Processing M81 M82 NGC3077 image - hours of data.

[tooth_dr]

Hi

I've have been processing this image for a few hours now and it still just doesn't look right.  Captured on 24th and 25th February.  I added Ha to the red channel at about 50% using lighten mode.

22x600s + 100x300s with CLS filter 1000d

60x600s with 12nm Ha filter 40d

 

Advised welcome so I can change my approach either with capture or processing.  I've attached the colour stack straight out of DSS converted to 16bit and same for the Ha stack.

 

Many thanks

Adam

 

 

AutosaveSGL_Ha.tif

AutosaveSGL_RGB.tif

[MarsG76]

It doesn't look right?? Whats wrong with it?, looks fine to me.... I think that galaxies look excellent.... I'm always happy to see another DSLR imager, especially when they're coming out with top quality work.

 

[tooth_dr]

5 hours ago, MarsG76 said:

It doesn't look right?? Whats wrong with it?, looks fine to me.... I think that galaxies look excellent.... I'm always happy to see another DSLR imager, especially when they're coming out with top quality work.

 

Thanks! It just looks a bit brown, other images have more blue but I can’t seem to make that happen! I think I messed up the initial colour balancing at the start of the processing process (what I think anyway!)

[MarsG76]

5 hours ago, tooth_dr said:

Thanks! It just looks a bit brown, other images have more blue but I can’t seem to make that happen! I think I messed up the initial colour balancing at the start of the processing process (what I think anyway!)

You can always color balance or add blue now in photoshop/gimp...

 

[tooth_dr]

9 minutes ago, MarsG76 said:

You can always color balance or add blue now in photoshop/gimp...

 

Tried what I could but failed lol

[Cozzy]

38 minutes ago, MarsG76 said:

You can always color balance or add blue now in photoshop/gimp...

 

I hope you don't mind, but just had a quick play with your Jpeg whilst at work, the boss isn't watching. I used Microsoft office cos that's all I have here, slightly reduced saturation, slight lighten and contrast and pulled it over to the blue a bit and lifted the mid-tones. the core is a bit blown but I don't have access to my Photoshop. there is a lot you can with that image.

So the old adage: if at first and all that. good luck with your re-processing and best regards

Tim 

[Allinthehead]

Hi Tooth. Did you add the ha to the whole image or just to the galaxies?

It should only be added where needed, you can add it in to a duplicated image. Copy and paste original over the duplicate, then use your eraser where needed at whatever opacity you feel it can take.

Hth 

Richard.

[tooth_dr]

44 minutes ago, Allinthehead said:

Hi Tooth. Did you add the ha to the whole image or just to the galaxies?

It should only be added where needed, you can add it in to a duplicated image. Copy and paste original over the duplicate, then use your eraser where needed at whatever opacity you feel it can take.

Hth 

Richard.

Thanks Richard.  I copied out the red channel from the final RGB image, and pasted that onto the alpha image.  I lined it up, did a 50% 'lighten' blend, then flattened the image.  I copied and pasted that back into the RGB image as the new red channel!  I think this is the first time I've tried this!

[Aidan]

I've just had a very quick go in Pixinsight..  Lots of data there to manipulate and play with.  I've probably been a bit to aggressive with the background and made it too black, but certainly room there to get rid of the brown/red tinge you have in your image.

Keep plugging away with it, there is a very good image just waiting to pop out

[tooth_dr]

21 minutes ago, Aidan said:

I've just had a very quick go in Pixinsight..  Lots of data there to manipulate and play with.  I've probably been a bit to aggressive with the background and made it too black, but certainly room there to get rid of the brown/red tinge you have in your image.

Keep plugging away with it, there is a very good image just waiting to pop out

Thanks Aidan. The moon intereferred but still the guts of 25 hours data. 

[wimvb]

A simple process to show what's there.

Just the rgb image, processed in PixInsight:

1. crop

2. DBE

3. colour calibration (standard method: background neutralisation and colour calibration)

4. Mark Shelley's arcsinh stretch

5. colour saturation with a range mask to protect the background.

Since the Ha image didn't seem to be aligned with the RGB image, I haven't used that.

Images stacked with DSS have a tendency to be colour anaemic. I think you're right about the moon interfering. After removing the background signal, which was very high intensity, there was a lot of noise left. This is typical of images taken in a light polluted/moonlit situation. I would suggest trying to image the RGB again when the moon graces us with its absence.

[wimvb]

I've had a closer look at your data. There's good and there's bad in it.

The bad: the Ha doesn't add anything to it. All the Ha information seems already in the red channel.

The good. The RGB image contains some unexpected detail. After massive gradient removal, including trying to reduce what's probably an unevenness in sensor response, your data reveals Arp's loop. Imo, it's amazing that you managed to get this unde so much light pollution/moonlight. But here it is:

Arp's loop is just about visible as an extremely faint ring extending from M81 to the lower left. There's also a weak structure visible just above M81.

That this is real structure and not just a noise artefact can be verified by comparing to this image from Terry Hancock.

https://www.flickr.com/photos/terryhancock/9025623984/lightbox/

The real challenge is of course to reveal it in the RGB image.

My advice: reshoot RGB with no moon. You could get a real winner.

[tooth_dr]

12 minutes ago, wimvb said:

I've had a closer look at your data. There's good and there's bad in it.

The bad: the Ha doesn't add anything to it. All the Ha information seems already in the red channel.

The good. The RGB image contains some unexpected detail. After massive gradient removal, including trying to reduce what's probably an unevenness in sensor response, your data reveals Arp's loop. Imo, it's amazing that you managed to get this unde so much light pollution/moonlight. But here it is:

Arp's loop is just about visible as an extremely faint ring extending from M81 to the lower left. There's also a weak structure visible just above M81.

That this is real structure and not just a noise artefact can be verified by comparing to this image from Terry Hancock.

https://www.flickr.com/photos/terryhancock/9025623984/lightbox/

The real challenge is of course to reveal it in the RGB image.

My advice: reshoot RGB with no moon. You could get a real winner.

Great stuff Wim. I was desperate to image something as it was clear. But I knew it would impossible to get anything good from my past experiences. I tried Ha obviously as it’s ususlly ok in moonlight, and RGB subs were 300s the second night when the moon was brighter. I’ll try it again for sure.

I appreciate you taking time to look at this. I’ve been trying lots and every time my image looks brown despite being balanced in PS using the sampler/dropper tool. 

[wimvb]

2 minutes ago, tooth_dr said:

Great stuff Wim. I was desperate to image something as it was clear.

I know the feeling ...

But waiting for clear moonless nights will be worth it.

[tooth_dr]

I took your Jpeg and added it to my image which was brown, but I like your colours:

 

[tooth_dr]

9 minutes ago, wimvb said:

I know the feeling ...

But waiting for clear moonless nights will be worth it.

 

I've nothing against the moon per se, save that I have absolutely no interest whatsoever in imaging it!

[wimvb]

Me neither.

[tooth_dr]

A final attempt using APP this time, before I put this one to bed.

 

[wimvb]

... and a final attempt using PixInsight.

No Ha was harmed in the making of this photograph. (Or used, for that matter.)

 

[ollypenrice]

It's a good image but I have some suggestions regarding the addition of Ha.

Once you have it placed over the red channel in Blend Mode Lighten, set the opacity to 100% and leave it there. More on this later...

Use the eyeball to blink it on and off to see what it's adding. Does it lighten the background? (My guess is that it did in your first attempt.) This must be avoided or you'll get a red sky, so you can clip the black point in a bit or drop the very bottom of the Curve while looking at the result in real time. It must not affect the background. Now is it adding enough in its regions of strong signal, like the Cigar's jets? If not, pin the Curve low down but lift it higher up till the Ha comes into play. Only the bright parts are getting into the red channel so you don't have to worry about noise. Remember that this is 'Ha for red' and not a standalone Ha image, so you can process it very differently. 

Flatten and call it something like HaLRGB.

Now is the time to decide how much of the Ha you want in the final image, so copy and paste it onto a copy of the original LRGB. You can vary the opacity to select how much Ha you want. Also, if the Ha has damaged parts of the image (stars, background, whatever) you can simply erase it. Nothing could be easier.

Why do I love layers??? 

Olly

[spillage]

Sorry on work laptop so very poor and had to bin the image. But here goes for my 5 min effort in startools. @wimvb has managed to pull much more detail out than me.

[wimvb]

4 hours ago, spillage said:

But here goes for my 5 min effort in startools. @wimvb has managed to pull much more detail out than me.

 But in all honesty, I spent quite a bit more than 5 minutes pulling.

[tooth_dr]

Amazing guys. I feel even more hopeless at processing now. 

[spillage]

Do not feel hopeless. I keep learning new things all the time with startools even after using it for a year. What software do you use ? It might be worth trying others as most come with some form of free trial, you can then choose the one that you find easier to use.

I have never been much good at using photoshop and although I appreciate pixinsight is a great bit of kit I found it a bit odd and very set in the way it works. I found startools very relaxed in the way I can jump from one process to another in any order I feel like.

I have been looking at astroart but cannot find that many guides.

[wimvb]

5 hours ago, tooth_dr said:

Amazing guys. I feel even more hopeless at processing now. 

Don't. It just means that you have good control ovet data gathering. And once you have the data, you can relax and learn the processing part. The data isn't going anywhere. Just consider my and @spillage's processing efforts as a quality check of your data, and examples of what is possible.