M31 burn out

[Philip Benson]

Last night I did some test exposures on M31.   Because of the wide range of surface brightness levels I took some very short exposures of about 20 seconds,. some mid range exposures of 120 seconds and some long exposures of 300 seconds.

In the case of both the 120s and 300s exposures the core region was completely saturated.   I expected this but I know there are processing techniques which can correct for this.   I have just tried processing these out using the DPP tool in Maxim but I notice that there is quite a significant amount of white clipping in the histogram.   Does this mean that the saturation is rather more permanent?

Is there anything I can do to correct this using Maxim, Photoshop or any other software tool?  I am becoming increasingly tempted by Pixinsight.

Equipment used:   Tak FSQ106ED with QSI 690wsg camera.

[ollypenrice]

Interesting question. I've spent a very long time on M31 and have some views which may not fit the orthodoxy. Here goes...

The core of M31 is very bright. (You amaze me Holmes!  ) The dust lanes within the core of M31 are very faint. (This does not receive suffient consideration...) How do you capture very faint data? You go for long exposures. Aarrgghh. We're doomed! So the problem is not to stop the core being very bright, the problem is that, within the bright core, there is data which is very faint. So...

In my view, using the deep wells of my 'insensitive' Atik 11000, I can do quite well with long exposures and the way to extract spiral structure into the core is to use brutal, shameful, bludgeoning techniques with regard to contrast enhancement. Very short subs didn't prove very helpful. it was in medium length subs that I found that I could pull out the faint spiral forms heading in towards the core. I was also interested in the very faint outer reaches because I'd once seen an image in which they showed a 'turned up' end. I never did relocate that image but I did find that feature for myself after much effort.

https://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-xbvjFDF/0/O/M31%20Outer%20HaloLHE.jpg

This image now needs to have its resolution improved by adding luminance data from a longer focal length but the core is about at the limit of what I can do with present knowledge. I certainly could never do a 'one stretch fits all' for this target. For me the core needs its own processing.

Olly

[pipnina]

Interesting question. I've spent a very long time on M31 and have some views which may not fit the orthodoxy. Here goes...

The core of M31 is very bright. (You amaze me Holmes!  ) The dust lanes within the core of M31 are very faint. (This does not receive suffient consideration...) How do you capture very faint data? You go for long exposures. Aarrgghh. We're doomed! So the problem is not to stop the core being very bright, the problem is that, within the bright core, there is data which is very faint. So...

In my view, using the deep wells of my 'insensitive' Atik 11000, I can do quite well with long exposures and the way to extract spiral structure into the core is to use brutal, shameful, bludgeoning techniques with regard to contrast enhancement. Very short subs didn't prove very helpful. it was in medium length subs that I found that I could pull out the faint spiral forms heading in towards the core. I was also interested in the very faint outer reaches because I'd once seen an image in which they showed a 'turned up' end. I never did relocate that image but I did find that feature for myself after much effort.

https://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-xbvjFDF/0/O/M31%20Outer%20HaloLHE.jpg

This image now needs to have its resolution improved by adding luminance data from a longer focal length but the core is about at the limit of what I can do with present knowledge. I certainly could never do a 'one stretch fits all' for this target. For me the core needs its own processing.

Olly

I think you've pulled out more contrast in the core than I've seen in hubble images.

A question, though, with galaxy images we rarely see the dust / molecular clouds in our own galaxy that obstruct the view, but I've seen huge expanses of the stuff in images of the pleiades, orion, cygnus and M81/82, does this image just not go deep enough or is there no dust / moecular clouds around the M31 area?

[Philip Benson]

Thanks for the feedback.   I have never dabbled very far with image processing so I wasn't sure whether there was some magical solution to this.  PixInsight is seemingly the post popular processing software now but I have always only used Maxim DL5 and Photoshop.  My knowledge is limited with both. 

​When I look at the 300s exposures the core region is as you would expect just a white elliptical shape.  This stays the same regardless of what DPP settings I use in Maxim.   As I understand it the histogram is what tells you how the brightness in the image is distributed and this in turn gives an indication of how much you can get out of the image.   The white clipping suggests to me that the software is seeing the core pixels simply as fully white and therefore has no idea what to do with them.   This is why I feared that I probably cannot do anything with these particular images.

I also took some subs through the R,G and B filters and these show a much more even histogram.   Yes the core still a little saturated but much less so than in the luminance subs.   The mission last night was simply to see how the FSQ performed with my 690.   However as always you end up opening up a whole new challenge for yourself.   The QSI 690 incidentally has a relatively shallow well depth of about 15,500 so I wonder if that is affecting my results.

[ollypenrice]

When building up an LRGB image there is a lot to be said for using the RGB-only for handling the brightest parts. Bright stars are the obvious example. The RGB does not go anywhere near as deep, per unit time, as th L. (That is the whole point of shooting L, after all. So the RGB takes longer to saturate.)

Pixinsight is not the most popular DS imaging package, I don't think. I would suggest that the most popular is still Photoshop, but PI has a lot to offer. I refuse to be drawn into a holy war about this and use both. Ps suits me better but PI suits others better. I couldn't do an image without both.

In my view you can tweak DDP till you are blue in the face but DDP is one stretch and I don't believe high dynamic range targets can be handled in one stretch. I find I need multiple stretches layered carefully in Photoshop. There are lots of layer-o-holics out there! My name is Olly and I'm a... etc    Maybe somebody knows a one stretch approach but I don't!  

Olly