M31 in Black and White (68MP 6 panel mosaic)

[pietervdv]

Hello all,

I have finally processed my imaging data from this summer's (August) visit to Southern France; there I joined my trusted imaging companions Frans Kroon and Maurice Toet for a great week.

I'd set my sights on a 6 panel mosaic of M31; each panel is exposed for 2 hours, so 12 hours in total using a Luminance filter. Some nights were plagued a bit by high clouds so I had to deal with mild gradients and noise in some panels.  My processing skills are also a bit rusty, but I hope I managed to make it look ok. 😀 I did not do any major processing on it, mere stretches and curves and stitching the panels together.

I always have a preference for b&w as it reminds me of the works of the early astrophotographers (another explanation is; I am lazy and don't like to tweak endlessly behind a computer😎 )

 

Equipment: Homebuilt 10" f/3.8 Astrograph with Wynne corrector on a Mesu200 mount

Sensor: ASI-1600MM-cool cmos -10°C (it was T-shirt weather all night long)

Exposure: 6 x 2 hrs (5 min subs for all 6 panels), totaling 12hrs.

Filter: Baader Luminance

Capture and calibration: Sequence Generator Pro and Astro Pixel Processor.

 

You can find the highest res version on my website, the full size is 68 MP.

Full res M31

 

Thanks for checking it out,

Regards,

Pieter

Edited December 8, 2019 by pietervdv
Updated link

[groberts]

That's a truly great image of M31 and very interesing in black & white.  Well done

[ollypenrice]

That's really two pictures in one, both stunning. The small one posted here is beautifully balanced and natural whereas the full res is something you explore a bit at a time with wonderful tiny details to discover. Great stuff Pieter. I'd encourage other members to check out your website, too, for unusual targets superbly imaged.

Olly

[tooth_dr]

1 minute ago, ollypenrice said:

That's really two pictures in one, both stunning. The small one posted here is beautifully balanced and natural whereas the full res is something you explore a bit at a time with wonderful tiny details to discover. Great stuff Pieter. I'd encourage other members to check out your website, too, for unusual targets superbly imaged.

Olly

For sure, definitely click on the full size image link and have a good look around.  Beautiful image.

[vlaiv]

Really remarkable images (both scaled down version and full size).

Hope you don't mind me noticing following, but really that is in my view something that is robbing your work of perfection. If it were not for those small details, I would really consider above images perfection in AP.

You are slightly oversampled (at 100% large image does not render stars as pin points, but rather there is softness to them - which means slight oversampling. There is also issue of "fat" diffraction spikes which means atmosphere was not playing ball and there is no detail to justify such resolution). Going with lower sampling rate will improve SNR further - which is great already, this is probably the best rendition of this galaxy that I've seen (not talking about M31, but this little fella):

This is first time I've clearly seen bar in this galaxy and how it's twisted.

Going with lower pixel scale would give you additional SNR and smoother background while all things would still be visible in the image.

Second thing is obviously edge correction of your setup. It is fast newtonian astrograph and sure it's going to suffer some edge softness on larger sensors, but in this case, it sort of hurts mosaic because overlaps can be easily seen, like this:

also, not sure what software you used, stitching is not quite perfect - as this part shows:

And the third one is obviously blown cores of M31 / M32.

I'm aware that I might be considered too harsh with my comments since you produced some splendid images, but I do really think that above things can be easily rectified (add a few filler exposures for the cores, be careful about stitching / blending part and do bin of your data in software) and then you will be closer to perfection in your work.

[pietervdv]

1 hour ago, vlaiv said:

Really remarkable images (both scaled down version and full size).

Hope you don't mind me noticing following, but really that is in my view something that is robbing your work of perfection. If it were not for those small details, I would really consider above images perfection in AP.

You are slightly oversampled (at 100% large image does not render stars as pin points, but rather there is softness to them - which means slight oversampling. There is also issue of "fat" diffraction spikes which means atmosphere was not playing ball and there is no detail to justify such resolution). Going with lower sampling rate will improve SNR further - which is great already, this is probably the best rendition of this galaxy that I've seen (not talking about M31, but this little fella):

This is first time I've clearly seen bar in this galaxy and how it's twisted.

Going with lower pixel scale would give you additional SNR and smoother background while all things would still be visible in the image.

Second thing is obviously edge correction of your setup. It is fast newtonian astrograph and sure it's going to suffer some edge softness on larger sensors, but in this case, it sort of hurts mosaic because overlaps can be easily seen, like this:

also, not sure what software you used, stitching is not quite perfect - as this part shows:

And the third one is obviously blown cores of M31 / M32.

I'm aware that I might be considered too harsh with my comments since you produced some splendid images, but I do really think that above things can be easily rectified (add a few filler exposures for the cores, be careful about stitching / blending part and do bin of your data in software) and then you will be closer to perfection in your work.

Thanks for your comments, I don't mind at all as they are all 100% correct. I appreciate you taking the time to comment in such detail! The pic is indeed a bit over sampled and the seeing in Provence can be pretty bad at times. 

The small defects during stitching indeed need some more attention, I'll sit down for that to rectify that. I stitched the panels manually in PS as my pc nearly went up in flames when I fed the mosaic to APP. 

About the blown cores, I understand what you mean. But I often find it looking too flat & manipulated when the cores are dimmed. Personal taste I guess... 

Thanks, 

Pieter 

[tomato]

Superb, I’m trying something similar but alas not with the same outcome.

[Laurin Dave]

Stunning image Pieter..  particularly like the little barred spiral to the lower right.. Took Olly's advice and had a quick look at your website.. bookmarked for further investigation.

Dave

[geoflewis]

OMG Pieter, this without doubt is the most stunning image of M31 that I've seen, the detail is remarkable. Then I had a look at your website and it is full of equally stunning images. The detail seen in your B&W and inverse B&W images could keep me absorbed for years to come.....

Thanks for sharing your work, Geof

[ollypenrice]

Mosaics: one reason I like Registar is that, although it will not always make invisible seams itself, it will allow you, very easily, to repair them. This is my method...

1) Edge crop the linear panels and give them to Registar to build an initial mosaic. Save.

2) Save all the registered/calibrated panels as well for use as 'patches' if necessary.

3) Photshop or similar: give the Registar mosaic an initial stretch, not necessarily going all the way but far enough to show edge defects while recording this stretch as an action.

4) Identify an edge defect and open the 'patch' panel from Registar which will cover it. Apply the Action stretch to it. It should now be nearly identical to the area you want to patch. Slide it into place over the mosaic as a Layer, adjust it slightly in Levels if necessary, and simply use a feathered eraser to remove everything but the part covering the edge defect.

5) Flatten and continue to the final stretch.

I've yet to find an infallible mosaic maker, especially if working with linear panels intended for a hard stretch. This 'patching' method might also be possible in other software provided the component panels going into the mosaic can be individually saved.

Olly

 

[Laurin Dave]

I’ve found Pixinsight’s mosaic registration (which produces the matching panels by registering to the master) and gradient merge mosaic process works well provided the gradients aren’t too bad (which for my location means narrowband)..  for RGB I have to do a light stretch first carefully matching the panels remove gradients and then use Olly’s method.. 

So Pieter it might be worth having a go with the trial version of Pixinsight to see how that works..   certainly faster than APP which nearly blew my computer up too! 

Just been on your website..  the equipment page .. I think I know where that is  

 

Dave

Edited December 9, 2019 by Laurin Dave

[coatesg]

20 hours ago, vlaiv said:

 

This is first time I've clearly seen bar in this galaxy and how it's twisted.

 

UGC 394  Very nice here! It's only 0.22 arcmin along the major axis... There is some great detail here (edge correction on what is a very fast newt with a huge sensor aside!)

I would agree with @vlaiv about the filler exposures for the star cores and galaxy core - the core of M31 is saturated (not a surprise!). Using short exposures and then an HDR combine into a 64bit depth image should sort that, though of course, it doubles the filesize again in doing so... 

Great shot nevertheless!

[pietervdv]

Thanks all for taking the time to look and comment. I took into account some of @vlaiv his great feedback and improved the seams of the mosaic. I also overlaid an old M31 taken some years ago and tried to reduce the burnt core. I improved the contrast in the spiral arms ever so slightly...

Perhaps not perfect, but time to put this one to rest.😀

Here is a new link for the Full size

Thanks & regards,

Pieter

Edited December 9, 2019 by pietervdv

[celestron8g8]

Great capture !