Meh101-First Galaxy Attempt

[MarkAR]

Well I thought I'd have a go with my 80D (standard) on the Skyguider with my Canon 100-400 at 400mm (I think my lens might have a problem). Took 207 x 30 sec subs + B,F &D's

Looks like I might be better off with longer subs, shame Canon won't let you use the internal intervalometer in bulb and mirror lock, will have to hook up to the laptop next time.

Done the best i could with some Gimp processing.

 

Added the original if anyone cares to have a bash at processing.

Pinwheel 2.tif

[MarkAR]

Had another go without playing with the colours.

[gilesco]

To be honest, I found your TIF file didn't look like flats / darks and bias had been applied, so difficult to really get anything from it, once I tried to remove vignetting, remove light pollution, calibrate the background and calibrate the star colour I found a lot of anomalies that I would normally expect to have been removed with Darks / Flats & Bias.

Actually, if you want to provide all data, darks, flats, bias & individual exposures, I will quite happily try and put them through an semi-automated process and see whether we get anything better? Not sure how much data that would be, probably a big download.

Edited March 30, 2020 by gilesco
suggestion

[Paul2019]

Well as we are in lockdown and I'm looking to get back to things, I thought I would have a go 

 

[MarkAR]

4 hours ago, gilesco said:

To be honest, I found your TIF file didn't look like flats / darks and bias had been applied, so difficult to really get anything from it, once I tried to remove vignetting, remove light pollution, calibrate the background and calibrate the star colour I found a lot of anomalies that I would normally expect to have been removed with Darks / Flats & Bias.

Actually, if you want to provide all data, darks, flats, bias & individual exposures, I will quite happily try and put them through an semi-automated process and see whether we get anything better? Not sure how much data that would be, probably a big download.

I thought it looked a bit suspect after it was stacked, I just used the basic script in Siril. Could be that the flats are wrong, here's one to look at.

How's the best way to get you 200+ subs, I think I can email them via a Dropbox?

IMG_5162.CR2

[gilesco]

It looks like the flats should have done some compensation. Just looked like they hadn't been applied.

I've not used any of the cloud file-sharing systems, but I'm sure that some cloud sharing system can host them temporarily, I guess they provide a link for others to download them. Maybe Google Drive can do this?

[MarkAR]

Just had an idea, send it from one of my email addresses to another and it should create a dropbox link I can post on here.

Shall I delete all the fits files?

[gilesco]

1 minute ago, MarkAR said:

Just had an idea, send it from one of my email addresses to another and it should create a dropbox link I can post on here.

Shall I delete all the fits files?

Just your RAW Lights and Calibration (Darks, Lights etc..) frames, prefer to have them in a ZIP archive as a single file.

[MarkAR]

Will see what I can do.

 

[MarkAR]

https://drive.google.com/file/d/1ff-ag8iJOua4hn259sUyl755kqGu3dLI/view?usp=sharing

Hope this link works, not used G'Drive before.

[gilesco]

19 minutes ago, MarkAR said:

https://drive.google.com/file/d/1ff-ag8iJOua4hn259sUyl755kqGu3dLI/view?usp=sharing

Hope this link works, not used G'Drive before.

OK, well I am downloading a 7.5GB file, so it seems to be working, will take a while to download.

I will take a look and try processing with APP which is what I'm more acquainted with.

I think you were using Siril, so I may also try the same process with that, although I'm not so acquainted with that.

Do you have any objections to me video'ing the processing session and posting it online on my website?

[MarkAR]

Excellent, took over 2hrs to upload.

Go ahead and use anything your acquainted with, in Siril I just ticked debater, changed the directory and used the DLSR Processing script.

Absolutely yes go ahead and video it, post up a link when done. Every tutorial has something to help someone.

Just a thought, put all the subs into one folder so here are the image numbers if it helps.

Biases- 4269 to 4299

Dark- 5142 to 5161

Flats- 5162 to 5181

Lights- 4932 to 5138

[gilesco]

OK, going to do a quick processing in the background with most of the defaults with APP. Might be able to post an initial result when that is done, then will look at using this dataset in a video with both Siril, perhaps Sirilic, and APP, would be a good comparison of the tools.

[gilesco]

Well, I have an integration result, on the registration at least one of your frames was way out of alignment with the others.

I think I need to repeat this with a blur of some kind as there are some reddish artefacts appearing on the background.

 

[MarkAR]

Looks like red rain !

Zoomed in on the galaxy and there doesn't seem to me much more that can be pulled out from the image.

I'll have to find and delete that odd frame. Not sure how that happened unless it was the initial framing shot.

Edited March 31, 2020 by MarkAR

[gilesco]

By the way, your original assessment of longer subs, although resulting in less frames per hour, will really help here, as it will help your Signal-to-Noise ratio, I found it difficult to assess your subs individually as they are quite faint, and stretching brings in a lot of noise from which the source is difficult to discern.

I'm taking another pass at your data, while adding algorithms to what I normally do, although if I add everything I will just goose-step all over your exposures!

It is an interesting experiment for me!

[MarkAR]

Longer subs in the future, I was using the laptop trying to figure out my main rig and Ekos and having problems so just thought I'd leave the DSLR off to run on its own.

Its a pain that Canon won't let you use the intervalometer in bulb mode.

Also not sure if my laptop would cope with running the main rig on Ekos plus the DSLR either on AstroDSLR or anything else.

[gilesco]

12 minutes ago, MarkAR said:

I'll have to find and delete that odd frame. Not sure how that happened unless it was the initial framing shot.

I might be able to identify it for you, on the next run I am going to omit outlier / low quality frames - I've also posted on the APP forum as to whether they could introduce a feature to export in CSV format the quality information that it calculates for each frame after they've been analysed, as this would help any third party to the software get an insight into the data.

[gilesco]

1 hour ago, gilesco said:

I might be able to identify it for you, on the next run I am going to omit outlier / low quality frames - I've also posted on the APP forum as to whether they could introduce a feature to export in CSV format the quality information that it calculates for each frame after they've been analysed, as this would help any third party to the software get an insight into the data.

Starting first integration, I took a note of some of these though, here are the worst frames in each of the categories, if the Integration doesn't produce anything better then I will manually exclude these and repeat it:

Lowest Quality Images:

5137
5135
5053
5138
5131
5100
5107
5122
5066
5106

Noisest Images:

5137
5135
5053
5138
5131
5100
5107
5122
5100
5107

Most Poorly Registered:

4932
5041
5097
5006
5008
5019
4984
5075
4953
5137

Worst SNR:

5138
5100
5123
5107
5125
4983
4970
4936
5053
4995

Worst on Star Shape:

5019
5006
4953
5066
5122
5137
4093
5008
5053
5000

[gilesco]

So dug myself into a ditch here, the "red rain" as you put it, still remained after integration, despite various methods to try and remove it.

I can only suggest that you increase exposures to maybe 60s, or if your guiding can take it, even longer (I love my 180s exposures on a CMOS camera, when I get the chance), what we are battling with here is noise in the sensor - and, less noise, or more signal will definitely bring better quality to the show.

I've not been able to improve upon the overall integration result that I've already posted before.

It is always worth keeping old raw data though, as new techniques do come up all the time, and sometimes you wonder - "I wish I had put that through that"

[gilesco]

oh and, here are the master flats, darks, bias + a bad pixel map for your Canon EOS 80d

https://drive.google.com/open?id=15qrX-pL9jPcUsLsA2vvk1Iudd_ki-s5q

The Bad Pixel Map might come in handy if you haven't one already.

[cuivenion]

Here's my attempt from the file in the first post.

[MarkAR]

@gilesco A huge thnkyou for looking at this and taking the time and effort, I'm going to try a re-stack without the worst frames and see how that looks. I definitely need to get longer subs, though 60s at 400m f5.6 on a sky guider might be pushing the envelope. Should be possible at 300mm though.

@cuivenion Great effort as well, you've managed to get better colour than I did.

Thank you all.

[MarkAR]

Well I removed all the dodgy images and stacked again in Siril with the same script, only took 36 mins rather than the 1 hour previous. Though that might be because a parcel arrived with my 64GB memory package upgrading from 8GB 🤣.

As for the image, it looks slightly cleaner. Will have a tweak later and post it up but I still don't think there's much more to see in the Pinwheel.

[gilesco]

48 minutes ago, MarkAR said:

Well I removed all the dodgy images and stacked again in Siril with the same script, only took 36 mins rather than the 1 hour previous. Though that might be because a parcel arrived with my 64GB memory package upgrading from 8GB 🤣.

As for the image, it looks slightly cleaner. Will have a tweak later and post it up but I still don't think there's much more to see in the Pinwheel.

Well it's been a pleasure trying to get something more out of it, look forward to the results of something with 60s or higher exposures.