Stub Mandrel

Here's one that did seem to work well in Pixinsight, but it clearly wasn't a suitable subject for background neutralisation or gradient removal...

Instead, I just tweaked the colour balance in PS and ran a round of star reduction, then denoised in Astra Image.

Today I revisited some of my initial attempts at Pixinsight while I was away in Photoshop. A possible disadvantage for PI was that I processed on my laptop, whose screen though good, is rather different and fussy about the angle it is viewed.

I found some very consistent differences:

PI seemed to really flog the data hard, pulling out faint nebulosity but at the expense of poor control over stars which bloat and poor control over noise. I suspect this is why star masks are so critical in PI (I haven't mastered them yet). It may also be why many suggest noise reduction on the linear data before combining it in PI.

PS gave much more natural, less noisy results, but with less faint detail.

I felt difference was more subtle and marked than simply PI stretching more aggressively than PS/DSS, but has to do with differences in curve shape with PI (and Sharpcap's preview) using a simple curve while DSS and PS encourage a non-linear S-shaped stretch.

PS has more sophisticated control over colour balance, although PI brings out faint and subtle colour better, especially when Oiii or Sii signals are weak.

In all three images, I ended up using a blend of the PI and PS images with PI mostly contributing colour and lifting faint nebulosity, while PS mostly contributed luminosity for much tighter stars and better definition in brighter nebulosity.

Noise reduction in PS is easier to use, but the jury is out on which is more effective.

One area I found a huge difference was that GradientXterminator is not just easier to use than DBE but gave consistently much, much smoother and better results. It also did a more accurate job of balancing background colour - I actually had to use it on some of the PI images whose backgrounds were severely mottled no matter how I applied DBE before they could provide an acceptable colour layer.

It's clear I have a lot to learn about getting the best out of PI, especially about controlling stretches and noise reduction. It seems using screen transfer function and transferring it to the histogram stretch is a crude way to proceed, but good for finding what is in the image. I also need to understand star masking and reduction. DBE seems hobbled by the default approach and using bigger control areas seems to work better but even then placement is critical, and it seems small variations can introduce artefact gradients that spoil the background.

I don't think it's fair to do side by side comparisons yet, but here are the first three 'best of each' images I've done.

Both my brother's and my house are given a Bortle 5 by CO.

I haven't seen his sky under astro-dark without the moon but before astro dark and without the moon it was almost as good as the best I have ever seen it here, just hints of the milky way. Most of his local street lights go off at 12 and they are incredibly well-directed LED lights.  The light in front of his house was shining between the two houses straight on my scope and I didn't even notice it for a few nights. The bit I could see lit up was not as bright as Jupiter, and the level of lighting on street doesn't seem much brighter than fullmoon, much better than here.

On 06/08/2020 at 16:53, wormix said:

Hey @Stub Mandrel - I’m currently wrestling with a similar design myself. 
 

Do you happen to remember what the thread size on the Finderscope is?  I’m currently just doing a bit of suck it and see - I’m guessing 52mm diameter, pitch 0.8 as my 51mm I just printed was too small and pitch was off

I had completely forgotten that!

I think that the thread is 0.7mm pitch as it appears to be the same as the T-thread.

If 0.7 doesn't fit, try making it 0.2mm undersize on diameter - or cropping the crest of the thread by that amount..

Thanks, but I really stuggle with video tutorials. I find step by steps like the Light Vortex ones bewtter, I use them like a reference book as I have a non-linear brain.

Back home at last, time to process the data with PS, compare the results and identify weak and strong points.

Last night's data, thrashed.... I will return to these when I get home, I'll process in PS, compare and see what's inexperience with PI and what's deficiency in the data.

You can see some of the colour is badly posterised.

Well the most useful thing about those tutorials is they tell you WHY not just WHAT which makes the lessons learned transferrable.

I'm still to get into masks but this Iris from a couple of nights ago doesn't seem too bad!

3 hours ago, Jkulin said:

Hi Neil,

As promised: -

Shawn is really very good as explaining the manner in which he works https://www.youtube.com/channel/UCIwWoJv_vi1US5zjyjRs_lg

His website is: - https://visibledark.ca/

I have also found  to be https://www.lightvortexastronomy.com/ excellent.

And finally Chuck gives you a rough and ready process tutorial that works: - https://www.youtube.com/channel/UCCllh0nUmlREEvoskaq9b3A

Hope that helps?

Thanks, The lIght Vortex tutorials look like what I've been searching for.

Some good points!

I didn't explore other calibration frames. I use bias for DSLR and dark flats for CMOS images.

2 hours ago, ollypenrice said:

I routinely use L flats for everything, without issue, ninety-odd percent of the time.

The difference between my RGB and narrowband filters is really noticeable. These images both use the same G-channel data. The first one uses a flat made with the Ha filter and the second with the R filter. Incidentally my L filter, a Baader, is pretty close for focus to the narrowband filters, rather than the RGB ones. I've noticed this problem several times before but never realised the cause. It may actually be that the amount of vignetting is different between filters, rather than a focus issue.

ZWO Green image stacked with Baader Ha flat:

ZWO green image stacked with ZWO Red flat:

In my experience, few things cause as much confusion and odd problems as flats, the extra set of frames you take to compensate for any defects in your imaging train like dust, vignetting or optical oddities.

Flats are images that capture these 'defects' without any otehr content, and when applied to stacked images they can make the most ugly of shadows or dust doughnuts disappear like magic.

This post was inspired by a presentation by @Whistlin Bob that covered the subject really well and made me realise why I was getting a particular defect in some of my images!

Flats are simply a set of images made using a plain background and exposed so as no part of the image is under or over exposed (so the histogram should be near the middle). With a DSLR the 'Av' mode will do the job reliably every time, with a CCD or COS camera you need to pay attention to the histogram display.

The target can be any smoothly illuminated surface, I have used teh following, all with success:

• A plain painted wall, illuminated by diffuse light.*
• A white t-shirt over the end of the scope pointed at the sky (away from the sun).
• An LED tracing panel.*
• The sky (as long as it hasn't got any clouds as these will focus and affect the background).

For the marked sources* I found it best to rotate the scope (or source) to avoid any gradient affecting the flat.

Witha DSLR I take 16 to 24 flats. With a dedicated astro camera, I take 64, vbeacuse it is easy to do so.

You can combine flats, DSS does this automatically the first time you use a set, then you can delete the originals and just use the master flat.

SharpCap has a special routine which takes flats for you. It always checks 'apply automatically' - I always make sure this is unchecked as you end up with (wrong) flats being unexpectedly applied to future sessions. It also allows you to apply a different flat (if soem dust appears during a session or something moves).

Which leads up to 'when should I make flats'?

Some people make them for every session. This is the 'gold standard' and is what i do when working with Sharpcap as it only takes a few minutes to get a good 'sky flat' while you are waiting for darkness.

Otehrwsise, it's possible to make flats before or after your session with the key thing being that you make no (or as little as possible) change to the setup.

If, like me, you leave camera and scope set up and are meticulous about cleaning dust from sensors and filters, then a set of flats could last you several sessions, even a month!

But sometimes it becaomes apparent that dust has got in and you need a new set of flats or master flat. I've learned that it is easiest in the long run to get theat master flat at the start of each session (or at the end if you use something like a panel). Flats taken the following day are fine - but only if nothing changes. The problem is moving scopes around can dislodge dust.

Optical changes. Many people create a flat for each filter they use. The presentation made me realise that filters do have an effect. I have a filter wheel and usually make my flats using the L filter, which is parfoical with my (Baader) narrowband filters. For a good while I have seen odd, circular artefacts on my (ZWO) RGB filtered images and it's finally dawned on me why. For my last few sessions I created a master flat using the Ha filter, which worked perfectly on all my narrowband images. The RGB ones showed a dark,circular artefact near the top left. As everything was still set up, I created a new flat using the red filter and an evenly illumineted blanket. This worked perfectly.

The lesson is that flats can work for multiple filters but only if they as parfocal, or nearly so.

I hope this is useful for those beginning with imaging and helps you get your head around how and when to use flats faster than I did!

I generated a lot of data last night, mostly RGB and my my flats were with the Ha filter... so I need to take a new set of flats before finishing thiose images.

The one narrowband image was the Crescent. I stacked in DSS then did a very basic STF stretch and combine in Pixinsight, background neutralised and then tried to use a mask for a final curves.

Boy, it's a foul user experience!

Here's my first result, which looks OK but feels like what Pixinsight wanted to give me and that I have far less control than usual. I'm sure there are ways and means. Can anypone point me at good step by step tuorials (not video ones, as even the simplest leave so much out or unexplained). I'm afrioad I went o Astra Image for NR... and I think it needs more curves/star reduction etc...

On 31/07/2020 at 14:06, BrendanC said:

@Stub Mandrel Yes, I am using a coma corrector. Thanks for posting those images, but I'm not sure what they're telling me? 

Your examples are all pretty much like the Baader example. This suggests you need to increase the spacing. Is it the Baader MPCC - this is much fussier about spacing than the skywatcher one.

It's not clear but there appears to be some tilt or the collimation may be out as well. Try adding a 1mm spacer (card will do for a trial) and recollimating and see if it improves.

Once you get the spacing right you can get a permanent spacer (or look on amazon they have some spacer sets that are cheap).

This is only my second mosaic, so it's been a bit of a struggle, especially as poor fits left the western part with a quite noticeable gradient. To my surprise the Sii signal was almost as strong as the Ha, so I used selective colour to make the non-Oiii parts a bit redder, they wanted to be yellowy brown. The TIFF version of this is 137Mb!!

2 hours ago, vlaiv said:

My approach is to remove background and then stitch images while data is still linear. Once you have stitched larger versions - then proceed as usual.

Thanks Vlaiv I had another go using more or less that approach, although they had all be given the same stretch in Fits Liberator.

I found this, which is rather unfair on the MPCC as increasing the back focus should bring it to the same as the others (which are all possibly spaced a bit too far):

20 hours ago, BrendanC said:

Hi all,

I have a problem with my 130PDS (I think - or it could be the mount, the software, or any number of things, but it's probably the optics).

I've posted elsewhere about this (see https://stargazerslounge.com/topic/358322-what-on-earth-has-happened-here/) but was advised that the resident 130PDS experts (that's you!) might be able to help. So, apologies if this counts as a double post, I'm just doing what I was told...!

So, the problem is, I'm getting subs that look like this (zooming in to show the details at the corners and middle - these are all from one, unprocessed sub, at actual size):

Top left

 

Bottom left

 

Bottom right

 

Top right

 

And finally, middle

 

There are several things going on, as you can see. The directions of the blurring changes. Some stars are out of focus. And none of them look at all right to me.

A previous similar problem seemed to be fixed by making the mount east-heavy, but as you can see, it's far from fixed the problem.

The latest thinking is that it could be the focuser impinging on the image?  This is what someone has found when zooming in:

For comparison, here's a sub I took a couple of months ago, exact same setup except that I've disassembled the kit a couple of times during bad weather, I'm now using EQMOD directly connected to my mount, and I'm also trying to improve my polar alignment by using Sharpcap (click image to see larger version then save that to see original, albeit jpg). The problem doesn't seem to be he

You can see my kit list in my signature.

I'm tearing my hair out (or at least what's left of it) to figure this one out. If anyone has any ideas, I'd really appreciate it.

Thanks, Brendan

 

Those shapes are definitely coma, shaped like the reflection of light slanting into a mug of tea! Exacerbated by poor collimation which is probably why they all point pretty much the same way..

Have you got a coma corrector?

22 hours ago, RolandKol said:

Location Location Location....

My Bortle 8-9, probably shine much brighter even if a slight thin cloud line is in the way.

I can see that being an issue.

Planets and the Moon are always a good standby when conditions are tricky.

I've got some decent data for the two ends of the veil nebula. As you can see my first crude attempt is a bit iffy, although I can align it OK.

What's the best way to proceed?  I'm cautious of spending ages on one approach when I could have taken an easier route.

How much processing should I do before I combine the images? Should I combine as individual layers or combined into RGB images?

I have Ha, Sii and Oiii for each half, all the same amount of data and reasonably consistent ( more moon and better flats for the second session).

Here's my crude first attempt:

Should have a 12V power adaptor.

10 minutes ago, DBushell1 said:

Just saw that delivery for an HEQ5 Pro is 40-60 working days! Crikey! Is that a normal lead time, or is it just a result of Covid.....?

Deliveries from China have been shot to hell by Covid, so most importers are working on long lead times.

They are generally rather frustrated as demand for hobby equipment is booming...

2 hours ago, RolandKol said:

On another hand, - if you are UK based, - you probably noticed, we usually have around 20 - 25 clear nights per year. 

That made me curious, here are my counts since January 2014. I only really started regularly in 2015, and many of my early attempts went undated.

2014  - 3 (actually quite a few more)

2015 - 40

2016 - 43

2017 - 42

2018 - 66 (12 nights in June alone!)

2019 - 41

2020 - 18 to date

There's  tendency for me to push my luck if things have been thin (so I might  a short hour or two's window for some planetary or lunar) , and to run out of steam if there are long periods of good conditions. My only blank month is February 2017.

But on the whole, a consistent 40 days a year of actual imaging (not opportunities).

Try downloading CCDinspector

Two completely new versions, right from the 'bottom up':

9 minutes ago, MarkAR said:

I came across something like this, I think it was one of Adam Blocks' videos either on YouTube or his website.

There's a lot of threads here and on cloudy nights but no obvious resolution. I'm going to reduce the dynamic range of a flat by half and see what happens.

I've tried a star reduced version and done another round of selective colour (followed by reducing saturation). These are worthy experiments as I want to add the eastern half as a mosaic, then I will have to reprocess the colour of the whole image.

But better do some work for a bit...

10 hours ago, MarkAR said:

Some interesting colour combinations, I'm tending to like the third one. They all do seem a bit saturated though.

Also a hell of a lot of stars in the area, star reduction often works wonders in bringing forward nebulosity in these situations.

Other than stretching, the top one is pretty much as it fell out of stacking. For the others I've done selective colour but nothing else.

I'm not happy with the flats, I'm getting 'ASI1600 flat overcorrection syndrome' - described by many and it seems never solved. If today goes according to plan I will take a new set as I haven't touched the setup.