M31 My best effort, help to improve please?

[DrRobin]

Hi,

Please find below a couple of images of M31 using different processing. These came from 50 frames of 2 mins each using a Skywatcher Quattro 200mm, Canon 600D, EQ6 and guided with PHD.

The seeing was generally good, there was some cloud but most of the subs came out okay. I stacked the best 80% after selecting the best one as the reference frame in DSS. I took 5 darks, 5 flats and 5 bias frames.

I don't have a coma corrector, looks like I need one next but any comments on how to improve would be gratefully received.

[SJJ]

There's a big gradient in the image, try using gradientXterminator if you are using Photoshop before processing the image too much. Again, if you are using Photoshop try using a layer mask on the core when stretching the image to try avoid burning it out. Also i would increase the number of calibration frames to maybe 20.

[nmoushon]

I would also suggest much more darks. I would try to shoot for 20 or so. will help keep the noise down when youre proccessing. Also when stretching with levels and curves do it in small amounts at a time. You'll have to do many more times but it helps to better control everything.

[Gina]

At least 20 darks and 20 flats I would say. I use something like 25-35 of each. Also you don't say what ISO you used.

[DrRobin]

Hi,

Thanks to everyone who has responded, some good advice there which I am going to give a try next time. Gina, I used ISO 800. My darks didn't show very much at all, suggests that 800 was okay? In fact I can see about 3 hot pixels. If I stretch the dark frame, I can see colour noise, but I guess this is normal?

I now realise I have made a couple of mistakes, No 1. mistake was I tried to set the exposure to get a good spread on the histogram in APT. However, I had forgotten that my camera is only 12 bit and I am guessing that the histogram is set for 16 bits? I spotted this when I did my flats, the peak didn't go more than 40% of the way across. This means that I almost certainly over-exposed leading to a compression of the scale, which then makes it difficult to get any contrast back in to the picture.

Mistake No 2. was not to use my Astonomik CLS filter. I can see light pollution in the original and think I would have got much better contrast if I had used this filter.

Based on what I can now see in the cool light of day, I reckon I need to reduce the exposure time to something like 60 - 80 seconds, possibly slightly more with the filter, take a lot more calibration frames, oh and buy a coma corrector.

[Gina]

Hi,

Thanks to everyone who has responded, some good advice there which I am going to give a try next time. Gina, I used ISO 800. My darks didn't show very much at all, suggests that 800 was okay? In fact I can see about 3 hot pixels. If I stretch the dark frame, I can see colour noise, but I guess this is normal?

800 is fine and yes colour noise is what you get if you stretch the noise.

I now realise I have made a couple of mistakes, No 1. mistake was I tried to set the exposure to get a good spread on the histogram in APT. However, I had forgotten that my camera is only 12 bit and I am guessing that the histogram is set for 16 bits? I spotted this when I did my flats, the peak didn't go more than 40% of the way across. This means that I almost certainly over-exposed leading to a compression of the scale, which then makes it difficult to get any contrast back in to the picture.

Yes, you want to make sure your subs are not saturating in the whites - you should be able to see some detail in the bright areas. When you have more experience you may wish to take two (or more) sets of subs, one to expose the bright parts and another with longer subs for the faint parts. Stack separately and then process as layers in PS (or whatever you use) and combine. There have been threads on this process elsewhere on here and links to tutorials etc.

Mistake No 2. was not to use my Astonomik CLS filter. I can see light pollution in the original and think I would have got much better contrast if I had used this filter.

I always use my CLS filter as it's very effective and doesn't affect the wanted light.

Based on what I can now see in the cool light of day, I reckon I need to reduce the exposure time to something like 60 - 80 seconds, possibly slightly more with the filter, take a lot more calibration frames, oh and buy a coma corrector.

Good thinking There's lots to learn.

[Jessun]

Robin, I'm guessing that you had a bit of drift between first and last sub, that you didn't dither, and that you took the subs with very little pause inbetween. The streaks throughout the image suggest this.

Using a 600D I got this same result until I changed to large dithering and waiting 20 - 30 seconds between subs. That really helps to combat the streaks.

You have a nice image there for sure!

[Shibby]

That's a really great start on M31 - you've captured plenty of good signal.

Gina, I used ISO 800. My darks didn't show very much at all, suggests that 800 was okay? In fact I can see about 3 hot pixels. If I stretch the dark frame, I can see colour noise, but I guess this is normal?

The darks actually add noise to your final image - they remove "dark signal" such as those 3 hot pixels. The other noise you are seeing when stretching the dark frame adversely affects your image when the darks are subtracted. This is why you need to get plenty of darks regardless of how many light frames you have gathered. The good news is, you can build a library of darks up and match them by temperature to your light frames. Saying that, I often don't use darks at all - the "natural" dither means that the dark signal stacks away quite effectively with enough light subs.

The gradient and colour balance can be fixed in post-processing. I had a quick play with the jpeg - rarely works well with a jpeg but gives an impression of how the LP can be reduced:

A good way of fixing the colour balance is to stack in DSS with "RGB background calibration" selected. GradientXTerminator can help with any gradient the flats have failed to remove, but M31 is probably the most awkward target in the sky for automatic gradient removal in post-processing!

[DrRobin]

Jesper - Yes only 5 seconds delay plus the download time so well short of your 25 seconds. That and dithering is something else I am going to have to look into.

Lewis - Quite a few things for me to get my head round and try. Your modified pic looks really good, more detail and much better colour so I think I might go back and try a re-stack with your suggestions. I suppose I could go back and take more darks tonight? The focus position hasn't changed, even though the scope is packed away and temperature seems similar today as it did last night, so that might work.

I am stacking my pictures on a server at home, so can easily add/change the stacking details and leave it to churn through, whilst I do something else.

Thanks everyone for the help, this is why I love this site, so many people with much more experience than me. I realise I am trying to run before learning to walk, sorry crawl, but hopefully I will get there.

With darks and flats does everyone take them at the end or part way through the session? If I take them part way through I can do a stack with some images on my server whilst capturing more on the laptop, that way I get an earlier view of how things are going and it may save time to have half of the lights already stacked?

[DrRobin]

Jesper, I have looked in to dithering and that now makes perfect sense to me, move the move a small amount between frames to get the light to fall on a different pixel and hence reduce the noise by averaging. I can see how you can avoid using a dark altogether if you get enough lights with a suitable spread of dithering, the noise on the sensor will be averaged out. This is what I am already used to with imaging planets and the sun using an AVI and Registax. I am still going to take darks and then try processing with and without to see which gives the best result.

I use APT with PHD and see that this has that facility so will definately give it a try next time, thanks for the tip.

[Jessun]

Precisely. I don't use darks on my OSC, but that chip is cooled and not too many pixels misbehave. I use a simple bias instead. A few hot pixels may remain, but dithering and finally some sigma reject stack takes care of those.

I'm sure there's plenty of advice on darks with a DSLR on the forum. I for one did find them worth the hassle.

[ollypenrice]

If that's your sky I'd go for Pixinsight. I only use a few of its tools, in truth, but the ones I use I need! Dynamic Background Extraction strikes me as being well ahead of Gradient Xterminator and can rescue the most gradient infested of images. Gradients are rare here but I try to help on other images as well and DBE is outstanding. It balances colour while removing the gradient and is highly adjustable.

You've got lots of great little core details there, by the way.

Olly

[DrRobin]

Since everything was still set up I took 20 darks, 20 flats and bias shots on Friday night. I have re-processed and improved the image a bit but as I can't do anything about the over-exposure, try as I might I can't improve on Lewis's version.

I had another go last night, but after a late start nothing went well, my camera battery went flat, both spares quickly went flat, both batteries for the mount went flat, to cap it all even my torch battery went flat!

My guide camera wouldn't work, had to use a colour camera, I got everything set up only to knock the focus whilst changing the camera battery and then my secondary fogged up. I only managed 10 subs before the cloud rolled in.

Still I did get PHD to dither, I tried my filter and I didn't over expose, so progress. Thanks everyone for your help.

[DrRobin]

Hi,

I decided to have an extensive scope tune up session last weekend and only managed to completely set the secondary in to the wrong place. To correct the problem I found an article on using a webcam and a piece of software called Mire de Collimation which is a great way of getting the secondary in the correct place.

Once done, it was partially clear last night so I decided to do a star test. After a final tweak with the laser, the star test showed perfect collimation. Then to my surprise the skies cleared so decided to have another quick go at M31. I didn't use my filter or guidescope as none of that was set up, but did manage to take 60 subs at 60 secs on 800 ISO with 10 darks and 10 flats, sorry ran out of time. The picture shows great improvement and I am quite pleased with it.

I have a coma corrector on order, might be here today, so next time it is clear a full test with guiding, dithering and my CLS filter, should just about get me there.