Xiga

14 minutes ago, Felias said:

I have made it work! The problem seems to be that I was copying the background from a layer of my previous processing, into a new file. PS does not seem to recognise it as the background even if you put it at the bottom and lock it. If you open it from a saved file and just add the starless image as a layer, it works! And yes, the curves were there to the right, as usual, I don't know why I didn't see them before. 

Very useful tool indeed, thank you!

Brilliant, glad it's working now! And thanks for the feedback too. I'd like to fix that, although i'm not quite sure how tbh. If i find a way to make it a bit more robust i'll post up an updated version. 

ps - Care to share a before and after pic of the star reduction? 

Hmm, i'll have a play around with it myself later and see if i can get a similar error message. Although i'm struggling to think of what would prevent PS from duplicating a layer. 

Yes, the action stops at the 'Curves' stage and a message box is supposed to pop-up with some notes. Also, the Curve is a Curves Adjustment Layer, so the settings should be on the right hand side panel somewhere, where you can adjust it yourself. If you find it's a bit high then just bring down the curve a bit and then press Play one more tie on the Action to finish it. 

Thanks for the feedback so far guys! 

7 minutes ago, Felias said:

Thank you! This could be very useful for us PS users.... but I couldn't make it work. It keeps telling me "The command 'duplicate' is not currently available" , "The command 'invert' is not currently available", and "The command 'select' is not currently available". Any idea what's going on? I'm using the last version, PS 2022,and I'm quite sure that I set the layers as instructed. 🤔

 

Hmm, that is odd. Why is the Crop tool visible in the picture above? 

Did you have anything copied to the clipboard before you ran the action? 

9 hours ago, Laurieast said:

A Re-boot fixed it, or because I did a search within CS6 for atn, found yours clicked it and it appeared in PS, who knows!

Now to try it 😀

Edit: Works a treat, used your image above. Great!!

Thanks for the action!

Excellent. Glad you got it working! 🙂 

11 minutes ago, Laurieast said:

Yes, I have run the wrong action, sorry! 🤦‍♂️

But I  am not seeing your action in the actions window, I put it in the plugins folder. I expect I'm doing something completely wrong. Where should it go? 

Thanks!

Did you restart Photoshop after copying the action into the plugins folder? I'm pretty sure you need to close it down and reopen.

Also, if you're not sure which exact folder to put it in, just find the folder where your existing actions are saved and put it in there (sorry, I'm at work so I can't check the exact folder on my own version until later). 

30 minutes ago, Shibby said:

Just tried this and it's worked quite nicely for me, I think! I've only given it a very quick test with an existing, fully processed, fairly poor image. My results are below.

Thanks very much for the action!

 

Very good Lewis! I also think using it near the end of a workflow is the way to go. 

Nice Cocoon btw! 👌

1 hour ago, Laurieast said:

It keeps telling me this?

Hi Laurence

I don't recognise the layers in your image above. It looks like you have selected a different Action above, called 'Star Mask'. Are you sure you didn't accidentally run this action instead? 

Hi guys

After seeing Bill's amazing new method for doing star reduction using Pixel Math in PixInsight, i immediately knew i wanted to make a Photoshop Action that effectively did the same thing, because not everyone owns (or enjoys using!) Pixinsight. I've attached the PS action below. Please feel free to experiment with it and post your results so i can see how it works out for other data sets. Essentially, it is by far the best method of doing star reduction i've come across so far, as it uses a lesser-stretched version of the original image to reduce the stars, which means you won't end up with any artefacts at all. 

First things first, a big shout out to lukomatico on Youtube, which is where i first saw this. Regardless of whether or not you're a PI or PS user, I definitely recommend checking out the video below, where Bill explains the process very well. 

https://www.youtube.com/watch?v=rM3-yAcAbZc&t=178s&ab_channel=lukomatico

Ok so to make this work, there are a few caveats. Namely:

1. You will need a starless image. No big deal these days hopefully, now that we have StarnetV2 and StarXTerminator available. 

2. Make sure the original image (with stars) you are working from is a standalone file that's been flattened. In other words, don't just copy and paste a layer from another image into a new image and flatten. This seems to break the action. I will try and fix this if i can! 

3. Don't process the original image or starless image in any way before trying this. They must be the same image, apart from the stars being removed, for this to work best.

4. The original image with stars must be sitting on the bottom as the Background Layer, and the Starless layer must be sitting on top as 'Layer 1'. Don't rename the layer, and make sure both layers are visible (using the little Eye icon on each one). Basically, if you just open the original flattened image and paste on top the starless image, you should be good to go. Sorry, i don't really have any advanced Photoshop skills in terms of making actions that are impossible to break! But i can confirm if you follow these instructions to the letter, then they will definitely work! 🙂 

So you're workflow should look exactly as such:

Then just run the action, and it will do it's thing up until the point where we need to apply a Curves adjustment to set the amount of star reduction needed. The action will pause at this point, and you should now see a Curves Adjustment Layer on the right hand side panel with a suggested curve (just my own idea as a starting point, feel free to try different ones!). Once you're happy with the level of stretch, press Play again on the action to finish it. Note, a too aggressive curve may result in the stars having too harsh an edge to them, or even some halos around them. I strongly recommend a one-point curve to keep the stars as natural looking as possible. Also, in extreme cases you may get better results by running the whole thing more than once, with a less aggressive stretch. In which case, you will need a new starless image each time. So far, i've only tested this on a single Lum data set of my own, and one pass was enough for my own needs (see example below). Of course, depending on your own equipment YMMV. 

Here is the PS action:

Star Reduction Bill Blanshan Method.atn

And here is a quick before and after animation of a WIP Lum stack of the Cocoon Nebula i took last week with my 80ed, using the default settings above. Note, this is with a cheap SVBony Lum filter, which i really need to replace due to excessive bloating. No fancy processing here, just a DDP stretch straight out of APP, then a pass of Noise-XTerminator, before doing the star reduction. This is a 100% crop of a portion of the image. Another benefit of this method that i've found, is that it doesn't affect the background at all, so any faint dust that is residing in the background doesn't get reduced at all:

Really interested to see how it holds up with other data sets, and also with other curve adjustments, as the amount of star reduction one likes is often down to personal taste. 

ps - Not sure if this is actually the best place to post this, so Mods feel free to move if needs be. 

🙂 

Hi Michael

You're probably right. I'll likely continue with what I've got for the foreseeable future and will only change the OAG if I keep getting guide star lost messages for apparently no reason. 

It's just annoying really. After 5 years of routine data capture, most of which were 20 min subs without a single lost one, I now suddenly can't even do 2 min ones with any great confidence. I'm probably losing anything from 1/3 to 1/2 of a night at the moment. I can definitely improve things guiding wise, but I'm also pretty convinced now that the problem is mechanical in nature. 

Another suspect on the list is the stock focuser of the ed80. I'd tightened it down before, and did see some improvement, but not a massive amount. I've just tightened it again, so much so that the DC hand controller now can't move it! I can still move it manually though, so let's see how the next session goes. 

Are there any other obvious candidates, other than the focuser, that would lead to stars being bad post meridian flip, when they are fine pre-flip? 

1 hour ago, michael8554 said:

Hi Ciarán

 "although Phd2 still apparently manages to pick up close to 12 guide stars while doing multi-star guiding"

I take it you mean PHD2 is capable of choosing up to 12 stars.

"the aperture of the OAG stalk is actually really small. It's only a circle of about 5 or 5.5mm in diameter,"

Your ASI290mm has a sensor 5.6mm x 3.2mm, so a pretty good match for the stalk, I can't see any need for a different OAG.

Is the guidecam well focused, what HFD do you typically get ?

What exposure are you using ?

You could post a PHD2 GuideLog that includes the "Star Lost" situations.

Michael

 

 

 

Hi Michael

Thanks for responding. From memory, i think Phd2 usually reports something like '9 / 12' (although the numbers can and do change) which i've always just taken to be the number of stars it is using for guiding out of the number it has picked up. The ASI290mm has a diagonal of ~6.45mm, which is about 1mm more than the diameter of the stalk, so i think i'm getting a good bit of vignetting in all 4 corners. I also suspect, that Phd2 is selecting some stars that are on the periphery of the central area, and that might be why it occasionally loses them. I should point out, it's not a problem that happens an awful lot, just every now and then, but i'd still like to get it sorted if i could. I think on my last session i was getting a HFD of around 3 using 2s exposuresso it should be pretty well focused. I should also point out, i use the main camera (Qhy268m) for doing Polar Alignment in Sharpcap, but if i try using the ASI290mm then Sharpcap can't solve the image. I find that weird, as its' such a sensitive camera, but maybe it's because the FOV is just too small? 

The reason for all of this, is that ever since i got my mount tuned i've been having real difficulties with data capture. I just can't seem to get a full night of subs with round stars on both sides of the meridian. After much hair-pulling, i deduced that the problem was differential flexure, so i ditched the Guide Scope in favour of an OAG. I'm still having problems though, but weirdly it seems to quite random in nature. I had the scope out 3 nights last week, and each time i had varying levels of success on both sides of the meridian. The worst night meant that not a single sub was useable after the meridian flip, even though the reverse was true only the night before. It was around this time i had the guide star being lost issue, so the next thing on my list is to just try and help Phd2 as much as possible, and i think that means an OAG with a bigger (ie rectangular) stalk aperture. 

I really should post up some Phd2 logs, just like you said. The mount is so sticky that finding optimal balance is next to impossible, so there's always going to be a good chance the guiding will be affected by that. Even so, the guiding was so bad on occasion there has to be something else major going on as well. 

2 hours ago, almcl said:

Hi Ciarán

I have the ZWO V2 OAG and use it with an ASI 120 MM, but not sure I have understood which dimension you need (just a bit slow this morning).

If it's the one shown in the image below, then the thread depth is 5.19 mm.  If that isn't what you need, let me know (perhaps by noting on the image) which one?

Thanks so much for this mate! That's exactly what i needed, and it does look like the tread will just be long enough. 🙂 

Hi guys

Recently i've had some situations where Phd2 has been losing the guide star on me due to low SNR, even though the weather has been completely clear. Visually, I also can't see that many guide stars on the screen (in both Phd2 or Sharpcap), although Phd2 still apparently manages to pick up close to 12 guide stars while doing multi-star guiding. I found this very weird, as I am using an Asi290mm, which is very sensitive, and i'm even binning it x2 as well. Note, i'm using it with an OVL OAG on my SW 80ed. 

Shortly after, I then realised that the aperture of the OAG stalk is actually really small. It's only a circle of about 5 or 5.5mm in diameter, which i think is causing a lot of vignetting of the Fov in Phd2, and i think this might have something to do with the issues i've been having with guide stars. One option i could try, is filing away the stalk aperture to make it more rectangular. The problem is, there is a teeny tiny grub screw holding the prism in place, and i can't see any way of removing it safely and getting it back on. I do have some very small allen keys, as small as 0.7mm, but none of them fit. I'd also have to find a way to then paint the inside of the stalk black again. Basically, that sounds like a lot of faff, especially for someone as averse to DIY as myself!

So i started looking at the ZWO OAG v2, as it's only 16.5mm wide. The OVL one i'm currently using is 16.3mm wide, but i am also using a 0.5mm M48 spacer on the scope side, so technically i can support up to 16.8mm wide for a replacement OAG (note, i've no option to find any more backspace, i'm currently at my max for the SW FF/FR). 

Now the OVL OAG has a really long M48 female thread, which allows it to take the full M48 Male thread of the SW FF/FR. I've tried to find out the length of the same M48 Female thread on the scope-side of the ZWO v2 OAG but i can't find it anywhere online. If someone has this OAG, i would be extremely grateful if you could precisely measure the length of this thread. Or, to be more exact, the distance from the outside of the OAG to the point where contact would be made with the prism stalk. If it turns out to be shorter than the OVL, then what would people recommend? One idea i had, was to try and shave down 1 or 2mm from the SW FF/FR thread. It's unnecessarily long imho, i don't need it to be 5mm, 3mm would suffice for my needs. How easy would this be to do though? Could it be done safely with a simple tool (which one?) without this tool (ie me) running the risk of ruining the threads on the FF/FR? 

Very nice Olly. I especially like the star colours, mmm lovely.  Out of curiosity, what is it exactly you're not sure about? Could it be something to do with the Tak Lum not being quite enough to handle the massive amount of signal coming from the RASA? 

Very nice Dan! That's a great M57 you've got there. 

I'm only looking on my phone, but I like the detail of the 2nd version and the colour of the 1st version. Maybe try a simple mix of the two to see how it turns out?

Ps - Love France btw. We're currently on hols just outside Perpignan. Aller France! 🌞🇨🇵

Amazing work Olly. Nice and rich, and the stars here are top-notch. Starless processing (whether it be Starnet v2 or StarXterminator) really is a game-changer for RASA's. I recently gave NoiseXterminator a go myself for the first time, and i too was very impressed by it. 

Can i ask, how did you stitch the 2 panels together? Did you do it manually in PS by any chance? I think i can see a slightly warmer tone to the panel on the right, which could be down to that panel having a slightly different stretch to the one on the left, but i'm not sure. APP and P.I can stitch panels together at the linear stage, which makes processing mosaics as easy as single panels. 

Another thought Brendan. How quickly do you cool your 1600? I think most people choose to cool over something like 5 to 10 mins. If you do it too quickly there's maybe a chance it could lead to frosting on the cover glass. 

Next time you are out, I would suggest doing another test (an easy one). Don't immediately cool your camera on connecting it. Get yourself all set up and then capture a few subs without cooling. Then turn on the cooling and just keep capturing while it cools down. See if the first few subs look ok (apart from the higher level of thermal noise and hot pixels of course). 

Could the problem be due to fogging on the sensor cover glass? I know earlier versions of this camera didn't have a heated window. When you set up Brendan, how long does it roughly take until you actually start capturing subs? 

If it is, then the problem would show as getting progressively worse throughout the night. However, this data set was started when the sky was still relatively bright, so the fogging could have already set in by the time the sky brightness had levelled out. It might be a good idea to see a different data set shot earlier in the season during Astro Dark, and see if the first few subs look ok. 

Haven't had much opportunity for AP this year, so decided to dust off the processing cobwebs and have a crack at your data Rodd. And very nice data it is too! 

I used the full (90 sub) Ha stack and apart from using APP to remove some slight gradients and save some DDP stretches, everything was done in PS. The workflow was roughly as follows:

1. Saved a DDP stretch of both the Ha and Oiii, then ran Starnet2 on both, followed by a pass of NoiseXTerminator (first time using this, will definitely be buying once the trial is up). 

2. Made a Master Lum using both the Ha and Oiii above. The Oiii was added in Blend Mode Lighten, and then masked to only apply it selectively (otherwise it just made everything too flat looking). 

3. Added the Ha stars to the Master Lum, by stretching the linear Ha stack manually and adding it in Blend Mode Screen. 

4. Created a HOO blend in APP, and added as a Colour Layer in PS. 

5. A whole bunch of the usual adjustments, Levels, Curves, some use of the Camera Raw filter etc etc, too many to list (or remember!). 

If it were my own image, i would probably downscale it to 80% at the end of processing, but just decided to leave it full size. 

Thanks for sharing, and giving me the impetus to get back on the horse! 😃

Thanks Colm, Mike and Steve!  🙂

For anyone interested, here is what the Ha stack looked like, with zero processing other than gradient reduction in APP and saved as .jpg. 

3 hours ago, Padraic M said:

That's a super 101. It's not an easy target but the Ha makes a big difference. 

Thanks Padraic 👍

2 hours ago, tooth_dr said:

I've just been gathering data since before Christmas, havent got round to processing any of it yet.  Nine panel mosaic on Markarians, 12 panels on Orion, and a single panel on the Witch Head nebula.  I dont think I'll swap over to a longer focal length this year, so just continue adding to widefield galaxy data in Leo.

Wow Adam you have been busy, some big projects on the go there. Looking forward to seeing the finished articles! 

3 hours ago, tooth_dr said:

That’s a great M101 Ciarán. It’s looking clear tonight again so hoping to get our myself. 

Thanks Adam. Have you managed much imaging over these past few weeks? What's on the menu tonight? I'm going to struggle for NB targets now that i've finished this one. 

1 hour ago, tomato said:

Great image, lots of fainter regions visible and the Ha really complements it.👍

Thanks tomato. The Ha made a big difference once i added it in. 

1 hour ago, teoria_del_big_bang said:

Nice image, far better than any of my galaxy attempts.
Thanks for the great detail of the acquisition and processing 🙂 
How much detail does the Ha add ?

Steve

Thanks Steve. I did actually add a small percentage of the Ha as Luminance too, but was very careful to only apply it to certain areas in the core. It was subtle, but it definitely helped to boost the detail near the core. Then adding it to the Red channel in Lighten mode really made the Hii regions pop with colour. I thought the image was looking a bit anaemic before i added it, so i'm really glad i went to the bother of shooting it now. The conditions for the Ha were really not good at all, but the beauty of adding Ha to an LRGB image is that it doesn't need to be the best data ever, and you can even hit it with the proverbial processing hammer and it won't show through in the final image. I'm at work, but i'll post up an auto stretch of the Ha stack later on tonight when i get home. 

Since getting the new camera (Qhy268m) this time last year, i've only managed one proper image with it to date (Butterfly Nebula in NB), so i've been itching to have a go at shooting something in Broadband. I think this is my first full LRGB image so still getting to grips with it all. 

I have a pretty restrictive horizon all around (can't shoot anything that transits below 60 deg to the South) but M101 is well positioned so i opted for it during the clear spells we've been enjoying here recently in N.I. I captured the LRGB over 3 nights in Feb, but have been too busy to process it up to now. I'm glad i didn't, as it i gave me time to shoot some Ha this past weekend (during a full Moon and wind speeds approaching 20mph!). I was amazed to see so much detail in the Ha stack, and somehow pretty much all the subs ended up being keepers, despite the wind and subsequent dodgy-looking guide graph! 

I've been dabbling a bit with Pixinsight this past few weeks, but i'm not exactly loving it if i'm being honest. I still prefer to use APP for all the stacking and gradient reduction. I think i did use P.I to do deconvolution on the galaxy in the Master Lum stack, but apart from that i was struggling to have any joy with any other processes that i couldn't already do one way or another in PS. I'm not giving up on P.I just yet though, i'll persist with it and see if i start to come around to it as i get to spend more time with it. 

The processing in general i found to be quite challenging. Although i think this was mostly down to the limitations of the scope (the SW 80ED being only a doublet) and possibly because of the filters as well (cheap SVBony ones). I made a Synthetic Luminance from just the RGB subs, as well as a Master Lum from all the subs. Between the R,G,B and L, the star sizes differed by quite a lot (Blue and Lum were really quite bloated looking, but Green looked great). The Lum went nice and deep so i wanted to keep it for all the faint background stuff as well as the galaxy, but just not the stars, so in the end i used Starnet2 to remove the stars from the Master Lum, and then added back in just the stars from the Synthetic Lum stack. This made a big difference, but still some of the larger stars were a bit too big, so i used Starnet2 once more but this time i only replaced the handful of biggest stars with those from the Green channel. This gave me a decent base level Lum file to work from, and the processing became easier from here on. 

Full capture details:

SW 80ED; HEQ5-Pro, OAG with ZWO 290mm; Qhy268m; 2" SVBony LRGB filters; 2" Baader 3.5nm Ha. Mode 1, Gain 56, Offset 25, -10C. Bortle 5/6 skies. 

RGB - 30 x 180s each (4.5 Hrs)

Lum - 211 x 60s (3.5 Hrs)

Ha - 44 x 360s (4.4 Hrs) - Added to Red in Blend mode Lighten

Total - 12.4 Hrs

Stacked in APP, Processed mostly in PS. 

Thanks for looking and CS! 🙂

Full FOV:

Crop:

15 hours ago, GalaxyGael said:

M3 imaged under full moon, just about got enough subs before the moon rose over the houses on each of two evenings.

A mix of 55 x 60 s and 120 x 120 s subs.

Takahashi Epsilon 130D and ASI2600Mc Pro, Gain 0 at -10C

iOptron GEM45 with ZWO OAG/ASI290MM mini.

Stacked and processed in APP.

 

Lovely M3 Colm. 

Star colours have been very well managed. I'd say Gain 0 is working for you. 

On 15/03/2022 at 17:23, Iem1 said:

Hey guys,

Does anyone know how to preserve image quality when posting to facebook? I can post the image to instagram, no issues. If I post the exact same image to facebook, it appears fine on PC but gets ruined when viewed on mobile? 

I have resized the image to 2084 x 1389 and it is 3.06 MB PNG. This is what I upload to Facebook.

After I post this to facebook, this is what is actually posted (I downloaded this on my mobile from facebook and emailed it to myself):

Its a 42 Kb JPG 2048 x 1365, not sure if me downloading it affected anything, but it looks as it does here.

 

 

Any ideas? Thanks

Being a member of SGL, you are allowed to attach content (such as your lovely photo above) to your posts. I'm not sure what the total limit is per member, but it's A LOT. So there is a workaround if you want people to see the full res pic.

1. Click on your picture above. It will display in a window. Then click it again. It will then open in full res on a new tab. 

2. Now you can just share the link for this tab anywhere you like, and when the user opens it, they will see exactly the image above, in full res. 

🙂