barlow reduces your light capture. You want every single light photon!
RE the T adaptor⦠Yes, wasn't sure I'd made the right choice by going for a 2x Barlow option. My logic at the time was, 'I want to get closer to these celestial objects.' Planets look so tiny through my 'scope. Mars appears much like a tiny red star. Saturn is by far the best thing I've seen through it. Even that is teeny tiny but you can make out the rings. Don't think I'll be getting into planetary astrophotography any time soon as much as I'd love to give it a go - don't think the results will be anything to write home about. The Moon and the Sun are realistically the best things for me with my equipment (or so I believe).
Really would like to give deep sky a go. I've been reading up on the dark frames, bias frames, flat frames, etc. that you and Mike have been discussing. They make sense to me now though I've no idea how to actually apply them and judging by your conversations (stunning second rendition of Andromeda btw), it looks like I wouldn't need to touch them initially, which is music to my ears.
Manβ¦ I just want to try EVERYTHING! π
Bloody hell, Christina, thatβs setting a high bar for a beginnerβs attempt! Great image. A mosaic, Iβve never even done a mosaic so you gotta teach us that!
Thanks also for your kind words regarding my Moon mosaic. It's actually super easy to create! This was my workflow:
Et voila! Give it a go! Not sure whether this is the best way to create theses mosaics⦠There's probably another easier/faster method (if anyone reading this is aware of any, please let me know) but that's how I created mine.
Wow, guys. We're starting to get enough material here to turn this into a mini e-book: What You Really Need to Know (and Really Don't Need to Do) in Astrophotography. When I finish this course I might haveΒ a stab at it.
Christina, the course I'm on is a Beginners course without telescopes - purely camera-based (though as the tutor, Tom Kerss says, it's 10% taking the picture and 90% what you do with it afterwards). I believe there's an ROG course starting in the New Year for telescope shooters.
Does sound like a fantastic course, as does the one that follows.
This thread has been really insightful! I've learnt a lot just by reading through the posts. Definitely would be good to be able to collate all these tips and knowledge so that it's in one place and easy to recall.
Don't completely discount the need for calibration files... I have been playing devil's advocate with my "controversial" comments, though I do stand by my premise that, for those just getting into astrophotography, they are an unnecessary complication.
When you're starting out, you have enough to worry about... focussing, aperture, shutter speed, ISO, polar alignment, tracking... it's important to master this stuff first.
Calibration frames may only offer a marginal improvement in the image quality. Sometimes, as we've seen in this thread, calibration can actually add to the noise. There are methods for dealing with this (checking each dark frame to ensure that there aren't any poor ones, and averaging them all first before applying to your light frames). All of this adds to image processing time, and can sometimes take all of the enjoyment out of the hobby.
At other times, however, calibration can massively improve an image... but it's not always a given.
Hi Tej
Just saw all this after a busy week. Nice picture. Shame that even in Brecon you can see that light pollution creeping in. Where do you have to go these days to avoid it!
Couple of things to point out.
Darks. Your procedure for taking these is correct with one flaw. Don't assume that just because the lens cap is one that no other light gets to your sensor. A huge amount of light leaks from the viewfinder, via the prism and through the mirror flap. Difficult to believe I know but its true. A 60 second flat at ISO6400 might show something but when you get up to several minutes then it will leak light for sure. Always cover the viewfinder when taking darks!
Flats. These are one of the most important calibration frames and any old light source won't do. The Aurora flat fields are the only ones that actually provide a full spectrum of light without gaps.
Flattener. Definitely needed for any scope. You have bought it now but ALL flatteners need to be set a specific distance from the image sensor. SkyWatcher do not quote a distance for this bless them, which is a constant annoyance when people ask us to make adapters for the things. It should be between 80-110mm for you 80ED. If you want to get round stars out to the edge of frame its really important you take the time to find this out. Only way to do this is with a variable adapter and lots of fiddling or a helical focuser and lots of fiddling.....
See you soon
Rupert
...and now I have read the post properly.
Did you import into DSS as a RAW file? Did you adjust the white balance before this? It is always worth having a custom white balance for Astro images, particularly if you use a filter or are close to light pollution. It will then be correct in your RAW file, otherwise you have to adjust your RAWS after for colour temp.
With your Darks, you can take them while imaging but there is no time to waste it doing this. The best time really is after you session when the sensor is hot and bothered. Personally I just want to get to bed so I do mine the next day but take a few more to try and simulate the heat and hence camera noise in the sensor.
With you processing and DSS, Its advisable to 'Save adjustments' is DSS even if you don't do any. You may notice that if you don't do this the image as it appears in DSS does not look anything like that in PS.
For levels in PS, its enough to simply use the 'dark' sample tool. PS just needs to know what the starting point for the darkest park of the image is. You can use the other samples but the bright one won't do much and the mid point will change colour balance if your not careful.
Finally, I guess you used ISO6400 to keep the exposure length down to 60 secs as you were not guided. That's sensible but as and when you are guided, you should reduce the ISO down to about 1250 max, 800 in town. There is just too much noise otherwise.
Wow, thanks for all the advice and tips, Rupert, quite a bit to soak in but soak it in I will. I want to achieve the best I can in imaging the cosmic wonders, I've invested a lot of dosh and time for it! Thanks for the Dark frame tips, will cover the sensor next time! I took my darks at the same time as the light frames btw as I knew the sensor temperatures could change
Brecon Beacons was my first proper deep space imaging session so my priority was to master the precise polar alignment, levelling, balancing and star aligning the HEQ5 Pro mount which I think I am satisfied I have. Next steps are to image using guiding as I have all the neccessary stuff for it.
Tonight if clouds are clear, I will be sneaking my mount and scope into my Dulwich sports ground and get a couple of hours of trying to image Andromeda again but this time with guiding. I will take darks too despite the lack of imaging time but I dont think I can take as many as the light frames as Mike advises. I am not going to do the flats, I have no time as I have to pack up the telescope and get out of there before they lock me the grounds! I cant do them later either because all the focus, orientation of the camera will be lost, having detatched it.
I am undecided whether to image with the Astronomik CLS filter or not. I will need extra exposure time with the CLS filter but do I really need it, can the light glow be eliminated in post process as effectively as a CS filter?
However, I still want to continue the experiment of Mike's questioning how significant a difference it makes to take the time to use darks from a beginner's perspective...and in all honesty for anyone who's budget for clear skies and imaging time is very low. I fall into both categories, beginner and low budget for sky time!
The previous experiment is now inconclusive because Rupert (and Mike) raises my uncertainty about the quality of the darks in the first place. Perhaps my darks are destructive rather than beneficial all because I didnt cover up the finder! Also, I have to check dark frames individually and stack them separately according to Rupert and Mike's tips. This I will do with my next imaging session tonight. Rupert also advises the necessity of checking white balance but I think I need to learn a little more on that part.
In the meantime. I did process another set of Andromeda frames that I took at Brecon Beacons but this time not on my equatorial but back on my Alt-Az! (I think Mike will be smacking his forehead at this point "why?!).
Well, see I want to show those that own a Goto Alt-az (such as my Celestron Nexstar, in this case) that it is possible to image reasonable deep space images with this wonderful very beginner friendly system if you have reasonable dark skies. Of course exposures have to be shorter but actually not that short. This was taken using 15 frames of 25 second exposures at ISO 6400. Its clearly no where near as good as what I have achieved with my equatorial mount but I know I would have been highly excited if these were my first Andromeda pictures and given the general concensus that a Goto Alt-Az is absolutely no good for deep space imaging, I think it gets a slightly unfair wrap. Sure, its not great for deep space but it aint THAT bad π
I didnt spend any time on trying to make it look beautiful, just quick levelling and curves (I didnt even bother get rid of the frame with a satellite streak on it!). For what its worth (which is likely zero), I stacked both a calibrated and uncalibrated version but given Ruperts comment, the quality of my dark frame may not be good so we cant really conclude anything from the comparisons.
My step backwards Andromeda with an Alt-Az Goto π (click on pic for a larger size then click on "View" then Full Resolution for the largest size where you can see the noise levels)
"Calibrated":
"Uncalibrated":
Man⦠I just want to try EVERYTHING!
Ha, well, they do say you cant have everything...but no harm in trying, go for it!
And thanks for that Moon mosaic workflow...would love to do that...for both moon and the sun. Now I just need to wait for a full moon in clear skies π
Wow, guys. Weβre starting to get enough material here to turn this into a mini e-book: What You Really Need to Know (and Really Donβt Need to Do) in Astrophotography. When I finish this course I might have a stab at it.
Christina, the course Iβm on is a Beginners course without telescopes β purely camera-based (though as the tutor, Tom Kerss says, itβs 10% taking the picture and 90% what you do with it afterwards). I believe thereβs an ROG course starting in the New Year for telescope shooters.
Love that title! All your forum titles, rock. Hey, I didnt realise they split the astrophotography courses into non telescopic and telescopic. The one I did last year, was all combined and we got to use the vintage 28inch for an imaging session on the moon which was a real privilege.
Wow you all have been busy with this experiment - many apologies for not being in touch, I've been away and it's getting quite busy this week!
Thanks Tej for having tried all of these! Much appreciated! Very interesting to see the two different images and also Rupert's advice.
I'll have a much closer look as soon as I have time to sit down properly!