Xiga

2 hours ago, Richard_ said:

I followed the thread you posted and the first thing I thought about was sensor temperature. This is the key parameter which will drive your dark current when keeping your gain fixed.

In the past, I've had weird sensor temperature readouts when using the QHY268M, but this was resolved when choosing the driver (ASCOM versus native).

In terms of diagnostics, it wouldn't do any harm to check your power situation. How do you power you system? Are you using any power distribution systems (eg Pegasus box, ASI Air)? Are you using good quality power cables with solid connections? Before raising a ticket with QHY (if it comes to it) I would try hooking the camera up with QHY's EZCAP software to see if any errors come up.

If you still want to use the data, all is not lost. If you're a user of pixinsight, you can utilise the Cosmetic Correction tool in your workflow to eliminate rogue hot pixels. Try different sigma values for the hot a cold pixels and check to see if they equally eliminate your hot pixels in the best and worst subs. In my experience with my QHY268M at Gain 56, Offest 25, cooling temperature -5C°, using sigma values between 2 and 3 do a good job at eliminating any hot pixels not removed via dark & bias calibration.

Hi Richard

I've double-checked my power connections and am happy that they are all sound (i use AC power) but i guess something must have come loose somewhere (although i can't figure out how!). I'll just keep a closer eye on the subs over the next few sessions to see if everything is ok. If not, i'll be back on here no doubt!

The data is still useable for sure. It's just typical though, that on my best ever Lum session (7 hrs with great guiding) that this happened! The first half were affected, but at least the 2nd half were ok. I've stacked the data already, just to see, and it stacks fine, with none of the hot pixels showing. No need for any cosmetic correction (which APP can do, if i needed it). The dithering and sigma stacking takes care of all the (many!) hot pixels, but of course the resulting stack is no doubt a fair bit noisier than it would have otherwise been. There's always something in this hobby isn't there?! 

3 hours ago, vlaiv said:

I don't think meridian flip un-glitched the software

I'm more inclined that meridian flip un-glitched dodgy power connector for 12V to camera cooling. Without it - peltier won't work and although software might think it is cooling the camera - without power it won't.

Hmm, now that you say it Vlaiv, that does sound like a more logical conclusion! 

I run everything from inside the house, via an extension cord plugged into an RCD (for extra protection) which goes to a weatherproof box, which houses all the power bricks. I do have a 12v extension cable for the camera, as the one it came with is nowhere near long enough, but it has a nice snug fit and doesn't seem loose of flimsy. I've never had any power issues ever since i moved to AC many years ago (best thing i ever did tbh) but i guess something must have come loose somewhere, somehow. I've opened up the weatherproof box and checked all the connections just to be sure, but nothing was loose. I guess i'll just have to keep an eye on the next few sessions and see if it happens again. 

FWIW on the new darks i took - on the first 20 (uncooled), they started out all the same, ie noisy, then weirdly on subs 8 to 14 they suddenly became a lot cleaner, then on subs 15 to 19 they returned to being noisy again, but then sub 20 was clean, similar to subs 8 to 14. Very strange. But on the cooled ones (subs 21 to 40) they all looked the same, so i'm happy that the cooling worked. I won't bother posting them, unless anyone wants to see them! 

I looked at all 210 subs a bit more closely, and i've found a weird correlation between when the mount did it's meridian flip, and a decrease in noise. 

The mount was due to do it's meridian flip quite early, but because the guiding was going so well i just decided to let it continue on past the meridian. I've been having mount issues and i wasn't confident that guiding would continue as well on the other side, so i let it go on for maybe 90 mins past the meridian before i told it to flip. 

The first 99 subs were done on one side, and they all exhibit noise that suggests cooling wasn't fully working (despite SGP saying it was ). The mount then flipped, and then from subs 100 to about 115 there is a gradual but very noticeable decrease in noise with each passing sub, before it eventually levels off and the subs all look nice and clean as expected. So somehow, the meridian flip un-glitched the software and the cooling suddenly started working fully again, is my best guess so far! Note, i checked the FITS header for most of the subs, and the temp does fluctuate between -4.9 and -5.0 in all of them. I would have felt better if they all showed the same value, but they don't, which only confuses things even more, lol. 

ps - I've now taken some new darks with and without cooling. I'll post those next. 

Hi guys 

I have a Qhy268m and noticed a weird problem in some lights recently. I started a thread in the Imaging discussion forum, but think it's better in here. 

Link below. Interested to hear other owners thoughts, and the best way to test/analyse whether something is actual wrong with the camera.

Cheers. 

Thanks for the response guys. Onikkinen that's a good idea, I'll post a link to this thread in the Qhy268 thread 👍

I think you might be on to something Oddsocks. I use AC power for everything so I don't think it was a power issue, but I agree it does look like the cooling wasn't working properly for whatever reason. I took a long break (over a year) from astro, and this was only my 2nd session getting back into it. I looked over the subs from the 1st session from a couple of weeks earlier, and those subs were also affected. I also checked some subs from a year ago and they looked fine. 

I'm not convinced just yet that the cooling is booked, so i think I'm going to re-install the firmware and see if that helps. I think it's still running the version from when the camera was first released. 

Forecast looks awful here for the next while, so I might also do a test of shooting some darks with the cooler off, followed by some at -5C, and see if they look as expected. 

Cheers guys.

I was out the other night shooting some Lum on Vdb 152 with my Qhy268m. The rig captured around 7 hrs in total, but when i came to inspect the data, i noticed something very strange. 

I use APP for calibration, and have never, ever had any issues calibrating my data. I calibrate with Masters for Darks, Flats, Dark Flats, and a BPM. But when i inspected the calibrate subs, i noticed lots and lots of what look like hot pixels remaining. I think a small number of them are in fact hot pixels (i probably need to take new Darks and a new BPM, it's been 18 months since i last took some) but the majority can't be hot pixels as they are clearly moving frame by frame. What's even weirder, is that the number of them start out quite low, then as the sub count increases, they grow in number until there are absolutely loads of them, and then they start to decrease in number, such that by the end of the night there are not many at all. 

I'm totally stumped as to what this could be. The camera was only running at -5C and i've checked the FITS data for a number of subs and they all show -5 (or -4.9) so i don't think it's a cooler issue. 

I've attached a few raw subs below, if someone could kindly take a look. I've included a few from the early part of the session, a few from the middle, and a few from the end. I've also attached the calibration masters as well. 

Here's a screenshot  at 100% zoom of how Sub 60 looks like after calibration!

I haven't bothered stacking everything yet. I suspect that the dithering and rejection algorithm will clean things up a lot, but i'd still very much like to sort this out as clearly something isn't right. 

vdb 152_120sec_No_ISO_filter0_L_3_frame1.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame2.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame3.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame60.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame61.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame62.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame197.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame198.fit

vdb 152_120sec_No_ISO_filter0_L_3_frame199.fit

MF-IG_56.0-E_4.0s-QHYCCD-Cameras2-Capture-6252x4176-.fits

BPM-QHYCCD-Cameras2-Capture-6252x4176.fits

MD-IG_56.0-E120.0s-QHYCCD-Cameras2-Capture-6252x4176.fits

vdb 152_120sec_No_ISO_filter0_L_3_frame63.fit

It must be a full year now since i last captured or processed anything, but I've been trying to get back in the Astro saddle lately. Of course, the awful weather we've been having here in N. Ireland certainly hasn't helped, but when i saw this i thought it would be a good opportunity to get back to processing again. 

Workflow for this was actually fairly simple. I don't have P.I, but the stars really needed BXT so i used trials for both P.I and BXT to correct them. APP was used for gradient reduction (which i did find was needed) as well as doing the initial stretching, before manually combining as SHO in Photoshop. Lots more stretching then done in PS. Star-Xterminator and Noise-Xterminator both used along the way (in PS, not P.I, just because i prefer to do as much as i can in PS) and finally a bit of high pass filtering. Also left some green in too, as i personally prefer it on SHO images. 

Decided to really push it, given the size of the data set. 

Thanks for sharing! 

42 minutes ago, tooth_dr said:

Thanks for the feedback Ciarán @Xiga
I will run those changes you have suggested as they do improve the image!

And thanks again for your help, meeting up motivated me and gave me the necessary tools to improve 🙏

No probs Adam, more than happy to help! And it's great to see your processing improving all the time 👍

7 hours ago, Whistlin Bob said:

With a few other members of my local astro club I'm in a syndicate at Roboscopes. December has been pretty sparse, so we've not seen much come through. However, we were gifted a rather chunky 50hr data set on the starfield around the Cocoon nebula. Now, most of my images consist of a few hours of data, grabbed between clouds in light polluted Staffordshire and I like to think that I've developed a bit of proficiency in compensating for the conditions I'm imaging in. Working with a rich gorgeous data set turns out to be a different beast entirely. Let me show you what I mean- here's 35 hours of broadband data with a simple stretch on it:

This took no skill at all on my part: it's just a calibration and stretch- most of my effort was devoted to hanging my jaw open at the number of stars in it. Once you get past the staggering richness of the star field though, it's obvious that there's a heap of structure here, so I set Star Xterminator to work on both this and the Ha data. Somehow, in the gaps between all of those stars it managed to pull out this structure:

I always have a lot of trouble capturing dust at home, so this had me very excited. I also have to take my hat off to the software- that point and click star removal is one clever bit of computing.  I didn't want  to lose sight of all of this, so I dimmed the stars down as I combined the stars back in, so as not to lose the structure of the nebulosity

 

I'll hold my hands up at this point and admit that I'm very much in the pretty pictures school of Astrophotography, I'm not too concerned with making my images appear natural. My question is really to find out whether others also find this attractive, or would you prefer it toned down a bit- perhaps something closer to the original RGB?

The starless image is really quite striking. In particular the Ha 😲

7 hours ago, Knight of Clear Skies said:

What might work nicely is a slow crossfade between the top image and your final one.

I had the exact same idea when I imaged the cocoon a few months back 🙂

Wowsers, this one is a little bit special Adam. I've been taking a break from Astro recently so i haven't gone anywhere near the computer in a good couple of months, but i simply had to boot it up to take a proper look at this. 

Fantastic work, you should be really pleased with this. I definitely wouldn't say you went overboard with pushing the dust, it's pretty much spot on in my book. I also don't think it's too dark either. In terms of brightness, t's very close to how i would have presented it myself tbh. The histogram is very well controlled too, with no clipping on either end. 

I took the jpg into PS and played around with it for 15 mins (couldn't help myself!). I agree with the points above about removing the hint of green in the image (both ways, normal and inverted, to remove the slight magenta halos). Aside from that though, i did very little. I added a bit more Vibrance, then applied a bit more NXT to all but the brightest areas, then applied a bit of Dehaze (this actually darkens the image slightly, but helps the dust pop a little bit more), then finally downsized it to 66%. For images like this, without any fine filamentary detail, for me the real draw is being able to see as much of the dust at once. In this case I don't think viewing at 100% really offers anything extra over something like 66% so i like to downsize a little, but that's just my own personal preference. It also helps with SNR too obv. NXT has totally converted me now as well i must say. Historically i've always preferred a little bit of grain in images rather than too much NR, but NXT is just so good i'm quite happy to go with more than i normally would. 

Result is below. As you can see, not much different to yours as it was already excellent!

Stuart, that's simply amazing how BXT has transformed that data! 

A lot of very skilled imagers have been happily making jaw-dropping images for years now without using any form of deconvolution (mainly because of it's dark art form, I guess). So if this now makes it a trivial affair, then that's great! For me though, the real wow factor comes from how it can correct star shapes. That just blows my mind tbh, and your images are a perfect example of what's possible now with this software. I'm also starting to wonder whether we still need all these expensive flatteners anymore, lol 😋

Ps - I read a few comments where people where saying they weren't seeing much improvement in non-stellar details. I think this could be down to not using the Manual PSF option. Here's a video that describes it pretty well. You can skip to 10:30 for that specific bit

Came across this last night. Russ Croman has done it again, this time using AI to help with deconvolution\sharpening. Adam's video is excellent, i really recommend watching the whole thing. The improvements he was able to achieve are remarkable, so much so, i think i'm now going to have to finally give in and get Pixinsight! Up to now i've been happy using StarXT and NoiseXT in Photoshop, but this one only works in P.I as it needs linear data. It's a bit pricey mind, $100 (or $90 if you already own one of Russ's prior tools) but once i saw how it can even fix bad corner stars (or even stars affected by differential flexure throughout an image) then i realised it was something i need to have in my life. 

And the best part is, like all of Russ' tools, it looks easy to use. Deconvolution is a total faff to try and do, so this simplicity is a big plus in my book. 

10 hours ago, jjosefsen said:

Wonderful image! Nice processing for sure. Thanks for sharing. 

Thanks jjosefsen 🙂

On 04/12/2022 at 15:21, ollypenrice said:

Very nice, especially the dust. It has just the right texture to my eye.

Olly

Merci Monsieur! The main objective of the re-process was all about the dust so that's good to hear 👍

On 04/12/2022 at 15:58, gorann said:

Yes, a very nice image Ciarán!

I think quite a few of us have these mediocre images in stock that now can find new life with the new tools!

Absolutely Gorann! It's almost like having new data isn't it? Happy reprocessing 🙂

On 06/12/2022 at 15:25, old_eyes said:

Well you certainly got an improvement! The latest version in much better. The trick now, in my experience, is to understand what you did differently. I have been reprocessing some of my older images, and they are better, but since I did not keep the process history, I don't know whether it is being more selective in data, a software improvement, or me improving my eye and 'touch'.

Seeing as I use PS, what I do is I just make a new layer every time I make an adjustment. So I can always see a history of all the major changes I've made. It does end up in a huge Tif file though, and because of the max file size of 4GB there is a limit to the number of layers you can have (and which I usually end up maxing out) but it's good enough for my current workflow. P.I users definitely have an advantage here, but that requires using P.I 😋

1 hour ago, andy fearn said:

Wow, what an improvement. very nice.

Thanks Andy!

1 hour ago, Lee_P said:

Big difference, nicely done!

Cheers Lee 👍

1 hour ago, teoria_del_big_bang said:

Wow Ciarán, that's an incredible difference.
Do you think this is down merely just to new tools or has the operator improved considerably too ??? 😁

Steve

Hehe, yes fair point! The operator has definitely learned a few new controls since last time 😋 Even so, there's simply no way I could have made the new version without StarXT and NoiseXT, as the workflow relies so heavily on them. 

15 minutes ago, Elp said:

That's an excellent difference. My 20+ hours yielded a similar result to your first and I know there's more data there. In my instance the flats ruined pretty much every session so I can't push it very much. I found however using GIMP I could push much higher than I could in PS, maybe worth exploring though your new one looks quite natural and not over pushed.

Bad flats are hard to overcome for sure. I've heard good things about AstroFlat Pro though, might be worth looking into.

Thanks for the feedback. Good to hear it still looks fairly natural, as I did finish it very late last night. I knew I was pushing it quite hard so there's always the risk that you end up going too far and not noticing! 

36 minutes ago, old_eyes said:

Very nice image. Could you post your previous version so we could do a side by side. It would be really interesting to see the improvement from better tools, and your own developing skills no doubt.

Thanks! I never actually produced a finished image first time around, i was that unhappy with it. But i did find the version below, which seems to be the most recent one i had before i gave up on it. 

Pretty bad, i know! 😅 🙈

This data dates back to April 2020. In my mind it was basically a gonner, as i could never process it properly. I tried a few times to process it, but i was always disappointed with how it came out. Try as i might, i could never get the dust to really stand out, and any attempt to do so would just lead to the image breaking down, fast. I just figured i needed faster optics and darker skies, so in the end I threw the head up and deleted all of the lights, lol. Thankfully though, I had the sense to at least hang on to the stacks. 

This was taken with a Qhy163c. Truth be told, i never really liked it, as i could never take a good flat with it for some reason, so i sold it soon after getting it (I then moved on to a Qhy9m, before getting my current camera, a Qhy268m). So for this I had 3 stacks (shot on 3 separate nights) but one of them was shot during a full moon so it was no good. That left me with 2 stacks, each one shot on different nights and from different locations (Bortle 4 and Bortle 5/6). On both nights, the scope was pointing quite close to a street light, one of which was the old sodium vapour type, and on the other it was one of the newer LED ones. So when i stacked the 2 stacks the resulting gradient was truly horrible! Yet another reason why i always assumed the data was only fit for the bin. 

Well, fast forward 2.5 years and i thought i'd have one final crack at this, now we have amazing tools like StarXTerminator and NoiseXTerminator available.

Capture details:

Qhy163c

9.5 hrs (570 x 60s )

SW 80ed, HEQ5-Pro

Stacked in APP, processed in PS. Downscaled to 75%. 

Fairly happy with how it came out in the end, for what basically amounts to a bonus image!

C&C welcome, and thanks for looking 🙂 

Woah, that's deep Goran! I love the deep brown colour in your dust, and i really like the areas of blue reflection nebulosity you've picked up. Very nice! 

7 hours ago, tomato said:

First class image with modest integration. The RC add-ons are a real step change in processing IMHO. So easy to use but they give outstanding results.

Thanks mate. You're right about the RC-Astro add ons. There's absolutely no way i could have turned this data set into anything even half-way decent where it not for NoiseXT. 

8 hours ago, Knight of Clear Skies said:

Lovely image.

Can anyone else see the man taking his fanged python for a walk?

I do now. And now i can't unsee it! 😅

6 hours ago, tooth_dr said:

Absolutely wonderful Ciarán. When you said you only had 30mins of red I thought you must be mad, but it just shows you - an ED80 - the most humble of scopes - produces data that in the right hands matches anything out there even at 10x the price. 

Thanks Adam. I thought I was mad too tbh lol. I probably still am in fairness 🤪. It wasn't until I was about 75% of the way through the processing that I realised it was going to come together ok. 

9 minutes ago, Rodd said:

Noisext vs tgv denoise  how do they differ?

 

Not sure Rodd, i don't really use Pixinsight (apart from doing PhotometricColourCalibration). But i'd be willing to bet the farm (if i had one!) that NXT would beat it, or any other NR software out there. 

6 hours ago, Rodd said:

This is top notch. Very very nice. How the heck can you do this with only 6 hours total. Dark nebula for me requires 25+ hours. 

Cheers Rodd. It's all down to the software doing it's magic. StarXT to boost the dust, and then NoiseXT to bring it back from the precipice. I'm finding that I can usually make do with about 8 hrs of data, split as 5 for Lum and 1 each for R G B. I fell short of that here, which honestly made the processing very difficult, but luckily NoiseXT helped to hide all 1000 sins! 

3 hours ago, simmo39 said:

V nice! as always.

Thanks Simmo 🙂

This is my take on Barnard 150, aka the Sea Horse Nebula. 

The Lum was captured over 2 moonless nights on Sep 28 and Oct 1st. The RGB was captured on Nov 3 & 5, with a fairly bright moon about (although low down).

SW 80ed; Qhy268m; HEQ5-Pro

Lum: 125x120s (4 hr 10);

G: 20 x 180s (1 hr);

B: 20 x 180s (1 hr);

R: 10 x 180s (0.5 hr).

Total: 6 hr 40

Stacked in APP and processed in PS. 

Starless processing (using StarXTerminator) used to bring out the dust, and NoiseXTerminator used for NR.