Attempted widefield but stuck on stacking

[Delboy_Hog]

Hi All,

I'm just trying to get into the wonderful world of widefield, but I'm having a few issues, particularly with DSS.  I’ve little doubt that the issues I’m facing here are entirely down to user error, but grateful for you taking a look and pointing me in the right direction!  (Apologies for quality of images on here, I've compressed them so that they fit, but hopefully they'll give you an idea of what I'm on about!)

Here's a single frame of my widewield attempt at capturing Cassiopeia and the surrounding skies.  It was a 30 second exposure taken at ISO 3200 with Mrs_Hog's Canon 1100d camera, which she kindly lent me when we were in Cornwall back in April.  (Due to moving house I've only recently had chance to get the computer set up and get playing with my data).

 

Bit of light pollution in the lower left, and transparency actually not great – when I slide through all of the frames, you can actually see a very thin layer of haze moving slowly across the images – so my first question is: how much do you think that’ll affect the final image? 

Then there’s a question about my ISO setting.  I used ISO 3200, though looking back at a couple of trial shots I did at 1600, would this have been a preferable ISO level to use?  There are fewer stars at 1600, but a considerably less background haze / noise.  Does it not matter too much if you have a good number of darks etc?  Having looked at the many wonderful images on here, most people seem to be using much lower iso.  Do you reckon I can salvage anything from using this high iso, or is it destined to be too noisy to use?

So then the DSS questions:
Firstly: stacking images.  When taking the photos, I had the camera produce a jpeg and a CR2 file.  When I try and stack the jpeg in DSS, I get a message saying that darks won't work because of lossy compression.  Understandable, and no problem.  However, if you do continue, does it still try and stack the darks, even though it warns that it won't work?  This is the file it produces, approx. 70 x 30s lights and 60 darks.

If you peer closely at it, you can see there are plenty of stars in there, and I had been warned that this is how files would come out of DSS, but that there was a way of darkening the background at a later point.  Now, when I try and stack the CR2 (raw) files , the final image looks very different.  Does anyone know why when I use the CR2 files, it create a small vertical (portrait) sliced version of a stacked image, whereas the stack of jpegs produced a more logical, landscape image? 

I guess this is user error, because clearly hundreds of you run raw data through DSS regularly without issue! Grateful if anyone could point out where I went wrong!?

So from my lurking on here, I get that DSS is not a processing tool, though from a couple of youtube tutorials I found, I understand that moving the sliders can bring out a bit of detail from your image, darken the background etc – something I’m keen to try at the moment, given my lack of any processing-type software.  My question is, how do you know what to go with, or what is an accurate representation of what's in your images?  The images that seems to produce are a bit crazy, and don’t look a lot like the original frame!  With the data I captured, my stack image could look like any of these, depending on what I do with the sliders in DSS:

From stacking the jpegs:

And from stacking the raw files:

So finally, are there any DSS tutorials that you folks would recommend?  I'd love to get the hang of this, and am determined to do so!  Am I doing something wrong, or was the mistake I made in the data gathering, rather than the stacking?

Please do say if none of this rambling makes any sense, but many thanks for any light you can shed on all of this for me!

[AstroFrog]

_{Hello mate,}

_{I'm no expert, but I wonder if the ISO was too high, and it's having a hard time pulling out the detail from the light pollution / haze?  I'm sure more knowledgable folk will be along to point you in the right direction though!}

_{Might sound like a silly question, but are you sure the stack is made up of the correct files?  The first time I tried I had lights in with my darks and vice versa, plus a couple of my test shots at different ISOs and exposure lengths in there too.  No idea what effect that has on DSS's ability to do the job, but worth double checking!   I can't help with the whole 'raw files making a portrait image' thing I'm afraid.  I've not seen that before.}

_{There's plenty of stars in there though, so I guess the skies must have been pretty dark?  Keep at it, it's worth it in the end!}

[Knight of Clear Skies]

Hi. I'm no expert on DSS, but I can help you with a few things. I've hardly changed the default settings but I'm getting some reasonable results out of it.

Firstly the narrow vertical image when stacking the RAWs: I had exactly the same problem with my 100D and you need a newer version of DSS. It's a bit hidden on the download page, you want the one with "Latest DSLRs support (DeepSkyStacker 3.3.4)", there is a download link a couple lines down from that.

On ISO levels I don't think there is a 'right' setting as such, it depends on the lens speed, sub length and sky brightness. The advice I've heard is to take some test exposures and look at the histogram (press 'disp' when reviewing the picture). The peak of the histogram should be no more than a third of the way across (I've heard some people say at 20% is better). This helps avoid overexposing any details.

Your example sub looks pretty good to my eye. A bit of light cloud will have some impact on image quality but won't ruin the image. Here's a quick Andromeda I did the other day with some wispy cloud scudding across. Most of it was on the right hand side of the frame but some did cross the galaxy in a couple subs.

When DSS stacks an image the output tends to be much too dark, but the detail is hiding in the .tif file. You then need to adjust the colour curves in DSS or another imaging processing program to stretch the data.

Hope that is some help.

[jgs001]

The over brightness looks like the screen review that the DSS has created and guessed at the settings for. Try saving the output file to a TIF, adjustments embeded (not applied), and see how it looks then. I'd expect it to look pretty dark, and this will then need some processing to pull out the detail. 

As has been said with the RAW's you need to use the beta version to stack those, the RAW processor in the version you have doesn't support your camera. 

[Delboy_Hog]

Thanks for your help guys!  Great, ok so that version of DSS will get downloaded this evening for sure! 

Like you John, I expected the final version to look a little darker, even with the noise that will have been generated at ISO 3200, but hopefully when I run through the stacking with the proper version I'll get the result that I'm sure must be lurking in there somewhere!  I'd had one successful run through back in April stacking a handful of Orion frames (albeit jpegs) so I know it's possible!  One I get the hang of the DSS stage I might start looking into some proper processing software.

Knight of Clear Skies, your images are coming on really well!  Have you got the camera on your EQ3-2?  That Cygnus image is excellent - how dark are your skies?  You mention the LP from London, do you have to use a filter for it?  Loving that Andomeda too - a few more subs in that and it'll really start to shine!  I took a whole bunch of almost identical subs the other night, so that's in the DSS queue as well!     I'm limited to 30 second (max) subs at the moment, but it still shows up nicely in that time, so it's a definite target for more work over the coming weeks hopefully!

Many thanks again for your help guys - it really is appreciated!  Hopefully I'll be back here with a reasonable image before too much longer!

[Delboy_Hog]

Cheers AstroFrog, yes, I had to double check the subs and there were a couple of bogeys in there!  Nothing too drastic, but a couple of shorter exposures I did to test the skies and a couple of different ISO levels.  I don't know whether that would mess it up or whether DSS discounts them, but I've tidied up my sub-filing system now, so hopefully I won't get all muddled up any more!

I'll definitely be sticking with it, I love the widefield images that you can get, so I'm determined to master it!

[Knight of Clear Skies]

Thanks, I feel I'm making progress and I'm enjoying myself. Yes, I have the camera mounted on the EQ3-2. I'm consistently getting 2 minute subs at 50mm, however I'm going to need to fit a polar-scope to make full use of my longer lenses.

My imaging site is a mile out of town and there is a fair bit of light pollution around, but how bad it is depends on where I point my camera. Since I posted the Cygnus image I've realised I was a bit turned around and was looking east, so the LP at lower right is probably Harlow. I don't have a filter but it's one of the things on my wish list.

have fun with your widefields.

[Alien 13]

I do find widefield to be the fun end of astrophotography with fast camera lenses 30s to 2 mins is often all you need and in my case is about the limit with the LP i have but it does mean i can get a session done in about an hour and that includes setup time.

Processing is a bit of a black art but i find i learn with each image and often revisit old data with new techniques.

Knight of clear skys i do like your widefield Andromeda it would make a cracking image with some more data.

Alan

[PorkyB]

As an alternative to installing the newer version of DSS, you could use the Canon utility to convert your CR2 files to TIFF files.

[Delboy_Hog]

That's interesting Knight, and good that you can still get that quality of image even with the LP there.  And definitely keep us posted on that Andromeda if you put any more data on it!  I do like seeing it framed a bit more in widefield.  Don't get me wrong, the close-ups where it fills the frame blow my socks off every time, but I do like seeing it with a bit more surrounding sky as you've got it there. 

Thanks Porky, I hadn't thought about doing that.  I must admit I'm still trying to get used to which bits of software can do what with various formats, I must make more of an effort to understand it all!

I actually haven't had time to get any further with this yet - work keeps getting in the way.  Hopefully I'll get a couple of quiet evenings this week to spend some time on this - it doesn't look like I'll be getting out with scope / camera for a few days, so I should use the time to get more familiar with DSS.

Like you say Alan, the processing seems a bit of a dark art - a bit daunting to me at the moment, but I'm sure it'll be ok once I get used to it!  I sure hope so anyway...

Thanks again all for your helpful responses.

[kalasinman]

Hi Derek, not an expert certainly, but here's some ideas. DSS has a very nice manual included. It's at the bottom of the left hand section. Don't change any settings at first. Make a folder for each imaging session, under a master imaging folder. I like to keep a folder under the master folder called example which contains sub folders "light, dark", then when I start a new session file I just copy example and re-name it. Bias and Flats can each go in a folder under the main, named by ISO and lens.

Place your bits in the appropriate folders, then when you open DSS just click on first of a set (lights for example) and shift-click on the last of the series to select all. Then drag and drop to the bottom pane of DSS. Answer whether dark. light. or whatever. Once you have them all in there they should all be checked. Then click on "recommended" on the left pane. DSS will show a series of setting recommendations for your data. If there are any red lines, click on the green lines below them to select a better setting. Then click register and it will register and stack.

As I recall, the finished image will auto save in TIFF to the file where the lights were stored. I don't like to do any further work at this point and close DSS.

Processing. I use Lightroom and Photoshop together, but you can download GIMP for free and it does a good job. Load your TIFF into your processing software and there you are, ready to produce an image.

BTW don't worry about your histogram, if it's all to the left. The data is there. That's the first stage in processing, to stretch your histogram.

ISO. You can usually recover data from a too dark image, but the blown out light areas from too high ISO are gone.

There seem to be two approaches to wide field, one off, and stacking. One off generally needs higher ISO. Most cameras have a sweet spot usually ISO 400-800 where the dynamic range is good and the noise level not excessive.-Jack

[Delboy_Hog]

Hi all,

Sorry for the slow response, work has taken over again, so I've not had enough spare time.  Anyway...

Jack, thank you for your very helpful response.  That certainly seems like a very sensible folder system, thank you for that!  I'm definitely still working on trying to get the right ISO setting for the situation, but hopefully that'll come with experience!

GOOD NEWS!  I've had some stacking success at last.  New version of DSS downloaded and it handles the raw files without a hiccup this time.  I've still not been able to get anything useable out of the data that I was talking about further up, (I think the ISO was just too high) but here it is:

It's such a shame as I can see LOADS of stars in there, but hey, at least I'll know for next time.

Having had the chance to get a bit more data, I have managed to start getting closer to the kinds of results I'm after.  Here's an example:

Widefield Andromeda, around 30 x 30 second lights and the same number of darks.  The jpeg version looks horrific on here, but the tiff file that DSS churned out looks a bit better!

So what is that grey dome in this picture?  Is it light pollution, or an issue with the camera?  Is that kind of thing correctable in processing, or is this data doomed!

I think I might try and download GIMP later this week if I get chance, as I hear that's a fairly useful tool for processing.  I wonder if I provided a Dropbox link to a couple of widefield tiff files straight out of DSS stacking, whether someone on here might fancy a crack at seeing what they can get out of them?

[Delboy_Hog]

I've now created another post with the Dropbox links in to the tiff files if anyone does want to have a go at processing any widefields!

http://stargazerslounge.com/topic/221886-any-volunteers-to-try-processing-my-widefields/

[kalasinman]

Hi Derek, I'm learning too. I hope you don't mind my suggestions.

First, it would be helpful to include EXIF data when you post a picture so folks know what lens, camera and settings made the picture.. Not sure what ISO you use. Don't worry if resultant output from DSS is a little dark. When you get around to stretching the image , it will all appear.

I'm just learning processing but I'm brave, I'll give it a try. ;-)

ISO...try taking  10 set subs at reducing ISO to see what you like the look of. 3200, 1600, 800, 400.

Always try to use a lens hood.

A 30sec exposure would be marginal for a wide angle lens , less as the FL gets longer. Try some sets at 15 secs, and 10 as well. Shorter (10 image) sessions will give you a quicker tune-up. Once you have it, then go longer. I find on a static tripod that 25-30 images per stack pushes the limit of DSS to correct for field rotation.-Jack