Iem1

Thank you for all the comments and advice guys, managed to push integration time to 2 hrs 51 mins, quite happy with the result! Working on process but I will finish in the morning (Working on blending, taking @ONIKKINEN's advice)

New starless of the 2hr 51 min data for comparison to earlier, I see a lot more sharper detail for sure

1 hour 10 min starless;

15 minutes ago, vlaiv said:

I'm not very good at discerning focus with Bahtinov mask - that is why I never use one. I prefer FWHM and looking at the stars on screen (not sure how would that work with DSLR - if you can see stars on computer screen or you don't use computer at all).

Btw, data is good. SNR is good. I did quick processing and I did not need to use denoising at all. In fact, I was able to do some nice selective sharpening and still keep background smooth.

Only issue is with dark calibration (to be expected when camera is not cooled) - there is some unevenness in the background.

Here is the result of processing:

(for some reason I prefer this orientation on M31 - it took me a while to figure out which way it goes ).

I am glad you noticed the difference in symmetry between the long spikes and the balls of light close to the star too, it has been giving me a headache for a while! Everything I read suggests that the long lines need to be evenly spaced (no mention of the balls of light), but I do wonder if I am doing something wrong or the B mask might be slightly damaged/bent or something. I use the B mask that came with scope (WO own for the Z73) but I will order a second generic mask and compare. if the weather holds tonight Il also try test both versions of focus and a single frame, evenly spaced spikes v's evenly spaced balls of light

That is a first for my data!  ..Usually it is emergency damage/noise control, I think the manual dithering is absouletly worth it now. Done it for both the Iris nebula and this attempt at M31, a lot of extra work for sure, but the results speak for themselves when compared to my older, non-dithered images. That and stringent screening of individual image quality. It hurts spending 4 hours out in freezing conditions, having to delete 30%-40% of data, to finish with 1 hour and 10 minutes of data, but it will be worth it in the end   

This was round 1, clearly before I knew back focus was even a thing! (first ever DS image!):

This was round 2, before I knew processing was a thing:

Current progress in round 3 with more data to come, with a very quick 10 minute process just to see what is there:

All the same equipment used, but with around 6 months worth of coaching and guidance from SGL folk in between 1st and last  

5 minutes ago, vlaiv said:

How can you tell if you have good focus there?

If I look at this part of image - that looks like serious asymmetry to me:

(look at three little balls of light right next to star - two on left are closer together than one on right).

I agree, but if i adjust to get those balls of light evenly spaced, I find the actual long lines themselves become unevenly spaced, this is something I have noticed for a while and I am unsure as to which i should be making evenly spaced, the balls of light or the long spikes/lines?

Looking at the original image above, I think the lines themselves are pretty evenly spaced, not perfect of course, but reasonable. Compared to noticeably unevenly spaced balls of light

Afternoon guys,

I started a project on M31 last night, my 3rd attempt at it. It was my first ever attempt at a DSO, and shortly after I tried again, improved a little but neither were much good. Bad time of year, low in the sky and light summer nights. So I thought with my first 6 months in astrophotography almost done, and it being better conditions, I thought I would take another swing.

Not overly happy with results to be honest and wondering if it is worth adding more data tonight.

I only actually managed around 1 hour and 10 minutes of data (30 second subs with a WO Z73 on a SGP, iso 800, unguided, modded 600D) due to a slow start up, a meridian flip, and being extremely fussy about what data I kept. I must have deleted about 45 minutes worth of images due trailing, wobble, even the slightest of imperfections as i wanted to create a solid image. It was a bit windy and I am manually dithering between frames with only a 10 second pause between images, so I have to be quick or I can introduce wobble, will probably bump to 12-13 second delay between images to allow more time for the system to rest. It is a fairly short integration time, but does the data look ok to keep adding to? I feel it is still a little soft/slightly blurred. Doesn't seem sharp to me for some reason.

Full image with some processing;

Just the galaxy, having removed stars in StarX

I quite like the starless image, but I have not figured out a way to blend the stars back without introducing a grey cast to the image and blowing out the core too much, working on that!

Raw tif

result.tif

And here is my locked focus before starting (Stupidly forgot to take another at the end to see if the focus had shifted)

Wow @ONIKKINEN, that is a brilliant effort! And thank you for detailing how you did it and what you used, helps me with direction on ways to progress my own data handling skills.

I also noticed a big improvement between 16bit and 32bit, especially when I removed stars in StarNet, there was a lot more of the dark dust clouds clearly visible. Thank you for pointing that out! Learnt a lot from this imaging session.

I would post how I got on with my own 32bit attempt...but It is shambolic compared to your work, so il save myself the embarrassment 

And thank you again to everyone who gave me pointers. 

the original fit file

result.fit

1 hour ago, ONIKKINEN said:

I noticed that the raw stack you posted above is in 16 bit mode while still linear. Its very likely you have lost a not insignificant amount of signal to this! Stacking in 32bit mode takes a lot more processing time and power but you should still always do this.

I have noticed that stacking my 26 megapixel frames takes around 1 gigabyte per sub with SIRIL. Easier and less straining on your PC to do in DSS, but SIRIL stacks a bit better IMO. If you have no extra harddrive space to free up, try stacking in DSS to get the 32bit result as i think it would be better?

Thank you for the suggestion!

I just restacked in 32bit, but with compression. No way I could do without it. And I was getting pretty horrendous results in DSS, not sure what I was doing wrong but it was coming out a bit of a mess in comparison to SiRiL

result.tif

Il have a play with this too and see if I notice any improvement  

Final completed image.

Best I could do, basic soft initial stretch while masking stars incrementally, removed stars completely in Starnett and isolated the dark clouds and the core, saturated and stretched separate, then blended with the stars.

Then I split channels in Siril into rgb and opened up each in PS and did my best to remove noise before recombining back in Siril.

Thank you again for the help guys, always appreciated. Quite pleased with the result considering the situation in which the images were acquired, lows of -2 degrees (with a cheap USB Dew heater) and ~90% moon  

1 hour ago, Lee_P said:

Just five minutes of editing on your version 2. The longer integration time has really helped! 

Nice job Lee!

I am going to try a little stretch before attempting to remove the stars, see if I can get them a little smaller. Might also try split the RGB channels to remove noise separately, if I can remember how. Think PS has updated since the last time I did it, throws me off  

38 minutes ago, StuartT said:

Hey, great that you got to grips with Siril. I must say, I totally love it and can’t imagine using anything else from now on. Prob slightly simpler for me as a) my camera generates FITS files, b) I don’t shoot any calibration frames. So I just use one of the available scripts. It’s a question of pressing a button and bingo! A few mins later I have the stacked image.

note: Siril generates a huge folder called “process” which you should remember to delete when it’s finished stacking. So you do need quite a bit of spare disk space for it to work in (especially in my case as I’m not guiding so I’m stacking several hundred, short exposures).

I then have a standard workflow post stacking;

1. crop

2. Background extraction 

3. color balance (photometric) - this really is like magic in Siril!!

4. Simple stretching

5. remove green noise (not always necessary)

6. Color saturation

7. save as TIFF

Yeah that I noticed! I had to delete a bunch of 'in process' images several times and restart it several times before having enough space for it to complete, as well as converting from 32 bit to 16 bit and using RICE compression I think it was called. Also had to scarp half my calibration frames, each light frame was being duplicated at 300 mb a pop...and I originally had 50 of them alone, plus the 50 bias and 500 light frames

but I got it done in the end and the result was definitely worth the hours of tweaking!

The photometeic calibration is amazing, and worked with most of the other stacks I did previously, but I couldn't get it to work with the new stacking! Had to do it manually as best as I could 

Finally figured out how to stack in Siril! Though not separate flats, so I have used the flats from night 1 across all the data. Not ideal, but I will take what I can at this point  

What Siril spat out;

Much better field of view! Bar my exceptionally dodgy processing, an improvement. Thank you for the suggestion @StuartT!

Struggled to process this as the tif was a lot different than I am used to, the green channel was exceptionally strong and the red weak, the opposite of what I usually get.

Not entirely happy with my finished product, will revisit and keep tweaking the processing. It has a bit of an overcast I need to get shot of.

Here is the new raw data from Sirils stack if anyone wants to have a play 

result.tif

1 hour ago, Turbocoo said:

What monitor are you guys using (besides dslr lcd screens) to view and focus etc? 

ATM I'm using a 21 inch TV via hdmi from the pc. My canon 250d displays focus on the Canon eos software on x5 or x10 zoom 

Problem is the view on the TV looks very pixelated with the mask on 

I've seen lots of bahtinov focus screeshots with really sharp display. 

Any help appreciated 

Just the lcd screen on my 600D  

I tried using SiRiL to stack the images as suggested, I somehow managed to stumble my way through making a Black and white stacked image with no calibration frames. I then did a little research and made an appropriate pathway to the folders containing lights/bias/flats and got the script for OSC with flats and biased but no darks, but I keep getting image dimension errors.

I am having to open images individually and covert them to .fits for it to accept. It was an issue with Bias, then flats, which was ok, 20 of each converted. But now it is taking the same issue with the 500 light frames. Not sure where to go from here.

The black and white image actually looked really good too!

Yeah I might have a look at stacking in SiRiL, see if I can improve on it. I had a quick look but seemed a little confusing, I am sure I can find a guide online!

I think I might have over done it on the manual dithering, ever so slightly exceed the 'few pixels' required..

Night one data Stacked on its own;

Night 2 data stacked on its own;

combined data;

What I ended up with;

And this is the unedited, but cropped image, the one I proceeded to process

Iris combined(cropped but no edit, for ps).tif

I managed to get around another 1hr and 30 mins last night, the data seems good, stars no longer blown out as much. 

But it is a nightmare trying to stack the two sessions together! They do not align well at all, which i was expecting to some degree because of the manual dithering, but I wasn't expecting the result I got.

I did my best to mitigate all the damage and work with what the stacking would reasonably allow, so here is my current take on the Iris nebula.

~x450 30 second images, ~3 hrs 45 minutes of data over two nights, during a 94% moon and 88% moon respectively;

it is quite heavily cropped, more than I would like. And stars are still large due to 1st nights data unfortunately

I tried quite a few methods of stacking in DSS (Standard, Mosaic and a custom rectangle), here is an example of what I was getting back;

Stacked and untocuhed.

Autosave002.tif

Not the greatest of outcomes, but it is a learning curve!

1 hour ago, Lee_P said:

There are three main things you can do to combat noise:

1. Longer total integration times. The more good quality data that goes into your stack, the better your signal-to-noise ratio will be. Here are some tips. 

2. Take calibration frames (Darks in particular). Having a cooled camera sensor does help to reduce noise a little, but the real benefit of a cooled camera is that you take set it to a particular specific temperature and then create a library of calibration frames that can be reused. Calibration frames are still important with a non-cooled camera like a DSLR, they just take a bit more effort to produce.

3. Noise reduction during post-processing. What you can use depends on the software that you're using. I find that Topaz DeNoise AI can work well, particularly on starless images (you then add the stars back in afterwards).

Thank you lee, I am looking forward to hopefully doubling integration time tonight, providing the weather holds!

Be good to see the comparison.

3 hours ago, StuartT said:

Thanks.

I guess the noise is mostly down to the fact that you're not using a cooled camera (but I am no expert on that!)

As for your star sizes, not sure. I used Siril too and I tend to only use the autostretch (not the Asinh). I don't take any calibration frames - my image is lights only.

I'm sure adding more data will improve matters - it will bring the nebula out more anyway. How were you focusing with the 600D? Could that be it?

Yeah, true.

Though it felt pretty cool last night as it was between 1 and -1 degrees xD

It shouldnt be the focus, I used the B mask that came with the WO scope, although I am not a fan of it. It seems as though sides of the diffraction pattern are not symmetrical sometimes. But I focused and locked it down. 

About all I can muster with my measly processing skills, lost some of the dust and messed the stars up, but it was a bit of fun. 

Nice Stuart! 

I was very suprised I managed to get a somewhat respectable result! 

My image is still very noisy though, not sure what I could have done to have improved it.

And I don't know why my stars are so chunky either. Maybe it's because the data is in there, and SiRL detects it and brings it out, but in order to do so it needs to stretch the hell out of it, blowing up the stars? 

I'm wondering if I get another 2.5 hours tonight, will it help, or make it worse? It will either add more data and perhaps SiRiL will have to stretch less to display data, cleaning it up some more...or will it just compound current issues? Hmm.

Having a play with it now, struggling to stretch myself and keep noise under control xD 

Thank you for all the replies guys!

Decided to bite the bullet and take a stab at the iris nebula  

This about x270 30 second light frames (~2.5 hours), x50 Flats & x50 Bias. Taken with a WO Z73, Modded Canon 600D, all on a SGP, unguided, taken with a backdrop of a 94% moon no less!

Looking noisy and in need of a lot of work, I also tried manually dithered between each image, hopefully helped somewhat. Hoping to get another 2.5 hours tomorrow night to improve 

I haven't done any real processing yet, just chucked it into SiRiL and done a few auto bits and bobs as it is almost 4:30 am, but I couldn't wait to have a sneak peek!  

As always, The raw for tif if anyone wants to take a look;

Autosave.tif

I have a second question regarding correct exposure during what will be ~95% moon tonight.

I was out a couple days ago where the moon was a similar brightness, I took a few shots in order to try and find the iris nebula and see the kind of impact the moon had.

Here is a single 60 second image opened up in PS, untouched;

Does this look washed out? Debating if I should be doing either 30 or 60 seconds under these conditions.

Il be under the same sky again (Bortle 2-3, depending on which site used) with near enough the same moon.

Actually amazed I found the Iris nebula so quickly, but I will admit, I sat there staring at my 600D screen trying to decide if that was just a funky star or the core of the iris  ..But the surrounding darkness gave it away.

6 minute (x6 60 seconds) of it with a little stretch and levels balance;

On 20/11/2021 at 07:49, GalaxyGael said:

Shoot between NW to NE, below zenith and you have good options with that moon. by that time, it will be less than full, which helps a lot. The open clusters are nice when they get a lot of integration time, more than most people usually do. NGC 7789 (Caroline's Rose) is a good choice. For something with dusty features, I would recommend the E nebula (B142/143). 

 

Thank you for the advice GalaxyGael, will check out the suggestions now in telescopius  

Hey guys, hope you are all well.

At the moment, we apparently have two potential clear nights coming this Sunday and Monday...But of course these nights will be accompanied by a near full moon (typical!)

Haven't had much luck with the weather/forecasts for a while, so hoping it stays clear. I am eager to get out and try and image regardless of the moon as I have not been out properly for a while.

I was aiming to try and image the pleiades as it is quite bright and would hopefully stand up better than most other targets in heavy moonlight, but unfortunately the moon is positioned pretty close to it (see below)

So I am weighing up my options and just thought to seek advice from others regarding possible projects over the two day period given the circumstances. Would love to try a 2-day project and combine data to break my current ~2 hour max integration time, given the circumstances. 

I will be shooting with a WO Z73 (430mm) + Modded 600D on a SGP, unguided (Limited to 60 second subs) from a bortle 3-4

I would love to try the iris nebula, shoot for 4+ hours as a target, but I am not sure just how worthwhile it will be with the moon so full. Seen others have success with OSC and short exposures during a bright moon, but not sure about it. It is further from the moon, but also a little low.

Any advice welcome, want to get a plan in place before travelling  

@chewie That's a pretty image! 

Yeah, I might experiment with a single image v's a stack soon.

This was the first attempt at a milkyway shot I took, and it was taken only a short way down the road from where the above one was taken

This was I think either a single 60 or 90 second image of both the foreground and sky.

fellow SGL members helped me blend the two images, and I think they also processed the sky a bit, but nothing drastic...it certainly packed more punch than my stacked imaged above.

Though I guess there could be many reasons for differences..most likely me doing something wrong

looking forward to getting back out there for another go

(Still bitter that I started nightscapes just as milkyway season ended tool! :D)

5 hours ago, Elp said:

Nice shots, last time I did the Milky Way from a Bortle 4 I did ISO 2000 for 30 seconds unguided, 30s being the limit of the camera I had at the time. Normally for camera I use between ISO 800-1600.

Looking at the yellow skyglow have you considered a filter? This will have the effect of needing to expose for longer though you are already acheiving that. Also the more light frames you can stack the better and combine with flats, darks and bias to remove any artifacts and noise.

Thank you for the response!

yeah, I think modded 600D isn't exactly a top notch camera and the Samyang isn't top notch either, so I guess it's a case of making the best of it. If I can drop to 800 it might help with noise im thinking.

never thought about a filter, but will look into it. I silly travel to bortle 3-4 so it's never been too much of an issue I don't think, thank you

13 minutes ago, AstroGee said:

I sometime find it tricky to gauge the brightness also, especially when working from a laptop with variable brightness. I haven't yet but plan to get a set of images which I know are at the proper exposure, to use as a comparative reference.

Yeah it is a finicky business

I have done a bit more research and learned that with the long exposures with the tracker, I should be able to also go from 2.8 up(down?) A stop or so? To help with coma and the like and hopefully improve quality a bit? Something I may try

Haven't had much luck in the way of weather, I had a 20-30 minute window to get the images above between clouds...Wish I had more time to play around with settings

- swap from 2.8 to 4

- Drop ISO from 1600 to 800

- Raise exposure to 2 - 2.5 minutes 

- Increase total integration time and apply calibration frames

^ Plan of action during next outing. Having a lot of fun with nightscapes (While I am saving up for my EQ6-R Pro :D)

17 minutes ago, AstroGee said:

great start, well done!

I'd be curious tp know when star trails start to appear with the setup you have, perhaps a longer exposure would allow a less noisy ISO.

If it were me I'd choose a longer exposure on the foreground, to lighten it up a bit more. Not much though, I like the silhouette and the blue windows.

Would a colour tint adjustment remove the orange have in the lower sky?

looking forward to the next image. cheers for sharing.

Thank you!

I might try shooting for 2-3 minutes next time with ISO 800, hopefully help with noise a bit more  

The foreground seems to be darker when uploaded here and viewed on PC, the photo on social media or viewed on a mobile seems a touch lighter!

Hereis another shot, but I angled the camera up a lot more; Though I am not a fan as I feel like it is quite unnatural for some reason, but it does capture Andromeda I think!