Reaching my Limits!

[Marky1973]

Well, not my mental limits (although they may not be far off) but the limits in terms of what I can achieve with my set-up. I knew my set-up with a Starwave 70ED, EQ3-2 and Canon 1100D (unmodded) was always going to be limited, I am just quite sure if I have yet reached them with my attempts at Andromeda. And, if I have, what is the next step. (Bear with me, I do have some questions, but I'm a bit long-winded and might take some time to get there)

The issue is, I have tried several combinations of shots, starting at 45 second Lights, moving up to 90 seconds and then, most recently, jumping up to 180 second subs after carefully aligning the mount.

The first effort wasn't great, but the second gave me an image that I was pretty proud of for my second ever effort -

20x 90sec Lights, Darks, Bias and Flats (all stacked):

The third effort was 21x180 sec Lights, Darks, Bias and Flats (14 stacked):

Both were stacked in DSS with basic processing as per some guidance I found via this website and then (what I believe to be) histogram stretched in GIMP with some attempts at colour adjustment and a little sharpening.

I like the top image much more as it is a bit more detailed and "crisper", the bottom image suffers with some bloating I think as it was clearly at the limit of the tracking at 3 minutes (hence losing a 1/3 of the images during stacking). I will try again as I fine-tune my polar scope - I could see the tracking drift as the night went on but 2 minutes might be a better next attempt.

I am still learning with the processing and know I have a long way to go, and working with an umodded DSLR is going to limit me but I have a few questions, that I hope you guys will be able to help me with so I can work on my next steps.

1. COLOUR - I am struggling to get any colour into the image (regardless of the length of sub - there seemed to no difference between the original 45 second efforts and the 180 second subs) without over saturating it in GIMP so it looks artificial. I have upped the saturation to about 13% in DSS, but that doesn't make a great difference and I don't want to over-cook it. So, am I likely to be able to get any colour out of these sorts of subs? I was presuming not, and would expect to need longer subs, but if anyone has any tips for teasing out the colour, it would be great to know. I've adjusted the RGB channels in GIMP through the "LEVELS" menu as per a guide I found from Steve Richards, but it didn't make a massive difference and I am assuming this is because the data just isn't there. If anyone wants to have a go with the original TIFs from DSS, you can download them from here: https://db.tt/ryVrasV0 and https://db.tt/T6fpWqGs would be great to see what others might be able to get out of them?
2. I presume light pollution is going to be affecting the COLOUR, so need to think about a filter. Sky's the Limit have a filter for £39 which looks like a bargain (http://www.ebay.co.uk/itm/161339402068) but would it be that helpful (I have reasonably bad LP) or should I be thinking of the CLIP IN filter for 4 or 5 times the price? Has anyone had any experience with the STL filter?
3. Obviously a modded camera will help with imaging and, from what I have read, particularly with nebula where the red light would not be filtered out. But presumably a modded camera would help all round?

I know the obvious upgrades would be the mount and scope, but I want to make sure I know what I am doing before I upgrade those, and I am happy to learn the ropes with these for another year or so. There are plenty of targets to practice on, i just don't want to waste my time trying to get better and better pictures if I am already at the limits. And don't get me wrong, I'm very happy with the images I have got so far. I am very proud of the top image above and still show it to anyone who stands still long enough! ;-)

[steppenwolf]

The second image shows star trailing so the next stage is guiding which unfortunately also means a new mount.

Obviously a modded camera will help with imaging and, from what I have read, particularly with nebula where the red light would not be filtered out. But presumably a modded camera would help all round?

Purely to put this into some perspective, a modified DSLR camera would not give you more colour on this object - modifying a DSLR camera helps with emission objects to improve (dramatically) the sensitivity to Hydrogen Alpha (Ha) emissions. Sadly, at this point, grandfather duties call and 'grandchildren dinner' needs my attention but I will be back with some further comments as hopefully others will be too!!

[Marky1973]

Cheers Steve! I realise guiding is going to be the next big step, but I went with the minimum I could get away with to learn the ropes. Actually, I got the 70ED as a travel scope for a trip to Lapland this Christmas, but thought I might as well get some use out of it before we go! I'm planning to play with what I have for another year or so as I save up for something like an NEQ6 and an 80ED - and also find out if I have the chops for imaging for the long term!

I'm just checking I am right in thinking that I am at the limits of what I can achieve with the tools I have - and I am still very happy with what I am achieving - but if I can tweak them a bit more, I'd love to have some pointers!

Hope you enjoy(ed) dinner!

[Knight of Clear Skies]

That's a decent Andromeda and there are easier targets out there. I've also struggled to get much colour out of my images with an unmodded camera. I'd advise against using GIMP for adjusting colour though, it only handles 8 bits per channel so won't manage saturation boosting well.

[steppenwolf]

M31 is not a particularly 'colourful' object but longer exposures would bring more out. I think your first rendition is very nice indeed - you haven't burnt out the core too much but you have caught some of the fine detail in the dust lanes. I have had a quick play with the jpeg posted here and there is not that much colour to bring out further - increasing the colour saturation brings out a green hue in the core so there may well have been some moisture in the atmosphere adding to this which is also a colour zapper.

I think your biggest limitation currently is the mount, as you have already identified but imaging is a slippery slope of the deep purse kind and there is an argument for upgrading everything! I don't think you have reached the limits of your gear yet and you are obviously having fun so keep on going at a range of objects (brighter emission nebulae are still within your grasp as well) and take as many subframes as you can at the longest exposure time you can use without unpleasant star shapes.

Sausage casserole was well received all round

[Marky1973]

Cheers Knight. I've ben using the Beta GIMP package because of that reason - I was advised the processing would be 16bit in that version, but maybe I need to be saving up for PS as well!? Or maybe start dropping hints for Christmas! Although even if it is 16Bit, the auto saturation is a bit heavy-handed.

Forecast for clear skies tomorrow with a bit of luck so will have another go at Andromeda at around 120 sec subs and then start trying some clusters and some other easier targets. Just thought I would start with a biggy! :-)

[Marky1973]

Thanks Steve. I am indeed enjoying myself and there is still an inordinate amount to learn and plenty of targets. I'll keep going and, as I said, after recentering my polarscope, have a go with some 120 to maybe 150 second subs as I think 3 minutes will be pushing it until I really crack alignment. And so far, my Flower Pot seems to be doing a good job for the Flats, so I've saved a few quid there! :-)

mmmmmm...sausage casserole.....

[steppenwolf]

the auto saturation is a bit heavy-handed.

Auto-saturation in PS works well on daytime images but can be a disaster on celestial objects. For most processing in PS, Gimp and other similar software packages, 'Levels', 'Curves' and 'Saturation Boost/Enhancing' in particular can best be carried out by taking an iterative approach. A small manual adjustment repeated in several small stages can yield some very good results giving you all the control that you need and making it easy to see when to stop!!.

[Rico]

Marky, although new to this as well, i think that if you persist with doing more 90s subs, you might be able to stretch it more and more. You've only captured 30 minutes worth of luminance in that first image; most great examples of the andromeda galaxy are captured over many hours. Ever tried getting a couple hours worth of 90second subs with your current setup? I know, longer exposures are better, but worth ago, before you take the inevitable plunge into deep pockets.

[Marky1973]

@Steve....slow and steady it is!

[Marky1973]

Hi Rico, I think that will be the next attempt, shorter exposures, but more of them. Thinking of 120 seconds and try and get a couple of hours in.

Good luck with your efforts Rico!

[Rico]

I'd be more inclined to stick to 90-120, unless your PA is spot on. I imagine that if the PA isn't correct and there's significant trailing, DSS won't be as efficient in trying to get that colour out when finding the mean/median, since you've spread the 'information' over a greater region in the image. I'll be having a crack at andromeda tomorrow night as well, need to add some more subs to my data set as well.

[Marky1973]

Rico, if you take subs on different nights, and presumably darks/flats/bias at the same time, can you just stack them all together? Or do you have to stack each night and then batch them together after that? Or can you take your light subs over a few nights and just use one set of dark/lights/flats? I'm just thinking about you are supposed to take them in the exact same conditions as the lights.

Sorry to throw questions at you!

[Rico]

Marky ... I'm hoping to find the answer to that question soon as well ... Lol

[Astrosurf]

I'll leave you with these excellent comments. I just wanted to say your 90s image is very good. Keep trying!

I'm not 100% sure about this, but you should be able to get PS cheaper for your school kids. I got mine a couple of years ago for £200 for my teenage daughter. But I think Adobe demand subscription payments for their Cloud package now. Ask Adobe anyway. Be prepared for a hard sell if you have a choice. Don't get roped into Cloud! I had to get quite cross with the young man who wasn't listening to me. Good luck!

Then there's also StarTools and PixInsight. A friend uses one and swears by it, but I can't remember which! I'll find out and let you know.

Alexxx

[IanL]

I had a quick look at the first TIFF link you supplied and I'd make the following observations:

1. It isn't the pure output from DSS to me, it seems like it has been saved with adjustments applied.  Ideally if you want people to have a bash at your image to show you what is possible, you need to save the pure linear image with absolutely no adjustments applied after stacking so that the starting point is known.

2. There does not appear to be any colour in the image and it seems to be greyscale (even though it has three colour channels) either as a result of your stacking settings or as a result of adjustments applied after stacking or the file format used to save it.  The RGB channels seem to contain pretty much the same data which they should not, so you need to go back to square one with this.  I'd would expect to be able to apply a simple histogram stretch and saturation to the stacked image and get some colour (probably a lot of colour to be honest) but I can't.  This isn't a camera or LP problem, this is a processing issue of some sort prior to creating the TIFF.

3.  Looking at the stars (especially the brighter ones) they have hollow/doughnut cores, i.e. the centres are darker than the outsides.  This might be due to stacking or more likely it is due to processing - unlikely to be the raw data.

4. There is definitely a bit of a gradient in the image (brighter at the top than the bottom. It's not that severe and could be removed with DBE in PixInsight or another gradient tool

I'd think that you should be able to get a far better result out of the 90s subs you already have with proper processing (it is a steep learning curve, and all to easy to assume it is your data that is at fault when actually it is just the need to spend a lot more time learning how to process it).

Personally I'd invest a lot of time in learning how to process and a bit of money in better processing software.  Lots of people use PhotoShop but it is very expensive for something that is not at all optimised for astronomical image processing. PixInsight (my preference) is a lot cheaper to own, but it is a very steep learning curve and might well be an easy way to put yourself off the whole idea!  I'd suggest you take a look at StarTools (http://startools.org/drupal/). There is a free demo which doesn't allow you to save but will let you do everything else, and if you like it the cost is only about £40.

Using a package designed for the job and with a (by all accounts) not very steep learning curve will surprise you with the results you can get.  You'll definitely need a better mount sooner or later, but all the data in the world isn't going to help you if you can't process it. Personally I would have been very pleased with the data you have obtained if it had been one of my first attempts, so keep it up!

[Christopher Davenport]

Good advice above and well done on your image.

I don't think you reaching your limits at all, you have chosen a tricky subject that puts the most seasoned imagers into a spin.

Drift aligning your mount will help you get up to 180s plus targets in the north near polaris are more forgiving than those in the south. 

You can also look at stacking all the data that you have together using HDR in DSS.

This will provide a lot of depth and range to the image, this will allow you to push the DSS saturation up to 20% to 30% to extract a lot of colour.

Hope this helps abit.

[dph1nm]

If  you use 8-bit GIMP and remember to save the DSS adjustments (including 20% colour saturation) then I can tell you that everything works fine. I have not tried the beta version of GIMP yet.

NigelM

[Marky1973]

Thanks Alexx, I'll look into that possibility! Used to be able to get some software deals through work, but those days are gone! :-(

[Marky1973]

I had a quick look at the first TIFF link you supplied and I'd make the following observations:

1. It isn't the pure output from DSS to me, it seems like it has been saved with adjustments applied.  Ideally if you want people to have a bash at your image to show you what is possible, you need to save the pure linear image with absolutely no adjustments applied after stacking so that the starting point is known.

2. There does not appear to be any colour in the image and it seems to be greyscale (even though it has three colour channels) either as a result of your stacking settings or as a result of adjustments applied after stacking or the file format used to save it.  The RGB channels seem to contain pretty much the same data which they should not, so you need to go back to square one with this.  I'd would expect to be able to apply a simple histogram stretch and saturation to the stacked image and get some colour (probably a lot of colour to be honest) but I can't.  This isn't a camera or LP problem, this is a processing issue of some sort prior to creating the TIFF.

3.  Looking at the stars (especially the brighter ones) they have hollow/doughnut cores, i.e. the centres are darker than the outsides.  This might be due to stacking or more likely it is due to processing - unlikely to be the raw data.

4. There is definitely a bit of a gradient in the image (brighter at the top than the bottom. It's not that severe and could be removed with DBE in PixInsight or another gradient tool

I'd think that you should be able to get a far better result out of the 90s subs you already have with proper processing (it is a steep learning curve, and all to easy to assume it is your data that is at fault when actually it is just the need to spend a lot more time learning how to process it).

Personally I'd invest a lot of time in learning how to process and a bit of money in better processing software.  Lots of people use PhotoShop but it is very expensive for something that is not at all optimised for astronomical image processing. PixInsight (my preference) is a lot cheaper to own, but it is a very steep learning curve and might well be an easy way to put yourself off the whole idea!  I'd suggest you take a look at StarTools (http://startools.org/drupal/). There is a free demo which doesn't allow you to save but will let you do everything else, and if you like it the cost is only about £40.

Using a package designed for the job and with a (by all accounts) not very steep learning curve will surprise you with the results you can get.  You'll definitely need a better mount sooner or later, but all the data in the world isn't going to help you if you can't process it. Personally I would have been very pleased with the data you have obtained if it had been one of my first attempts, so keep it up!

Thanks for the detailed response Ian, some good stuff to think about there, and you are right with the TIF file, it does have some adjustments so will remember not to apply anything in future if I am asking for help like that - as you say, it is a steep learning curve, and I am working my way up from the bottom slowly!

I'll also have a look at those other packages as well. Have heard a lot of good things about PixInsight but, if it is that hard, it might be better for me to start with StarTools!

Thanks again

[Marky1973]

Good advice above and well done on your image.

I don't think you reaching your limits at all, you have chosen a tricky subject that puts the most seasoned imagers into a spin.

Drift aligning your mount will help you get up to 180s plus targets in the north near polaris are more forgiving than those in the south. 

You can also look at stacking all the data that you have together using HDR in DSS.

This will provide a lot of depth and range to the image, this will allow you to push the DSS saturation up to 20% to 30% to extract a lot of colour.

Hope this helps abit.

Thanks Christopher! Not sure what HDR is but I'll have a play tonight! I've seen the setting though and I presume it is only available with the lights? Seems to be when I have just had a peek, but will experiment!

[Marky1973]

If  you use 8-bit GIMP and remember to save the DSS adjustments (including 20% colour saturation) then I can tell you that everything works fine. I have not tried the beta version of GIMP yet.

NigelM

Thanks Nigel, perhaps I should go back to the previous version of GIMP for a bit? Do you mean to save the adjustments applied or just embedded? Which one is best to work with?

[uhb1966]

the current 16/32bit beta of gimp (2.9) is now stable enough for basic astroprocessing ...zero $ ... i use it (however the mac version) and like it very much

[dph1nm]

Do you mean to save the adjustments applied

That is what I do. I find if I don't I just get a black image (as I think GIMP just scales the 16bit DSS image down to 8 bits by dividing throughout by 256).

NigelM

[Marky1973]

GIMP is great for free UHB and still getting to grips with everything in it....trying StarTools as well as they offer a free trial and there is a big learning curve with that as well!

Thanks Nigel - so many options, I just end up with something different every time I play - which is probably part of the fun!