Andromeda

[edarter]

Taken over 5 nights a few weeks ago, I managed to grab about 8hrs on Andromeda. By far the most data I have collected so far on any target. The best 6.5hrs of that data stacked and initial processing in APP then finished off in Photoshop. I'm convinced there is a lot more data hiding in there somewhere but as I'm only in the foothills of the mountain that is Photoshop processing this will do for now

 

[AstroNtinos]

Very nice picture. Good job.

[wimvb]

Very nice indeed. I think that if you spend some time on getting the colours  balanced, you'll be surprised how much more there is in the data. There are hints of blue star clouds and Ha regions in your image already.

[edarter]

7 hours ago, wimvb said:

Very nice indeed. I think that if you spend some time on getting the colours  balanced, you'll be surprised how much more there is in the data. There are hints of blue star clouds and Ha regions in your image already.

Thank you

Yeah, struggling with colour balance. The curves seem aligned in photoshop but I think it's a bit too red. Is it a case of just eyeballing it and ignoring where the curves are? 

[Phillyo]

Curves are accurate so they should be lined up, however if you use the info button you should be able to hover around various points in the image and it'll give you an RGB value. Try to get the RGB at somewhere around 25, 25, 25 to 35, 35, 35 or as close as you can. That'll neutralise the colour cast in the image.

[edarter]

13 hours ago, Phillyo said:

Curves are accurate so they should be lined up, however if you use the info button you should be able to hover around various points in the image and it'll give you an RGB value. Try to get the RGB at somewhere around 25, 25, 25 to 35, 35, 35 or as close as you can. That'll neutralise the colour cast in the image.

Would I do that with a levels adjustment applied to the whole image? And does that not affect the curves? Sorry for the basic questions! 

[Phillyo]

Stick with the individual R G and B curves to try and get them even-ish. Don't worry too much if they're a little out as long as the background has a general 25, 25, 25 to 35, 35, 35 reading or as close as possible. It's tricky doing it manually. I use Pixinsight so it's much easier.

[wimvb]

I downloaded and examined your image. The histogram peaks are aligned close enough, but that only got you a fairly neutral background. You also have to get the white point right. I don’t know enough about astro pixel processor, but suggest you have a look at their tutorials.

[edarter]

Hmm... OK will have a look through the tutorials, thanks. What sort of value should I aim for with the white point? I'm assuming it should be less than 255 then? 

[tomato]

Great M31, the colour does look a bit pink to me. Have you tried the Calibrate Star Colours tool in APP? It’s pretty straightforward to use and generally does improve the colour balance on objects such as galaxies.

[wimvb]

9 hours ago, edarter said:

Hmm... OK will have a look through the tutorials, thanks. What sort of value should I aim for with the white point? I'm assuming it should be less than 255 then? 

What @tomato wrote. Oddly enough, you can't use a white point (255, 255, 255) as reference, because that may leave midtones with a colour cast. PixInsight takes (certain) non saturated stars as white reference, or the entire galaxy. I assume that the Calibrate Star Colour tool in APP does something similar. This is best done in the linear stage, before stretching throws the balance for midtones off.

[edarter]

Thanks for the tips guys, I spent the evening yesterday experimenting with different things I've seen on YouTube but have to say it didn't get me very far. I shall have a look at the calibrate star colour tool in APP and see what that does. APP stretches the image automatically, though you can turn the stretch off when you save. Would I be better saving a non stretched version and working on that in photoshop? 

Cozzy also messaged me with a quick edit he did of my jpg and I'm blown away by it, so much better than my effort here. Its spurred me on to try and nail the processing better because its shown there is indeed way more data to be had out of this image! 

Thanks again for the advice and encouragement, it's all helping! 

Ed

[wimvb]

1 hour ago, edarter said:

Would I be better saving a non stretched version and working on that in photoshop? 

I don't use APP, so can't advise on its use. I work exclusively in PixInsight. So, assuming that APP can work in a similar way as PI:

In PhotoShop you have to apply a permanent stretch in order to see anything on screen. PI (and I assume APP also) can apply a screen stretch. This means that the image is stretched on screen, but not in reality. Any processes you apply are done on the unstretched (linear) image. Only when you apply a permanent stretch, will it be applied to the image. All processes after that will be applied to the stretched non-linear image.

If you upload the tiff (or fits) file as it comes from your stacking program, ie without any other processing applied and unstretched, I can have a look at it.

[michael.h.f.wilkinson]

1 hour ago, edarter said:

Thanks for the tips guys, I spent the evening yesterday experimenting with different things I've seen on YouTube but have to say it didn't get me very far. I shall have a look at the calibrate star colour tool in APP and see what that does. APP stretches the image automatically, though you can turn the stretch off when you save. Would I be better saving a non stretched version and working on that in photoshop? 

Cozzy also messaged me with a quick edit he did of my jpg and I'm blown away by it, so much better than my effort here. Its spurred me on to try and nail the processing better because its shown there is indeed way more data to be had out of this image! 

Thanks again for the advice and encouragement, it's all helping! 

Ed

I tend to save the non-stretched version in FITS format and fiddle with curves in GIMP. APP does all processing on the linear version, which is generally the best, I find. The automatically stretched version is often rather good, so I do sometimnes save that as a jpeg to post a quick result

[edarter]

Hi All,
Well after 2-3 weeks of constant trial and error, then getting 'Dark Art or Magic Bullet' I've spent a lot of time over Christmas playing with my Andromeda data. Its better (I think) but I'm still not happy with it. I see images on here of other peoples Andromeda with much less data and way more colour. I'll keep experimenting though and see what I can pull out of it.
Main differences here are that I calibrated the star colours in APP and then saved the unstretched version ready to edit in photoshop. I then used a mix of procedured in DAOMB and what I've picked up from Trevor Jones and some of the other videos I've seen on Youtube.

In fact I've just noticed how different it looks on each of my two monitors! So not sure what it will look like to everyone here. Two identical monitors here but looks a bit over exposed on one of them. Can afford a calibration gizmo so will have to investigate other methods of at least getting it half right between them 😕

[geeklee]

1 hour ago, edarter said:

Well after 2-3 weeks of constant trial and error, then getting 'Dark Art or Magic Bullet' I've spent a lot of time over Christmas playing with my Andromeda data. Its better (I think) but I'm still not happy with it

Great work @edarter - that definitely sounds like time well spent on an already good image - a noticeable step forward in the detail, colours and general balance showing above.  Very nice 👍