ONIKKINEN

Some processing practice, and a look back at what i made then and what that same data produces today.

This one might be considered my first astrophoto. It was taken with a 550D piggybacking on an Astromaster 130MD with the RA motor engaged, lens is the canon kit zoom lens set to 100mm focal length and f/5.6 aperture. 32x 13 second exposures made it into the stack, i did take about three times that number but anyone who has been within 10m of an Astromaster mount will understand why only a third were usable. I was very proud of this one and i still am. It will always have a special place in my heart as the image that started my journey in astrophotography. I knew as soon as i saw the first sub that i had to get a proper setup and immediately fell down the rabbit hole (still falling, no bottom in sight).

My first process back in September 2020, in Gimp if i recall correctly:

And then what i processed just now in Pixinsight and Photoshop:

Then the second picture is the first somewhat successful night with my then new VX8 and EQM35 combo (not a very good match i should say) in January 2021. Camera was attached to the scope with a horrible contraption as i did not have the proper kit, it was a 1.25'' barlow barrel with the barlow removed clamped into an eyepiece projection t adapter with some thumbscrews and tape. The mount was powered through the interior 12v cigarette lighter plug of my Ford Focus with the engine running and the cable running through an open door, my batteries just couldn't do anything as it was very cold. MacGyver would have been proud to see that . No coma corrector yet, which should be obvious from the stars. I dont remember the processing, but judging from the file name it was in Photoshop. This was around 60 exposures of 30 seconds each, which was all i could get in the -20 degree weather with my aging Canon batteries.

And then what modern tools produce today, in Pixinsight and Photoshop. BlurXterminator recognizes the coma and just adjusts it away, its actually crazy how good a job it does. No flats and i had the back of the scope uncovered so there are horrible light leak gradients. I believe i also had the camera viewfinder uncovered, but i think i managed to hide most of those issues in today's process.

As a BlurXterminator showcase i thought to add the before and after view with aberration inspector mosaic view to showcase the difference better, since my original image has butchered stars to a point where coma is not so apparent.

And that's it for now, this was a good reminder on why you should never delete usable data no matter how bad it seemed at the time as future you will always produce the better image.

3 hours ago, Elp said:

Mele is another one. Minisforum also make small desktop replacements but might be overkill (also might not be suitable for a cold moisture prone environment).

I have a Minisforum Win10 mini-PC. Dont remember the exact model name but it wasn't very expensive (i think 300€ or thereabouts). Its not very good for planetary and has a habit of crashing Sharpcap pro with high framerate video captures. I think the i5-2000 something series CPU (dont know the exact model) in it cant quite keep up, would go for a different model if i bought one now. Its not a critical issue though, i just have to set the framerate limit to something like 10fps when starting the capture and then immediately unlock it once the recording starts. For deep sky never had issues.

Pretty significant improvement on that barred spiral between the 18 and 30 hour images. Some extra wispy bits of the messy spiral arms also much better shown in the new one.

Also, 4 cameras (or is it more?), 2 scopes, all those filters, 2 years and getting close to a thousand subs. Quite the ordeal to put the image together and not getting any easier if you keep going so i applaud the effort.

It takes me around 30 minutes to go from opening my car trunk to the first exposure being taken, could be a bit faster or slower depending on if i have to carve small dents into ice for the tripod to snugly fit in and not slide around. Using a newtonian so i also collimate in that time and like you i do calibrate every night. Sometimes its too windy to calibrate and i'll have to use the previous calibration, but these nights dont produce good data anyway so hardly an issue.

But this speed is for all intents and purposes meaningless, because my newtonian takes about an hour to cool down to ambient where tube currents disappear even with a fan running. So in essence i am 30 minutes too fast.

Before jumping to conclusions on sensor tilt you should attempt to eliminate all other sources of tilt. This could be focuser tilt or collimation (really these 2 are kind of the same thing). Its not a stacking problem, Siril has some fancy algorithm that will twist and turn every sub to match the reference frame, so all problems in a stack are things that are due to some issue with the scope/kit.

That said, adjusting that tilt away can be tricky so up to you whether you can live with that or not. BlurXterminator would certainly fix that, but it kind of costs an arm and a leg since you also need Pixinsight so probably not the solution you were looking for.

For your guiding its hard to say where the issue is without seeing the guide log. If you want to you could share the guide log for us to look at (just drag and drop it to the attachments, same way you do with images).

* Based on the image i would not actually bother tweaking anything, looks good. You can easily make things worse with tilt adjustments.

7 minutes ago, pipnina said:

If the brains are inside the mount then maybe I could plug my handset in and give it a factory wipe of some sort?

I'd say this is a possibility. My issues persisted even without the hand controller plugged in, so there is a chance the alignment model is stored somewhere on the mount head itself.

What does the raw recording look like? Check whether it has this banding or not to rule out capture and camera issues. If you captured as .SER you can download SER player to look at the recorded video.

Whether you have banding or not, check these settings:

If you do have actual horizontal banding, try enabling the Row Noise Correction option, and of course try to fix the issue for the next time as your camera really shouldn't have any issues with banding at all. If you dont have banding in the raw recording make sure this option is off, as it can actually create the bands if you have it enabled when there is no banding.

@Elpsuggested the alignment points, which is also something you need to see to. Its not apparent from your screenshots whether you had them or not, but you do need them.

Sorry, dont know anything about EKOS/Indi, but something like this has happened to my EQM-35 and AZ-EQ6 when using NINA. So there is a chance this is a Skywatcher thing and not a software thing.

For me the weird wrong way slewing happened randomly and wouldn't go away within the night it happened. I ruled out different things and was left with only one option, the mount itself being confused by sync commands from NINA creating a bogus pointing model of some sort. At the time i used the USB port on the hand controller as a way to connect the mount to the PC, and a temporary fix was to factory reset the mount through the handset. Most likely not helpful for your case, different mount and control setup and all but thought to mention that a very similar thing happened with my mounts.

5 hours ago, tomato said:

So I have obtained an SQM figure of 19.83 from ASTAP, but that was with with a sub captured about 40 degrees way from an 70% illuminated moon, but all of my recent M101 data was similarly moon affected. I'm assuming all of your data was collected on moonless nights so assuming your SQM was 20 and if I was to image under the same moonlit conditions,  I would need 1.425 times the integration time so around 35 hrs, about double what I already have.

Now I know why I went for a dual rig.😉

Im getting a different number for the factor, here is how i calculated mine:

For our scopes we get a multiplier of 1.47 times already, your Esprit 150 has an aperture area of 17671 square mm, my 200mm aperture newtonian with an obstruction of 70mm and mirror coatings of 97% (OOUK Hilux coatings) is an effective aperture area of 25937.

I went ahead and checked 3 subs from all my nights, 1 from the beginning one from the middle and the end and got an average SQM of 20.73 for the UV/IR nights and 20.43 for all the nights. I think the UV/IR data is doing most of the work based on the weighting calculations from Siril's plot drawing fucntion so lets call it something like SQM 20.6 as the average.

So with an SQM difference of 0.77 multiplied by the 2.5x per magnitude factor we get a 1.925x increase in required integration time.

In the end we get my baseline of 25h * 1.47 * 1.925 = 70,74h. So you would have to image for 70,74 hours with your kit and sky to reach the SNR of my 25h image, in theory of course. Not sure how well practice would follow this theory but safe to say your conditions would require significantly more integration time than with mine.

If you got some pristine moonless data i think it would not require nearly as much time since a 70% Moon probably adds somewhere between 0.5-1 magnitude to the sky if not more. Doable, since you get twice the data with 2 scopes at the same time.

An absurd number of galaxies, great image!

If you see only white and no detail you need to lower your exposure time. The Moon is very bright so youll use something like 5ms for capture.

This actually looks pretty good. If someone is picking up the hobby from scratch, then this thing covers your imaging camera, guiding, and control all in one and that price tag is not that bad either. That is if it really is 2000 USD (probably at least 3000€ here, yay taxes).

Yours is quite nice too, those fainter spiral arms are a pain to pull out of the noise and here they are readily available.

I think we could calculate an "equivalent time" between our scope/sky combos to see how big a difference there should be in SNR. Purely in aperture my 200mm is around 50% larger in area (roughly considering the obstruction and reflection losses from mirrors) but our imaging scales are almost the same (mine is 0.76''). The skies mine was taken from were probably around SQM 20.7 on average for the 7 nights i spent, some nights better than 21 and a few closer to 20, and integration time needed to reach some chosen SNR increases by roughly 2.5x per magnitude of sky brightness increase. So if you want to crunch the numbers you'd only have to figure out your sky brightness using ASTAP's sky quality measurement tool and then plot in the numbers to see how many hours would you need to reach what i got in 25h.

Isn't there a downloadable plugin for solar system objects in new versions of NINA? Check the plugins tab, it will be there if it exists.

Cant check for that myself as im using an older version of NINA, but i am 90% sure i saw a plugin like that when i last checked what was available.

Looks like a shadow, or shell of some kind within the PN. Bits blown off the star prior to becoming a PN maybe?

11 minutes ago, wimvb said:

There is a limit to that, set by your sky darkness. Going past some 22+ Mag/arcsec2 just isn’t feasible. But yours is an excellent image as is.

I think my magnitude record is in the 22 region, in my last year's 35h M81 image. Skies in the location i use are between 20.7 and 21.3 depending on whether we have snow or not so going much further than that might not be too realistic.

You can also take flats indoors by pointing the scope at an evenly lit wall or a tv showing a white screen (and a t-shirt).

I did the latter a few years back and it worked ok.

46 minutes ago, gorann said:

That is a truly outstanding M101 Oskari! Why more data if you what you have can produce such an image?

Up here in the north @wimvb and I have also concluded that we have taken our final images for the season due to moon and lack of astrodarkness. Good thing is that we avoid exposing our optics to the pollen season.

Thank you!

There is always this thought at the back of my mind "what if it were better/deeper still". So i have no trouble justifying the hours to try and chase that. I do think there is room for improvement, particularly in Ha because the Triband + OSC combo is very inefficient at capturing it (wide 30 ish nm pass and only 1/4th of the sensor). I reckon just a few hours of proper mono Ha will beat the current triband data.

The luminance aspect i am not so sure about. Part of me is excited to see if new faint background galaxies emerge from the noise, or if more stellar detail can be resolved in the galaxy. If the image doesnt improve much, well thats a problem for future me.

1 hour ago, callisto said:

Amazing... that's why I love astronomy ☺️ 

Mark.

Sure is a humbling hobby. There are countless objects in the image that are not in any catalogue, and this is a very popular imaging target so its strange to find so many of them. Many of those faint fuzzy spots are older than the Earth, all in just a single degree of field of view.

Thanks all for the compliments!

5 hours ago, tomato said:

As M101 images go it’s up there with the best, deep and detailed.👍

I love the background galaxies too, like this barred spiral:

One of the best things about going for longer integrations i would say. The background just tends to spawn interesting objects out of thin air.

For example here is a quasar with a redshift of 4.170:

Unassuming little speck, however that speck has spent more than 12 billion years on its way here.

5 minutes ago, WolfieGlos said:

7 sessions in 3 months must be demoralising (I thought you had been quiet), but I'd say from this one image it was worth pursuing. The detail in that core is really nice, Ha isn't overdone for me, and the 25 hours shows across the whole image. It could even be one you return to with more data next season?

If not, it's one of the best Pinwheels I've seen.

Actually its 7 nights in a year since the first 2 nights were in 2023 🙃. Of course its not all the nights that i had in the 2023-2024 season, just the ones where M101 was nicely positioned (17 nights in total since august 2023).

My plan is to get a monochrome camera for galaxy season 2025, and yep i would be shooting actual Ha and luminance for this one to make the best possible image i could reasonably make. I think there is a chance that the project will go on in 2026 if next year's weather is equally bad since i would like to get a similar amount of luminance to the current RGB stack (so another 20+ hours...), but whats another year when the project already celebrated its first birthday.

766 x 120s, around 50/50 split between UV/IR cut and an Antlia Triband RGB filter. Taken with an 8'' newtonian, TeleVue Paracorr, and a RisingCam IMX571 OSC camera.

Resampled to 75% of original resolution, so roughly 1'' per pixel.

The UV/IR cut filter data is all pretty good from better than average seeing nights and little or no Moonlight. The Triband data not so much, with some real uphill battle nights and 10+m/s winds and/or Moonlight. I think the UV/IR data is doing most of the heavy lifting here so not that big a deal. Imaged over 7 nights in total, 2 of which were in April 2023, 2 in February 2024, and 3 in March 2024. Haven't gotten a chance to image in April of this year mostly due to weather, and its looking like that is not going to change as the Moon is rising and the season closes in around 2 weeks so this will likely be the season closer for me. This year was significantly more windy than the 3 past years i have imaged, i think this might have been a 40+ hour image had the conditions been better but still shouldn't complain, its not like 25h is a short image.

Calibration and stacking in Siril, processing in PI and Photoshop. I used a 2 stack method for this one, with the first stack containing all the data and the second stack containing only the Antlia Triband data. The Triband only stack was used as an H-alpha enhancing layer in Photoshop which is a subtle but noticeable boost to the bright emission regions.

Feedback welcome

-Oskari

41 minutes ago, 900SL said:

Kiitos Oskari. Do you need dew heaters on the primary and secondary as well as the fan, or do you use a extended dew shield? I'll likely stick to a 150mm f6 with a smaller secondary for the 533, given the sensor size

And thanks for the offer of the Paracorr. Nothing decided yet and I probably wont get anything until later in the year, but I'll drop you a line if I go down that route. Have a collapsed storm drain to fix at the moment out in the yard, about 10m long to a blocked soakaway. There goes any astro-budget for the next couple of months. 

I've no heaters, the primary is so far down the tube it really doesnt get dew. The secondary is kept dry by a dew shield and flocking applied to the back of the mirror (so it doesnt radiate heat away as much). I think the fan also helps with dew prevention since there is constant airflow over both mirrors and dew is less likely to stick.

I'd go with the 8" newt, but then again as a newtonian user myself i am biased.

Something to keep in mind is that you need to install a fan on the primary mirror to keep the scope acclimated throughout the night. Temperatures can easily drop 10 degrees at night here in winter, which is too much for just passive cooling to do the trick. Its not an expensive thing to set up but one more thing to fix (just a computer case fan + some DIY for mounting and power).

I am planning on getting rid of my Paracorr because it has some field curvature over an APS-C sized chip, but i think it would be very good for your 533 cameras. Let me know if you are interested in that if you do decide to go for the newt.

14 minutes ago, Your Star Will Shine said:

Thanks and well figured out from my vague post

 

Yes Milkyway bortle 3 mostly

Bortle 3 is excellent, so no point in a light pollution filter unless your camera is astromodified in which case you'd need an extra UV/IR cut filter. If the camera is stock, you're all set for some great images and no filter is necessary.