My attempts at astro imaging

Posts tagged “pixinsight

Season 2015/2016 Orion

I say it every year – but you have got to image M42 every season, it’s such an amazing object and really lens itself to imaging.  This is not my best attempt, the weather this season has bee terrible, this was imaged during a nearly full moon but it was so long since I had the scope out that I had to take mu chances!  It’s just under an hour in 5 min subs, back on the 8″ F4 Newt again (I love this instrument, just pulls the photons in!). Processed in Pixinsight – DBE, a little HDR Wavelets to bring out the core and the slightest of touches of LHE…..



Pixinsight Tutorials

Just found this site –

Got links to some great Pixinsight tutorials 🙂


My second foray into HaRGB (See the earlier M1 image). I took my Ha image from late December and added it to the RGB I took a couple of nights ago. Not sure if it is overdone as the Ha was a bit oversharpened and I had to do a lot of noise reduction on it before adding it in but here it is anyway! I think it looks kind of posterised when viewed at 100% so best not to!


ATrois Wavelets/K-Sigma noise reduction in Pixinsight

I always wished that noise reduction could be applied at the linear stage, afterall why go to the bother of stretching all the noise only to then try and get rid of it? Better to remove the noise at the linear stage then it is not included in the stretched image at all!

Pixinsight of course, has the answer in the form of the K-Sigma noise reduction in the ATrois Wavelets function. Open up ATrois wavelets on an image that has only a screen transfer stretch applied (This stretches the displayed image but the underlying image is still in its linear form) and click the K-Sigma noise reduction and marvel at the dissapearance of the noise!  You can also apply some sharpening by increasing the detail layer bias by small amounts. Here is a before and after image, it’s from a 30 second exposure of the Orion Nebula (The sample is a tiny portion of the full image).





Here is the setting you want to enable:


Noise Reduction at this level also solves the patchy noise reduction artefacts that NR functions like ACNDR produce, the image below has been stretched then had ACNDR applied, it’s quite tricky to show in a low res shot but hopefully it shows some very dark, faint blobs, which are not present in the images above when K-Sigma NR has been applied.


First post in 2012 – IC410

This was actually taken before the end of the year but I never got round to processing it – it is very faint and could do with more time on target but you make do with what you have eh?  Its around 3hrs in 15min subs, with my Altair 8″ F4 Newt and QHY8, guiding was with an OAG and Lodestar. All stacking and processing done in Pixinsight.

So, what is IC410????  It’s an area of glowing hydrogen gas, the reaosn it’s glowing – there is a cluster (NGC1893) that is exciting the gas with the energy from it’s hot young stars.  IC410 is in the constellation of Auriga (The charioteer) and is nice and high in the sky at this time of year making it a perfect winter target. The most interesting thing in the image is the ‘Tadpoles’, these objects are around 10 light years long and are created by the stellar winds and radiation from the cluster – the tails point directly away from the clusters central star.


The Crab Nebula in Ha and RGBHa

This is my first proper attempt at Ha imaging, I have had the filter since March but have never really got round to giving it a good workout! Last Tuesday I fitted it again and had another go. With the Ha filter in place things get a bit trickier……. I needed to refocus but the Ha filter cuts down on the light from stars making this a lot harder – I use plate solving to centre on my targets, this involves taking an exposure, the computer then looks at the various patterns of stars and matches them to known patterns, thereby solving where the ‘scope is pointing. As the Ha filter cuts down on the light reaching the camera, I had to inrease this from 5 seconds to 60.  Next problem is focusing, same issue as before, as stars are so much fainter it took about 10 mins to focus as my exposures for fucusing had to increase as well. Finally, as I use an off axis guider, it is much trickier to find a guide star with an Ha filter in place, luckily this time I found a guide star around M1 but I have an adapter so that I can use a standard finder scope to guide with but this then introduces flexture between the guide scope and the imaging scope so I was very happy to see guide stars pop up on the guide screen with the Ha filter in place.

So, M1, see the earlier image for details of what this target is –

The image below is the Ha image, composed of around 3 hrs in 5 minute subs and the same in 15 min subs, combined in Pixinsight, I also got deconvolution working well on this image to really sharpen it up.

Here is the same image bit this time I added the Ha data to the original RGB image as a Luminance layer


Here’s one thats been sitting on the hard drive for a couple of weeks – NGC891, I took it from the back garden observatory, it’s 17x10min subs,processed in Pixinsight and Photoshop (Setup was the usual, Altair F4Newt, QHY8 and Lodestar/OAG for guiding). Was a bit of a pig to process as I seemed to have a red halo round the middle third of the frame, I have seen this before and it appears to be some of the red light in the observatory leaking on to the CCD Sensor of the QHY8.

NGC891 is an edge on galaxy in Andromeda, it’s around 30 million light years away! Hubble has also had a look at this galaxy, it got a little closer in…….

Get the Pixinsight ‘creamy’ background look

I find a lot of pixinsight images have a night smooth, almost ‘creamy’ background to them – I found out how to create this!

Select the background, open the convolution tool and use that to blur the background slightly…


I may have a went as little too far in the examples below, the effect should be subtle but this is to show the difference it makes, also this is a 100% zoom, it will be less obvious when the image is viewed at normal resolution.


Pixinsight Deconvolution tutorial

I have tried for years to get deconvolution working in various astro and non astro processing apps and never had ANY success, until now! After much reading of the docs and other tutorials I finally have a process that seems to work with my images…..

Deconvolution, what is it?  Think about it for a minute, if we could work out all the errors in our optics, and tracking, and atmosphere, we could apply the inverse of this to get a perfect image… Deconvolution attempts to do just that, the results do vary and the settings are VERY sensitive, but give it a try as I do think the results are worth the little bit of time it takes to get it right.

So, my method…. First up, this needs to be applied at the LINEAR stage

1. Create a synthetic Point Spread Function  – choose the Dynamic PSF Function from the Process menu

2. With your image, and a screen stretch only (remember this is to be applied at the Linear stage), double click on some stars to add them to the PSF function, say around 20, spread all over the image

3. click the Export Synthetic PSF button (Looks like a little ball)

4. Now, you want to create a star mask so…. apply the screen stretch to the histogram trasfer function then run the create star mask function, THEN…. undo the previous 2 steps so you are back to a linear image with a seperate star mask.

5. Open the Deconvolution process, enable de-ringing and local de-ringing, set the Local Support to the star mask you created previously

6. Now the tricky bit – Global Dark value in de-ringing, I usually scroll about the dark areas of the image and find the lowest values, then set the Global Dark value just lower than this, it does need a play around with but thats the good thing about Pixinsight and previews!

7. Leave everything else at default, other than the Iterations, I usually start at 10 and watch the preview, watch for ringing/dark areas starting to appear around the protected stars then back it off a few iterations


Leo Triplet

This is the Leo Triplet – A collection of 3 galaxies in the, wait for it……. Leo Constellation!  These are around 35 Million light years distanct from us, so the light that my little telescope gathered had been travelling accross space, interacting with nothing but the odd exotic particle or piece of dust since the first apes started appearing on planet Earth!

This image is 17 X 10 min subs, with the usual setup – Altair Astro F4 Newt/QHY8, processing done in Pixinsight (Atrois Wavelets for the galaxy core processing)