M31- Andromeda Galaxy

[gonzostar]

Hi

M31 taken with the following details

14 hours worth of 5min lights frames taken over several nights

With darks, flats and bias frames

 

I used the 102mm APO ES refractor on a AVX mount. Camera used was a canon 450d (un-mod camera) with a CLS filter. 

Guided with PHD2 and processed with PS. 

I am sure there are many ways to improve this image, but please let me know your expertise  

 

Cheers 

Dean

 

[Guest Tuomo]

Hi!

Spiral arms are nicely visible in this image. I guess its the total exposure time of 14h. I really need to aim for these kinds of hours. Only -and the biggest- issue is the blown out cores of meda and both of those smaller galaxies. Maybe you did something in the processing that caused galaxy center to be blown out (just pure white, nothing intresting). Maybe your sub exposure time is too much? What ISO you used? I guess with that camera max iso should be 800.

[gonzostar]

Hi thanks for feedback. 14 hours is a long run but worth it in the end. 

Camera setting was iso 800, I agree that sub exposure time of 5mins could have blown the cores  Also i guess i need more practice with Photoshop. Iys all a enjoyable steep learning curve

Cheers

[gonzostar]

Hi

I have  attached the  mosaic TIF file if anyone fancies having ago  

 

Cheers

Dean

14hr-mosaic-dssinitial.TIF

[wimvb]

Great result, Dean.

You can easily check if the core is overexposed. If the unstretched stacked image already looks overexposed, you're out of luck. My guess is that it won't be.

You can then just have two copies of the same unstretched image. Stretch one for details in the core, and the other for details in the outer arms. Then put both as layers on top of each other. With a suitable mask, blend the two copies. I don't use ps, but I'm sure there is a tutorial somewhere describing this procedure in more detail. Google the name Jerry Lodrigues (spelling?)

Btw, your image looks very blue on my screen. That's easily fixed with colour balance.

I may have a go at your image this evening. I'm all clouded in anyway.

Cheers,

[gonzostar]

Thanks Wim, Its being cloudy overhere for the past week also! 

Will research into layering and combining layers. In the initial stack the core didnt look overexposed. I guess i overstretched to get details of outer areas.

Blueness could that be because i was using a CLS filter. Also one of the nights the moon was around 70% wanning! Do you also think that a astro-colour camera from ZWO may enhance my pictures?

Cheers

Dean

[wimvb]

1 hour ago, gonzostar said:

Thanks Wim, Its being cloudy overhere for the past week also! 

Will research into layering and combining layers. In the initial stack the core didnt look overexposed. I guess i overstretched to get details of outer areas.

Blueness could that be because i was using a CLS filter. Also one of the nights the moon was around 70% wanning! Do you also think that a astro-colour camera from ZWO may enhance my pictures?

Cheers

Dean

Blueness may be caused by the moon. Moonlight, like sunlight, makes the sky blue. Just less bright. That's why you can do Ha (red) imaging in a near full moon.

The advantages of an astro cmos camera are:

1. more sensitive than most dslrs

2. bayer matrix already removed (mono camera)

3. cooled, so lower dark noise

4. lower read noise

The main disadvantage I have found, of ZWO cameras is the amp glow. This causes a bright area in part of the image. It's relatively easy to calibrate out with dark frames. But it increases the noise, and decreases the dynamic range.

Having had an unmodified dslr that I couldn't connect to my computer, I find an astro camera a real improvement. Nowadays nights, I don't even do star alignment anymore. I just use the plate solving routine in Ekos/Kstars to get accureate goto. Framing, focusing, filter change and imaging is controlled from Ekos. The automation has decreased my setup time, so I get more time for imaging = more subs = better quality.

If you go for an osc (colour) camera, you don't have the filters (which help some in decreasing light pollution effects), but you still have the other advantages.

Imo, if you have access to dark skies, osc will work. But if you are in a light polluted area, mono is probably better. Mono also has the advantage of enabling narrow band imaging, which is much less LP sensitive. This is the main reason I chose a mono camera over an osc.

Hope this helps.

Edit:

I just downloaded your stacked image. There was quite a lot of rotation between the subs. This can cause edges where the subs overlap, and these can become visible in the final image. To avoid this, you can align the camera sensor with the RA direction:

take a 15 - 30 second exposure. Five seconds into the exposure you slew the mount at 1 x sidereal rate in RA+. This will give you star trails with a larger star at the beginning of each trail. The star trail should line up with the image edge. If it doesn't, rotate the camera 'towards the trail'. Take a new exposure and repeat until the star trails are parallell to the edge of the image. Align either the long or short side of the image with RA, depending on the orientation of the target. Don't forget to recenter the target after each exposure. You also need to refocus after this alignment.

If you do this as a part of your setup, you will find it much easier to combine images from several nights. You should be able to get an alignment within a degree or two.

[wimvb]

Here's my go at your image:

Processed in PixInsight

- dynamic crop to get most of the stacking edges off. There was a trade off between edge removal and keeping the galaxy

- removing background gradient with dbe

- applied a simple stretch

- restoring detail in the core by using hdr multiscale transformation

- darkening the background with curves transformation

- star reduction

- resample

There was very little colour in the image, so I left that as it was. Colour saturation only resulted in a blue background and a pink core.

But since this image was more about whether the core was blown out, and to increase the dynamic range, I just did intensity processing.

You might want to check your debayer settings in DSS.

(click to enlarge)

[gonzostar]

Thanks Wim for all efforts and advice. Actually the image looks more impressive to me as  you have done it. Would increasing overall exposure time to say 20hours +  bring out more colour? or is it the fact that i am using a CLS filter. You mentioned about debayer settings in DSS, What setting is recommended. Should be it on super pixel?

 Looking at the mono route would be more ideal for  my location as there is lot of light pollution. At the moment there is the cost issue with this. The wife wouldnt be to happy  £1000+  So guess keeping to the DSLR for the time being. Maybe get a newer model. 

 

Also will try aligning camera sensor to RA, Possibly this weekend fingers crossed for a clear window despite the moon.

 

Cheers

Dean

[wimvb]

UHC filters (Ultra Hight Contrast?) that are used for nebulas, can strongly affect colours in galaxies. But I've seen several images stacked with DSS that were quite colourless. First, make sure that you use the correct bayer matrix (RGGB, BGGR, GRBG, etc). I wouldn't use super pixel mode, but rather bilinear interpolation or AHD interpolation.

Then if you post a stacked image with no adjustments applied (only embedded), it will be easier to process in another program.

Added integration time will mainly limit the noise, but not necessarily bring out more colour.

If you have the Astronomik CLS filter, it has this characteristic:

The colours that this filter passes through are mainly blue and green. The red that it passes through is very deep, from Ha and longer wavelengths (infrared). The problem is that a dslr that is not astromodded has its highest sensitivity at about 500 - 550 nm, which is in the green. At 650 nm, which the filter lets through, the sensitivity of the camera is so low, that very little red registers. On the other hand, with a modified camera the sensitivity to Ha (656 nm) is much higher, and more red will show in the image.

This image shows the transmittance of several filters. The Astronomik CLS is at the very top (below the spectrum of mercury, Hg) It shows how Blue/Green and deep red are passed through.

I used the Baader UHC-S filter (fourth filter from the top) with my Pentax dslr. This filter lets less blue/green through, but more red, so it has a better colour balance for galaxies. I also was lucky enough that Pentax cameras generally have a slightly better sensitivity in the red than most Canon cameras. So, for me the Baader UHC-S filter was a good match and I never had much colour issues.

If you're on a budget, you should consider having your camera astromodded. It will cost, but with the right modification it is still usable for daytime photography, and the increased sensitivity to red (Ha) should also result in better images with the filter in place. Another solution would be to buy a used modified camera.

[ziofrancotto]

15 hours ago, wimvb said:

You can then just have two copies of the same unstretched image. Stretch one for details in the core, and the other for details in the outer arms. Then put both as layers on top of each other. With a suitable mask, blend the two copies.

Interesting: I always thought was better to blend different exposures with different time lenght

[gorann]

Nice to see that the core is still there!

I strongly suspect that your CLS filter is the reason for loosing colour. My first light pollution filter was also a CLS filter and M31 looked as anemic as yours. I then invested in what is often considered to be the best light pollution filter: Hutech IDAS LPS P2. It is much more specific at blocking the unwanted light, see spectrum below. I am quite sure you do not need more exposures. 14 hours should be more than enough for both colour and detail. Below you can also see what I got out of only three 8 min exposures with a DSLR (Canon 60Da) at ISO 1600 (it was in November so a cool camera with not much noise). This was when I started astrophotography in 2015 and had just changed from CLS to IDAS. I was so eager checking out new targets back then that total exposure time suffered terribly. Nowadays I would spend at least a few hours on M31 (probably not 14 thought) and also do some shorter exposures (1 min or so) to get as much core details as possible).

Cheers

[gonzostar]

Cheers Wim & Gorann

Very informative and useful graphs. It looks like its worth experimenting with suggested filters.  I know you cant beat a dark site which is not a option where i live. I went to Devon a couple of weeks ago, and took the binoculars. The milky was so awesome. The best i have seen since i was in Portugal in 1992!  

Also thats a great picture of M31. Shorter then 5 minute exposures compared to 1minute  looks to make such a difference to the core.  Yet another lesson learned.  I  notice with the badder UHC-S filter it cuts off a bit of green towards th LP cutoff. Is that not to impotent. But then you gain at the red end is that more importents for nebula work? Then the IDAS filter looks good to. 

Cheers 

Dean

[ollypenrice]

On 06/09/2017 at 23:48, ziofrancotto said:

Interesting: I always thought was better to blend different exposures with different time lenght

Sometimes it is, but most of the time two stretches of the same data will do fine.

Olly

[gonzostar]

Going off topic slightly  Does this apply to globular clusters aswel? One exposure time for the core and another for the outer stars. Then blend them together in PS.

[wimvb]

I believe that @mike005 used this technique not so long ago on a globular cluster. It was chosen as a NASA apod. 

[gonzostar]

Thanks for the tip Wim. Although i cant find Mike005 globulars  

[wimvb]

https://apod.nasa.gov/apod/ap170711.html

[gonzostar]

Thanks Wim. Veery impressive image!

[gorann]

If you have a partially really bright target, M31 and the Orion Nebula being prime examples, I think it cannot hurt also shooting a series of short exposures. Does not take long and you will not have to bother with flats or darks or whatever since it is bright and you will only use it for selected parts of the image

[ollypenrice]

10 hours ago, gorann said:

If you have a partially really bright target, M31 and the Orion Nebula being prime examples, I think it cannot hurt also shooting a series of short exposures. Does not take long and you will not have to bother with flats or darks or whatever since it is bright and you will only use it for selected parts of the image

Certainly M42 needs short exposures but M31 surprised me. This is from a set of 30 minute luminance subs at F5. Multiple stretches of the same data, but the camera (an Atik 11000) does have excellent well depth.

However, I absolutely agree with your recommendation to experiment because there are so many variables in this game.

Olly

 

[gonzostar]

Again fantastic image, one day next millennium i might get to this standard  Part of my issue also is i need to practice on PS. And put into practice new ideas

 

Dean