Jump to content

sgl_imaging_challenge_banner_31.thumb.jpg.b7a41d6a0fa4e315f57ea3e240acf140.jpg

Oh No! Not another M31. Yawn.


Recommended Posts

Taken 14th/15th Sept 2013.  6 x 30 minute subs, all processing in PixInisght.

post-18840-0-57558100-1379272242_thumb.p

Imaging telescopes or lenses: Sky-Watcher Evostar 80ED DS-Pro
Imaging cameras: Canon EOS 500D / Digital Rebel T1i
Mounts: Sky-Watcher NEQ6
Guiding telescopes or lenses: ST80
Guiding cameras: QHYCCD QHY 5
Focal reducers: Sky-Watcher 0.85x for 80ED
Software: PHD guiding,  AstroTortilla,  PixInsight,  EQMod,  APT
Filters: Hutech IDAS LPS P2 2"
Dates: Sept. 14, 2013
Frames: Hutech IDAS LPS P2 2": 6x1800" ISO400 bin 1x1
Integration: 3.0 hours
Darks: ~109
Flats: ~101
Bias: ~330
Edited by IanL
  • Like 17
Link to post
Share on other sites
  • Replies 43
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

Taken 14th/15th Sept 2013.  6 x 30 minute subs, all processing in PixInisght. Big version here: http://www.astrobin.com/full/56573/?mod=noneImaging telescopes or lenses: Sky-Watcher Evostar 80ED DS-P

Yep taken in sunny Essex this very Saturday.  I am out of town by about six miles, and no street lights in our village (just the usual random security lights of course).  The LP is bad enough to the n

I did some tests and I'd say the easiest thing you can do is to make a mega-bias.  I made mine at the same ISO (400) as I shoot the lights at.  There is a noticeable difference in the fixed pattern no

Posted Images

Very nice. Not tempted to do a mosaic?  That would look great.

Well it was just a quick run out to test my gear after the summer break.  If I'd had more time to plan it I would have rotated the camera and framed it better so probably wouldn't have had to  do a mosaic.  The current setup is too hard to rotate on the fly as it requires tiny grub screws to be undone, which is not something I would want to do in the dark.  Was pleased that I managed to get six 30 minute subs though.  Only had to throw one away, but had a few problems with others that I aborted as I kept getting run-aways in Dec which has never happened before.

Link to post
Share on other sites

Might be just another M31 - but an excellent capture and process.     For 6 x 30 minutes subs on the 500D is impressive - it looks really clean on the astrobin full size version, which must speak volumes for very careful calibration.     Don't think I will ever get bored of Andromeda, so keep them coming :)

Link to post
Share on other sites

I'm very impressed by this image for all the reasons stated above.

Was this taken in Essex? Or have you found a dark sky location somewhere? What does the LP look like on the raw images?

Yep taken in sunny Essex this very Saturday.  I am out of town by about six miles, and no street lights in our village (just the usual random security lights of course).  The LP is bad enough to the north through the West (even get some of London's output in that direction).  The North Sea is to the South through East with not much between us and it, so depending on transparency, it can be reasonably dark (just about average for this image though, so not great but not bad compared to an urban location).  Hardly a dark skies reserve though.  Using an LP filter helps, though actually it doesn't help so much with galaxies (and the background generally), since you have to expose for longer but overall the result seems better so I am sticking with it for now.

Below is one unprocessed sub (just did a basic histogram stretch in PixInsight).  I did put in about 200 hours of learning/testing on my Markarian's Chain image (see post elsewhere in this forum) and it has really helped, but every image is a new adventure so some tips below.

post-18840-0-10753700-1379367224_thumb.p

- Long subs are better than short ones if you can do them without saturating the target or getting washed out by LP.  I am convinced of this now (I was before but this image has proved it to me beyond doubt). 6 x 30 minutes beats 18 x 10 minutes any day of the week, though the risk of losing a sub is worse (had to throw away one 30 minute sub due to bad guiding and abandon 3 others about a third of the way in).

- Don't be afraid of the noise on a hot DSLR chip.  Long subs and still relatively warm evenings produce a snowstorm, but it does process out if you do it right.

- Use lots of bias frames.  My master bias is 330 frames at 1/4000th of a second.  It was a challenge to stack in PI but worth it.

- Ditto for dark frames.  I used 109 dark frames (shot them in a fridge with ice packs).  The exif temps range from about 9C to 13C, but PixInsight's noise scaling routines for stacking the darks, and also for applying them to the lights works really well since it just tries to statistically match the noise between frames.

- Same with flats.  This had 101 flats applied.  I did them in daylight by using a perspex diffuser on the scope, pointed at a white(ish) curtain illuminated by daylight.  Got the histogram to just over 1/3rd from the left.  The flats are better than anything I have managed with a lightbox.

- For any PI users, I did about 30 tests on Markarian's Chain and evaluated the noise after each to find the best approach (tried every variation you can think of).  Make the master bias by stacking lots of frames, use Winsorized Sigma Clipping for rejection.  Make the master dark by stacking darks, don't calibrate with a bias, again use WSC for rejection .  Calibrate the flats with just the master bias and stack (don't use a dark frame), use WSC for rejection.  Calibrate the lights with master bias, master dark and master flat, test the different pixel rejection routines to find the best result.  Linear Fit Clipping seems best for lots of frames (30+) and in this case Percentile Clipping was used (as only 6 frames.

- The other thing I discovered from this image is that my previous PI colour calibration process was killing the colour pretty badly.  Normally the combination of LP filter and dodgy colour-cast flats from the lightbox made it hard to get a good start.  I was using BackgroundNeutralization and ColorCalibration to fix this, but it was just making the channels too well lined up and so there was not much colour left.  This time with a better flat (all three channels equally exposed as far as I can see) I skipped these two processes and just did a simple histogram transformation once I was done with linear noise reduction processed.  The colour balance is a lot better (though probably a bit too blue).

  • Like 2
Link to post
Share on other sites

Great M31 and very interesting reading about the amount of darks flats and bias you are stacking. I am not sure of the exact science of it all but have found out for myself just how much difference getting a good stack of darks flats and bias makes.

I used to think that after stacking about 30 it made no more difference but if you experiment you soon find that is not true with some images.

Link to post
Share on other sites

Wow that is amazing! Great shot! I hope to take a pic like that!

I'm still learning when it comes to long subs, are your darks 30 min also? And did you put your camera in the fridge to match the temp?

Cheers,

Chris

Sent from my Nexus 4 using Tapatalk 2

Link to post
Share on other sites

Thanks Ian for your detailed reply. I'm even more impressed now I know you used a light polution filter. Is that the Astromik CLS filter?

I used one once, but couldn't get rid of the green colour cast in the post processing.

Link to post
Share on other sites

That is a very nice M31, nicely processed. yes interesting how many flats etc you used. i only do 20 flats & flat darks. it is a pain with the QSI583 as each one takes 22secs to download, and I do them straight after an imaging run, as dont like to risk moving anything in the imaging train. 

Fay

Link to post
Share on other sites

Great M31 and very interesting reading about the amount of darks flats and bias you are stacking. I am not sure of the exact science of it all but have found out for myself just how much difference getting a good stack of darks flats and bias makes.

I used to think that after stacking about 30 it made no more difference but if you experiment you soon find that is not true with some images.

I did some tests and I'd say the easiest thing you can do is to make a mega-bias.  I made mine at the same ISO (400) as I shoot the lights at.  There is a noticeable difference in the fixed pattern noise as you add more and more frames.  Since the bias is subtracted so many times during processing, getting the best one you can is important as it reduces the noise contribution.  For PI users, the trick is to use the batch conversion script to convert the bias frames from CR2 (raw) to FITS before you stack them.  PI can't read raw files progressively in to memory, but it can do so for FITS, which makes it possible to do these massive stacks on a reasonably modest amount of memory (6GB in my case).

Wow that is amazing! Great shot! I hope to take a pic like that!

I'm still learning when it comes to long subs, are your darks 30 min also? And did you put your camera in the fridge to match the temp?

Cheers,

Chris

Sent from my Nexus 4 using Tapatalk 2

I did use a fridge to make sets of 10 minute darks at ISO400 with matching EXIF temperatures, and ended up with different sets of between 15 and 30 darks at 9 to 13C.  I made master darks for each set and then calibrated the matching temperature lights using PI's dark scaling (see below), stacked the lights and then evaluated the noise of the result.  I then made one master dark out of all 109 darks (different temperatures together), calibrated and stacked and evaluated the noise.  The latter process produced a slightly less noisy result, but there are two important points:

- PI does a noise evaluation of each frame and weights them when stacking the masters, and I think this is what allowed me to stack the different temperatures successfully (within reason, I wouldn't try a set of 9C and 25C for example, though it might work I suppose).

- PI calculates a dark scaling factor when calibrating the light frames.  Basically it tries to find a multiplier for the dark that results in the least noise in the calibrated frame.  This means that you don't need to match exposure lengths between darks and lights, or match temperatures either since it's going straight to the heart of the problem, i.e. how much noise is left after calibration?  This definitely works, as you can see the multiplier changes in proportion to the exif temperature of the light frame.

So the upshot is that I used 109 non-matched darks stacked together, and calibrated 30 minutes exposures with 10 minute darks.  Be aware that this wouldn't work in something like DSS as far as I know, and you'd be advised to match temperatures and exposures in that package.

Thanks Ian for your detailed reply. I'm even more impressed now I know you used a light polution filter. Is that the Astromik CLS filter?

I used one once, but couldn't get rid of the green colour cast in the post processing.

I use a Hutech IDAS LP2.  Not cheap (especially the 2" version I have) but it is well worth it.  There is a slight green cast, but PI does a good job of removing it.  I just used the auto Screen Transfer Function (channels unlocked), and transferred the parameters to the Histogram Transformation tool when I was ready to move out of linear processing.  I just did some masked saturation boosts and ended up with reasonably good colour.

That is a very nice M31, nicely processed. yes interesting how many flats etc you used. i only do 20 flats & flat darks. it is a pain with the QSI583 as each one takes 22secs to download, and I do them straight after an imaging run, as dont like to risk moving anything in the imaging train. 

Fay

I only have to move the scope and camera about 10 feet from the patio to the lounge and it's pretty light weight.  I ensure the focus screw is locked and I also make sure the Canon auto-sensor cleaning setting is turned off (as that has ruined my flats before now!)  With care it works just fine, but I would not make flats the next day if I was putting the kit in a car at the end of the night for example.

I will try to write up the whole routine on my blog over the next couple of weeks (probably in instalments).

  • Like 1
Link to post
Share on other sites

Thanks for the tips! I didn't know you could use variable length darks in PI. Love the idea of putting the camera in the fridge to get darks, more time for imaging on a good night! Can't wait to get back out now!

Sent from my Nexus 4 using Tapatalk 2

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    No registered users viewing this page.

  • Similar Content

    • By astrobena
      Hey everyone,
      I was out recently in what felt like the first clear sky in years and got ~109 min of data on M31, minus 76 frames due to a 12mph wind, which left me with 69 min of data (each shot is 45 sec with ISO 200 tracked with skywatcher star adventurer). As mentioned in the title I captured all these images in a bortal 8 location, used an unmodified canon eos 400d and the skywatcher 75ed as the scope (with a flattener). I've attached my edit (warning: it is not great at all + slightly overedited to see what details are even there), and to be my surprise it looked very similar to an image of M31 with only 20 min of data which i captured a month earlier (both of which i used DSS and photoshop for). Now this may well have something to do with the way i edited it in photoshop or a different setting in DSS or just the fact that 49 more data doesnt make much of a difference considering im in a bortal 8 location, maybe you guys could help on that. I've attached the link to the original files (in the folder called 18.2.2021) as well as the stacked image from DSS (https://drive.google.com/drive/folders/12NT4TmLCXvTfOXNPE_l8UWPRpgO2VjLe?usp=sharing). I didnt capture any flat images but have dark and bias frames, all in their correpsonding folders in the attached link. It would be greatly appreciated if you guys could see if there is more data in this then i have managed to 'extract' using photoshop. (If you use different software and try and edit these files please tell me what you used) If there isn't then maybe do you guys have any images of M31 (or similar) from very light polluted skies that you could share here? (If so i would if you could share the full exposure time and gear that would be great)
       
      Many Thanks!
       

    • By AstroM1
      Andromeda Galaxy Group
      Nikon D5100 -- Nikkor 70-300 at 200mm
      60x30 sec + DOF
      Star Adventurer
    • By F137CHA
      M31 25 x 5mins Williams optics megrez 72 Cannon Eos 40d
    • By Lancebloke
      So, although I had some issues with my auto guiding (which I found out afterwards) I did manage to get 9x180s exposures and 5 darks of Andromeda and give both stacking (using DSS) and processing (using GIMP).
      My first attempted DSO....
       
      I am pretty sure that people could get far more information out of the TIFF file from the stack.
      Hopefully I will get another clear night soon. I am in a heavily light polluted area so I do have a clip in filter on my DSLR which I think took a lot away.
      More practise needed!
      Lance
    • By beka
      Hi All,
      Having had to move to an apartment where I could not use my CPC 1100, I decided that I have to see what I could do with my NexStar SLT 102 (alt-az achromatic 102mm f6.47 refractor). Setting up on my narrow balcony was challenging and the altitude bearing was so loose that it almost moved from the weight of the Canon 700D. I could not see M31 in the estimated 3.5 magnitude sky so I did a two star alignment and used the live-view to focus on a bright star. I then took a 15 second exposure after slewing to M31 which allowed me to see that I had it in the field of view. After a few more 15 second exposures and playing with the motion controls I managed to get it centered. The resulting picture is from 39 subs of 30 seconds at ISO 1600, 9 flats. The images were stacked and stretched with Siril and then I played with the curves on Gimp, cropped and scaled. Not too unhappy.
       

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.