(Note: I’ll come back and add some screenshots at a later date!)
Astrophotography is great when it all comes together, but sometimes that road is full of frustrations and problems.
I bought my first “real” telescope in 2015 (a Celestron 6SE) that whetted my appetite for the hobby. Whilst the Alt-AZ based scope didn’t lend itself too well to imaging the night sky, ever since I’ve seen distant objects through the eyepiece I wanted to capture what I’m seeing and share with people; capturing images is a great way to get people to see there is much more to life than what we experience on Earth. However, it wasn’t until the Winter of 2017, a little more than a year since this blog post, that I’ve been learning how to to do astrophotography properly. I say properly – I guarantee there are some old-timers that will baulk at my technique and choice of software for how I do things. This just spurs me on to encourage more people down this route!
So what is this post all about? This is my routine for
- image capture using SharpCap Pro (£10 per year)
- image processing using Astro Pixel Processor (€ 60.00 / yr or
€ 150.00 own outright)
- image post-processing (Adobe CC / £9.99 a month)
I will go into detail of what I use, how I use it and why. There are variations in software and hardware to use, so see this as a guide only. The key thing to remember though – like anything that involves computers, and indeed any photography, is that:
INPUT DATA -> PROCESSING -> OUTPUT
The output (your final image) is only as good as the quality of the data you input (the images you capture under the clear skies). Don’t expect miracles if you aren’t utilising your equipment properly to capture images in the first place.
I would also point out that astrophotography is subjective in that it involves a lot of personality and personal opinion. Astrophotography is part science, technology and creativity. Whilst the science and technology are definitive (this is the kit you choose and their capabilities), the creativity piece is up to you on how you want the final image to be presented.
What You Need
- Suitable astrophotography scope, mount and cameras. I use
- SkyWatcher EQ6-R Mount (recently hyper-tuned as a StellarDrive 6R)
- TS Optics Imaging Star 65mm Quad Apo or Altair Astro Starwave 102ED-R
- ZWO ASI294MC Pro Cooled Colour Camera
- Pegasus Motor Focuser
- Various filters if needed, depending on location. Predominantly shooting with an Optolong L-Pro Filter under Bortle 5/6 skies.
- SkyWatcher EvoGuide 50mm Guide Scope
- Altair 130M GPCAMv2 as Guide Cam
- Windows PC for image capture
- A PC (Windows, Linux or Mac) for image processing (because Astro Pixel Processor (APP) can be run on any of those machines). You can always use the same PC you’re using for image capture too of course. Image processing is much more processor intensive than image capture as they serve entirely two different processes, so bear that in mind if you’re deciding on purchasing any computer kit. For example, I have a cheap laptop for when I’m out in a remote location, because imaging is not resource intensive; I have a small Intel Core i3 with 8Gb Ram and 2Tb hard drive very small form factor PC for image capture at home; my main image processing PC is a Ubuntu Linux 24Gb RAM, Core i5 with 256Gb NVMe card because it takes a lot of resources (Ram, CPU and Disk) to stack and process images – the bigger the better – and you don’t want a big, expensive laptop in the field where it could risk being broken when all you’re doing is capturing images.
- SharpCap Pro (pay the £10 a year for the Pro version to give you access to functions that will ultimately help you with your image capture – such as Smart Histogram that will help, amongst other things, determine if you should be doing 5 minute exposures or 10 minutes will be more beneficial without wasting time just for extra noise)
- Astro Pixel Processor for DSO image processing. Note this software is not for processing of Moons, Sun or Planets.
- Adobe CC for post-image processing. I utilise both Lightroom and Photoshop. If you’re much more familiar with the free GIMP image processing software, use that, however if you’re serious about astrophotography – there are lots of guides and features in commercial offerings that may give your images that extra needed help that warrant the extra cost.
- Software for controlling your mount from your PC
- ASCOM and EQMOD is the standard here. It will connect to your mount and give you control via your chosen software or the EQMOD interface itself. Remember, you are wanting to be away from your mount and telescope when you’re imaging. Knocking cables or moving near your mount can cause enough vibrations and movement that may be noticeable when doing long exposures. See this page here for setting up ASCOM.
- I use Cartes du Ciel for my click and point zooming around the skies. See this page here for setting this up. This is free software.
- Optional, but recommended – but may cause more problems than it solves if you’re trying to just get going: PHD2 for Guiding. See this page for details.
How To Do It
There are 3 main stages that will be talked through that were described above: capture, processing and post-processing – but first you need to make sure you understand your kit and capabilities, and set up correctly before you expect Hubble-like images to appear magically on your laptop screen. To setup your scope and mount, ensure it’s polar aligned correctly, etc. check out this page.
Image Capture in SharpCap Pro
If we could all just point and shoot, this would be easy! But unfortunately, just like terrestrial pics – you can take some good ones, and you can take some bad ones. What makes astrophotography much harder is that there are many more factors coming into play: You’re shooting long exposures at night with various pieces (and capabilities) of equipment and environmental factors that on one night worked great, but the next makes you look how much to put your kit up on Ebay! You also add in another problem with astrophotography in that it is like the days of film cameras but with an added hint as to how things may turn out: you have to wait for the processing to find out how good the data really was.
All that aside, this is my process of an evening session. Note the preparation. Remember, your output is only as good as your input! Getting your mount setup correctly and getting your imaging focus correctly will lead to a better imaging sessions.
- I load up all my software for controlling my mount and cameras.
- I go through my Polar Align routine.
- I load up SharpCap Pro and select my camera. I have a default setting for my ASI294MC Pro that I use for the initial setting up (focus and plate solving):
- Gain 121
- 1 Second Exposure
- Temp -20C
- I use CDC to slew to a bright star like Vega or Capella. This is so I can set my focusing and tell my mount where it is in the night sky.
- If the display looks a little bright or dim, in SharpCap Pro (not free version), find the histogram in the configuration box on the right of your image and hit the lightning bolt. This will correct your preview. Now I can see the bright star, I usually end up doing a good enough job of rough focusing here, ensuring the stars look visually sharp by controlling my Pegasus Motor Focuser.
- I then set the exposure to 4 seconds to allow me to Plate Solve as it brings out more detail and stars that will be needed for the plate-solving to succeed. Adjust this time appropriately for your kit and sky conditions.
- I Plate Solve to synchronise my mount with CDC so that when I go to more objects, the mount will end up where it is meant to. This should move your chosen star to the centre of the view.
- When Vega (or Capella) is in the centre of my view I pop a Bahtinov Mask on the end of my telescope and use the finer controls of the Pegasus Motor Focuser to get a perfect Bahtinov Mask looking star with star spikes evenly spaced. You can use SharpCap Pro to help here. The magnifying glass icon on the tool bar has a focusing help option appropriate for a Bahtinov Mask, however I found it too picky so I now just visually get the spikes perfect.
- Once focusing is good, I now take Flats. These are some of the extra images to take when talk about Calibration Frames in the processing section below. See this page here [TBD] on how I do this.
- Flats are used during processing to remove any permanent image artefacts from the imaging setup (e.g. vignetting) or temporary artefacts on the sensor or glass such as dust. PLEASE TAKE FLATS WHENEVER YOU IMAGE, SWAP FILTERS AND ADJUST YOUR CAMERA because dust can move. Just don’t make any changes when you have taken your Flats is the more simpler rule to stick by!
- Flats can also be taken after you’re imaging run BUT I do them first because I’m usually going to be taking images of multiple targets during a night. By taking Flats first, I can process, say M42, whilst my kit is then pointing at M45. If I wait for the end of my session, I have a lot of wasted time and at 3am, sometimes it’s easy to forget to take those Flats!
- We’re now ready to slew to a target and start imaging! In this case, we’ll head to M42. I choose this in CDC and click on this and selecting the “Slew to M42” menu option. Whilst its doing this, I set the exposure in SharpCap Pro back to 1 second duration so I can have more instant feedback when framing. Despite my mount correctly slewing to M42, you may want to adjust the fine position so it is framed correctly using the onscreen controls.
- When I’m good with the position, I do a test exposure. For M42 – it’s bright and my ASI294MC Pro is sensitive, so I set a 30 seconds duration. After the 30 seconds the preview may be too bright, so that same histogram in the control panel on the right – hit the lightning bolt again to correct this. This is reason enough to spend £10 a year on this software! Now the image may look too green or blue, but don’t worry – you’re going to be taking RAW/FITS images and these colours will be balanced out during processing. Do NOT assume what you’re seeing here is exactly what your end result will turn out to be. What you’re looking for is appropriate details and good stars.
- For DSOs you’re stacking images together (in our case, in APP). This means we’re taking the same length exposure images multiple times. I tend to choose 30 images and start from there. So in SharpCap Pro, first set the Target Name to be “M42”. This will mean all the images we capture will be put into a folder in our Captures folder under M42. When doing multiple targets in an evening, you’ll benefit from this!
- Ensure you are using 12/16-Bits FITS by setting the file type from the control panel on the right. Do not choose RGB32. These are compressed images, like JPEG – you will lose detail.
- Now press the Start Capture button from the menu toolbar. This will pop up a dialog box that asks you how many exposures you want. Set 30 and click Capture.
- This will do exactly as you expect – it will just capture 30 images of 30 seconds and store them. These images you have just captured are known as Lights.
- At this stage, you would then take Darks by covering the end of your scope with your cover and repeating the process above (naming them appropriately in the Target field as “Darks_30Secs” for example). Darks are another type of Calibration Frame images remove the amp glow or other light caused by the warm part of your sensor flushing the data from the sensor, and noise from your camera that are there that would show up as specs in your image. These are again used during processing. You need to take darks at the same gain, temperature and duration as your lights. However, if you have a cooled camera like the ASI294MC Pro, you can take these darks whenever and use them over and over again during processing. Just remember to do it at some point before processing! The obvious advantage of doing darks (using a cooled camera) at a different time to your imaging is that you’re not wasting valuable light time with your
- Now because M42 has lots of bright and dim detail, I then increase the duration to say 60 Seconds and repeat steps 12-16.
- I then increase to 120 Seconds, then 240 Seconds, repeating the above. The end result is lots of Lights of different durations for the same object, in the same position with two intentions:
- Either one of those durations could be “Good Enough” for what you want to achieve (so you know next time you can just use the one duration for this target for your kit)
- Can use the different detail gathered at each exposure to create a High-Dynamic Range (HDR) processed image. The reason this is chosen here for M42 is that the bright core can easily become over-exposed on a long exposure, but the detail of the nebula is only exposed on a long exposure. So your different exposure lengths are designed to pull out the detail for the appropriate part of M42.
Note that the above process, minus the multiple duration settings can be used for any target – some benefit from combining multiple exposures, some do not.
- finally take some Bias frames. These are needed because we’ll be using different duration Darks, and I didn’t suggest take Dark Flats too. Put the duration slider all the way to the left in SharpCap Pro and take 30 frames with the end of your scope covered/cap on (like Darks but shortest exposure possible).
- Once done, I download these to my processing PC whilst I’m choosing my next target to image. Repeat steps 10-20 for my next target!
Image Processing using Astro Pixel Processor
Now that you’ve spent hours trained on your chosen target, in this case M42, it’s time to spend many more hours processing the images to create a stacked image. Stacking is the process of combining the fruits of your labour from the night under the stars and combining all the good parts of the image to create your more robust, detailed image. You simply don’t take a single image in the night and expect it to be full of all the detail you expect.
My favourite software of choice for doing this is Astro Pixel Processor. It has been designed from the beginning to give very good results from sensible defaults using an interface that takes you through numbered stages up to “Integration”.
- The first thing that will pop up is asking you for a Working Directory where it will store master files and the combined image. I find setting something like “Processing”/”Target Name”/”YYYY-MM-DD”/ e.g. “Processing/M42/2019-01-01” is appropriate.
- For my One-Shot-Colour (OSC) ZWO ASI294MC Pro, I first select tab 0) RAW/FITS and set the option “force Bayer CFA”, else my images will turn out in greyscale.
- The 1) LOAD tab is where we select and load up all the image files we will be using to process our image. As we’re running through M42, we’re going to be selecting an advanced option: Multi-Session processing.
Ordinarily, you wouldn’t need to select this option. So if you’re following this guide and you just want to process your 3 x 240 second M45 images, then don’t select this option. What this option does is allow you to select different calibration frames (Darks, Flats, etc.) that match the Lights. Remember that if you took Lights of 30 seconds, you need Darks of 30 seconds too. As we’re adding in Lights of various duration, we need to match the Darks too. This is the Multi-Session processing option.
- We load our Lights by clicking on the Light (grey) button. Find your M42 30 Second Lights that you took in SharpCap Pro in Step 14 above. Now, when Multi-Session processing is selected, an extra dialog box will pop when you load them, asking which session to store these in. Choose Session 1.
- Repeat Step 4 for your other duration Lights, placing them into different Sessions.
- Where it says “sort” I choose “quality”. What will happen is that when we get to stage 3 (analyse stars), it will apply the calibration frames and analyse each Light and order them according to quality rating. This is handy to pick and discard good and bad frames for stacking.
- Now we load our Flats. Click on the Flat (grey) button and load up the Flats we took in step 9 of image capture. When it asks for which session to load these to, select “all sessions” because we didn’t change any of the imaging set up, or move any equipment between the image captures, making the same set of Flats appropriate for all sessions.
- Now we load our Darks. You may have taken them last week, or during the night after each capture run for that duration-gain-temp combo. Find them and add them to the appropriate sessions. For example if you have 30 Second Lights in Session 1, add the 30 Second Darks to Session 1.
- If you have Bias frames (not always entirely necessary, depending on who you speak to and what kit you may have) then load them too – and apply to “all sessions”.
- We’re now ready to head to tab 2) CALIBRATION. This creates Masters for your Flats, Darks and Bias as well as give you an option to create a “Bad Pixel Map”, also known as BPM. If this is the first time using APP and you don’t have a BPM, select this option. In the future we can re-use this created BPM for future processing. This is unique to APP and is a map of the bad pixels for your camera, thus allowing these to be used over again to be processed out. If you have a BPM already, head back to the 1) LOAD tab and load it where it says BadPixelMap and don’t select this option.
- I select on this tab:
- integrate: Average for all the calibration frames (Darks, Flats, Bias) (because I use 30 frames of each) and set Sigma outlier rejection.
- Create 32-Bit Masters
- Create MasterBias, -Dark, -Flat
- What this means is that we can re-use these Master images, rather than spend time going through these calibration processing for future runs. Dark Masters can be used over and over again, on different nights. I use the same Bias over and over again as I only use one camera. Flats are unique to your session, but you may be processing multiple targets during the night – so save time on the next image target processing by loading the Master, rather than 30 Flat frames.
- I select adaptive pedestal/reduce Amp-Glow. The camera I have seems to benefit from this. If its not needed, it usually doesn’t adversely affect your calibration frames, so select this anyway if unsure.
- Ensure calibration warnings are set
- When I push the “create Masters & assign to Lights” button, it will go through and take the 30 x Flats, 30 x Bias, and each 30 x Darks I took during the imaging run and make Master files of them, and then tag the Light frames to say these images can have Calibration frames applied. For all intents and purposes – this is Calibration complete – very easy!
- The fun now starts as APP does the real work. The tab marked 3) ANALYSE STARS is the first indication of quality. Keep all the defaults then hit the analyse stars button. This process will take a short bit of time and the result is the #stars & star density field of the file list filled in, as well as a few other columns of data around quality and noise. Scroll along to the “quality score” column and you can get a good idea for your good and bad frames (its all relative, and the idea is most of the frames should be of equal values. If you have some that are hundreds of points off the main Integration Ref scored frame, consider removing it by right clicking on it and select “Remove”. This will make it be ignored during registration and integration.
- The next tab is 4) REGISTER. Keep all the defaults and scroll down slightly to the button that says “start registration”. This will register all the calibrated and star counted Light frames which will form the integrated (stacked) image.
- The tab 5) NORMALIZE processes all the Light from as a single entity to normalise the light and backgrounds across all the Lights. This is unique to APP here. For our multiple duration exposures we’re doing with M42, I choose mode Advanced. Ordinarily just keep all the defaults. APP is very good with the defaults as a fantastic starting point. I also select the “neutralize background” option to keep a consistent background that I later correct slightly once my image has been integrated. Pressing the “normalize lights” will see APP process this for a short while – time to get a coffee!
- The final tab before you get a great hint as to how the very finished product will look after post-processing, is the 6) INTEGRATE tab. To us all, this is the main event! As we’re performing a multi-session integration, we have some extra options available to us in that we can choose to either integrate all of our Lights as a single final image, or integrate per session (i.e. you’ll end up with a 30 second integrated image, a 60 second integrated image, etc) which is great if you want to do your HDR processing/combining in, say, Photoshop rather than APP. Or you can choose to do all the above – perform each integration independently AND also create a final one. Obviously all the above just follows the same rules and the more integrations you choose, the longer it just takes. For my M42 I just chose “integrate all” which just produces a single integrated image.
- I select the following options here for my M42
- weights: quality
- integrate: average
- local normalization correction
- LNC degree 4th degree LNC
- LNC iterations 3
Note that I’d start off by not selecting LNC and seeing how the output performs. If you’re integrating from multiple sessions from multiple cameras or nights this will help normalise these images with a view to even all the images out. The result should be a more consistent looking integrated image.
- You shouldn’t need to select MBB – use this if you’re combining images over multiple nights. This blends the image edges together. So if you have different positions over different nights for example, you don’t want some square artefacts running through your combined image is how I see this!
- I choose MAD winsor clip for outlier rejection. It’s very robust and will remove satellites and planes with great effect.
- Keep the rest defaults!
- Now press the integrate button, and go get more coffee. The next time APP notifies you will be your final stacked image. The length of time it takes to do this depends on how powerful your kit is and how many images you’re stacking. This could be 5 mins to couple of hours!
- Now once this is done, you’ll see a stacked image. If it doesn’t look correct, or you get strange blended colours that you know are a pain to process out, try adjusting the LNC options and pressing integrate again. Whilst this guide is complete, you need to start using common sense and judgement as to what works for your data. If you’re happy with it, and it may not be completely perfect, but you’ll be amazed at what you’re seeing. If you see some strange integrated lights for the background, APP has a great tool that you can use in the 9) TOOLS tab called “remove light pollution“. Press this button and the view changes, asking you to draw at least 4 boxes around the light polluted areas. If you have some greys or reds that you know should not be there, draw boxes – big and small covering most of these colours then click calculate. You’ll see the image change. Be careful not to choose your nebula part of the image for example but experiment until you get a fairly good view. Don’t worry about there still being some artefacts, we can remove the rest in Photoshop – but you want to have APP do as much work as you can at this stage. Click OK & Save and give it a meaningful file name, or keep the defaults.
- When you return back to the main screen, in the file list scroll down slightly to the Other/Processed image that is now in there and double click on it.
- Feel free to use the drop down on the right next to the red question mark. This has some default options that affect some of the image and background, to save you nudging some of the strange named sliders! I tend to choose 15% BG, 3 sigma, 0,0% base as it brings out some nice contrast, but not too much.
- When you’re done with APP and your image, it’s time to save it as a 16-Bit TIFF image that we will then use to load into Photoshop. So, ensuring “stretch” is still selected (it’s a default) click the save button that’s under the histogram and choose TIFF as the type, 16-Bit as the bit value and I choose Adobe v4.0 as the colour type. Give it a meaningful file name – I put as much information into this like target name camera used, scope used, number of frames, and filters, and date taken so you can find it again on a later date.
Congratulations! You’ve just integrated your image in APP ready for us to tease out more details and fix background contrast and any additional noise in Photoshop.
Post-Processing in Adobe CC
I use a couple of plug-ins in Photoshop that are highly recommended
- I use Lightroom for some edits, and Photoshop for some astro editing plugins, so the first thing I do is load up Lightroom and select Import.
- I browse to the TIFF file I created in APP and import.
- Double clicking on this should load it up on screen. I then go to the Develop tab.
- I make a small crop just in case there are some artefacts at the edge, then I immediately load this file up in Photoshop by going to Photo… Edit In… Edit In Adobe Photoshop CC 2019…
- Select Edit A Copy With Lightroom Adjustments.
- This will load it into Photoshop where I now basically follow the instructions from the book “Guide To Photoshop Astrophotography Image Processing” by Dave Eagle FRAS. I wouldn’t feel comfortable repeating the steps here for obvious reasons, however as a high level:
- I first apply the filter AstroFlat Pro. Adjust the settings appropriately to remove some gradients or noise by adjusting the sliders, with the aim of keeping the nebula bright and detailed.
- I then run HLVG and select the Strong option to remove any Green from my image. As the tool suggests, experiment with what looks right in your image.
- Duplicate your background (default) layer as a luminosity layer
- Increase the hue slightly of the background layer
- Apply some touches to the curve of the luminosity layer
- Flatten image once done
- I then close this image that then sends it back to Lightroom for final edits.
- I prefer to make the rest of the edits in Lightroom from this point – for those familiar with Photoshop and prefer that, you will achieve the same results – but I find the process too advanced with far more options! So in Lightroom:
- I adjust the Clarity slider up a little, say 20-30. This brings out some detail, but also some noise
- I adjust the Dehaze option slightly – moving to the left will increase brightness and bring out more nebula overall (haze) and to the right will dim (dehaze) which will remove detail from the nebula, but the core may show up better.
- I add a touch of Vibrance and Saturation to my images – but be careful. You can easily over do things here, so be classy!
- Under the detail section you can do some noise reduction. Sliding Luminance option to the right is great for removing the noise from the background but at the expense of some smoothing of your nebula. So experiment. I try with a value of 50 and go from there. I find a careful balance of sliding this with the Clarity slider gives a good balance.
- At this point you’ve got a good idea of how well your image is – and you’re now doing the very small final edits. With M42, you want detail in the nebula, but don’t want to blow the bright core. I find that reducing the Whites and Highlights are useful here – but try to balance this with not taking too much away from the nebula.
- Once you’re happy with the results, you’re done I export this to Flickr using the Library tab again, or export this as a JPEG to the hard drive ready for uploading to my collection.
Congratulations, you’re ready to tackle another deep space target!