M20 and M8 - Double Trouble

[jase]

Hi All,
Been having some “reasonable” weather of late, though not the best for imaging. High level cloud and fog to contend with – a real pain when its new moon. I’ve been spending a few hours refining the FSQ. Still have a slight field curvature issues. This should be reduced with the new tube extender from Precise Parts. Finally, got around to recording a good FocusMax V-curve so automated computer controlled focusing is now happening. Regularly get down to FWHM around 1.50 which I’m happy about. Still need to set up the temperature compensation. I’ve found that over the course of an hour the FSQ will have shifted focus by at least 30 to 40 microns due to temperature changes – not good when the CFZ is only 55 microns.

I have been working on stepping my images up a notch by carefully considering composition. Anyone can take a photo of a DSO object, but I’m looking to frame the object in aesthetic manner (where possible). Based on the western world, we read from left to right, top to bottom respectively. I look to frame an object in the top left of the field – this serves as a common reference point for the viewer to explore other objects within the frame. In this instance, its M20, but a bright star or cluster would also serve the purpose. I try to break the frame down to thirds across the field. The goal is to immerse/captivate the viewer and provide depth. It is clear when images don’t flow correctly. I have a few on my computer that will never see the public because of this. I will still use them for data on specific objects.

Anyway, I present two new images for your viewing pleasure. ;)

M20 and M8 wide field (http://www.cosmicphotos.com/gallery/nebulas/index.php?path=./&page=0&img=M20,%20M8%20Wide%20Field.jpg&idx=2)

M8 – The Lagoon Nebula (http://www.cosmicphotos.com/gallery/nebulas/index.php?path=./&page=0&img=M8%20-%20Lagoon%20Nebula.jpg&idx=2) (100% scale - crop of the above image)

The image data was collected over two nights (struggled with clouds and fog). Total exposure: 160 minutes (L:70min, R:30, G:30, B:30). Luminance – 10min subs. RGB – 5min subs. I captured the blue channel when the object was close to zenith to mitigate as much blue extinction factor as possible. On reflection, I should have combined the L with Ha data to increase nebulosity and contrast. May collect more data on this another time.

Subs reduced (darks/flats) in MaximDL, registered in Registar and median combined back in MaximDL. Colour balance – R:1.35,G:1.0,B:1.65. Had some issues with PS curves on M8 – experienced white clipping as the nebula has a large dynamic range. I was going to use a multilayered approach allowing better control over the transition of brightness, but decided to use DDP in MaximDL to provide the non-linear stretch. MaximDL DDP by default also sharpens (which I didn’t want) so configured a custom 3x3 kernel user filter to ensure no sharpening was applied during the DDP stretch process. The L and RGB files were then brought into PS for final processing – unsharp mask and inverted layer mask for smoothing the dim areas.

Thanks for looking. All comments welcome.

[Phil]

Great shots Jase well done the colour is spot on nice and sharp to. I tryed the colour balance you have with my photos and worked great in Maxlm.
Phil

[Dr Nick]

Fantastic pics! ;)

[Garyh]

Jase,
You have a fantastic widefield shot there!!! Stars are pinpoint sharp and I think the colors are spot on!!
Well worth all your effort in obtaining and processing!
A big WELL DONE!!
cheers

[2020BC]

Jason, this is a magic shot. I followed the links around your great website. The setup you have there with your observatory in Gippsland looks ideal. Congrats all round.:thumbsup:

[iceman]

Beautiful images, Jase. A lovely pairing very well captured.

[montewilson]

This is great stuff. I know from experience you can certainly get a lot more with the Ha with our set-ups. I aslo know from experience including an Ha-L in a colour image is not easy. So you have done well.

I don't know if you have tried it yet but do a series of say 8x15 mins in Ha in that area and you will be amazed.

That is one of my planned images soon with a decent amount of colour data.

BTW - Ashley at Precise is a top guy. I visited him a couple of times in Miami. By coincidence he had lived a few blocks, in São Paulo - Brazil, from where I was living at that time although he had left a few years earlier to move to Miami. Good chioce Astronomy equipment is much cheaper in the US.

He is a Swiss born American who worked in Engineering and now does this stuff on the side. He has some fantastic machining equipment. Ask him about the cap he does for the STL. I had one made because they don't come with one. He'll probably remember that he did one for me about 18 months ago.

He made my STL-FSQ adaptor and it works so well. He also made a Pentax 67 to STL adaptor too but through no fault of his, it didn't work, the lenses were not suitable for the rigours of AP work :(

[Ric]

Hi Jase, what stunning images, very nicely processed and presented.

Cheers

[richardo]

Hi Jase,
man, .......cooking!!:D

Takes a bit to get new setups tuned, but this setup of yours looks like it's right there.
'Tak' sharp stars,( as we expect from Taka's) saturation to them looks excellent and very well processed.
You'd never know the conditions you spoke of...
Love the wide field giving us all an example of where these two favourite objects are situated from each other.
The M8 shows heaps of very fine detail as well.

I'd say you'd be rather ecstatic with the new scope:thumbsup: ..... lucky man:P

Look forward to more of your images with this setup.

All the best
Rich

[jase]

Thanks Phil:thumbsup: Just a note on colour balance - this is specific to my imaging configuration. If you are using the Custom Scientific LRGB filters with the STL-11000M then you'll come close to the right balance. Even if you had the exact same combination (filters and camera), you'd still need to do you're own G2V star calibration as the NP101 would have different light transmission characteristics to the FSQ-106ED. By all means, go ahead and use RGB ratios I use, but be careful as it may not provide the right balance. If you're using Astrodon's or Astronomik filter sets, I wouldn't recommend you use this balance.


Thank you Dr Nick.


Thanks Gary. The more I work at reprocessing my images the greater pleasure I get out of this journey. Collecting the data at the telescope is cool, but seeing it all come together during image processing is a real buzz. Working with luminance and chrominance information is amazing. :)


Thanks Bill. I really enjoy your website too - you've got some incredible images. Keep up the good work. Imaging from a dark sky is very beneficial when collecting chrominance information as it minimises contamination from light pollutant sources. My only problem is distance. Take two hours to get to the observatory from where I live! So I need ample planning and good fortune that the skies will be clear.


Thank you Mike. The two objects indeed work well together.


Thanks Monte. :thumbsup: The FSQ suits these rich star fields very well. It has taken me sometime to get computer controlled focusing to work well through the 13nm Ha filter. Though I did have some success over the weekend. I needed to significantly increase the exposure time in FocusMax. I managed to take only 30 minutes of data through this filter of another object. I will work on collecting an hour or two more Ha info before I consider adding the chrominance components.
Yes, Ashley from Precise Parts is a good guy. Really hanging out to get the right adapter for the FSQ as the optical train at the moment is producing some serious camera x/y tilt. Not really noticable in the recent images, but its there and its bugging me a little.
Combining the L or Ha luminance data with chrominance data isn't too difficult. There are many techniques available to ensure you don't wash out the colours. Sometimes I blend, other times I layer or do both at the same time. Different nebula types responded in different ways so there is no right or wrong. The trick is to make small adjustments in curves to the luminance data. Don't immediately stretch the data to its final level. Slow and methodical until you reach the right balance.



Thanks Ric for the positive feedback. :thumbsup:



Thanks Rich for the encouraging comments. If there is a refractor made for imaging - its the Takahashi FSQ. Nothing compares to its super flat field. No field flatteners required as its integrated into the modified Petzval design. Something the TV NP101 and NP127 can't compete with. This simplifies the imaging train and accommodates large format cameras with an image circle of 88mm in diameter - massive. Yeah, you could say I'm stoked :D. I only wish Takahashi made a larger aperture version...
I'm still working on the set up. Not totally tuned as yet, but as you can see its more than capable of pumping out nice images. Still working on the processing skills - wide fields are a pain with gradients.


<==========>
For the record, having another look at The Lagoon nebula image (with fresh eyes), I'm not entirely happy with my processing efforts (looks a little "overcooked" in processing. The open cluster stars have become too sharp and harsh when I was bring out more detail in the nebulosity. I will reprocess this image using a different technique when I get a chance. I'm not searching for perfection (as this doesn't exist), but I feel I could have done better on this object.:rolleyes:

Thank you again for your comments.:)

[JohnH]

Jason,

That is a stunning image - you should be very pleased - athe result shows mastery of all the kit and processing techinques.

[jase]

Firstly, Thanks for your comments John H.

Spent a few hours Monday night reworking the Lagoon Nebula image.

Original (http://www.cosmicphotos.com/gallery/nebulas/M8%20-%20Lagoon%20Nebula.jpg)

Reprocessed (http://www.cosmicphotos.com/reprocessed/M8-RGBFinal.jpg)

So how did I reprocess the latest image?

Re-combined cropped and reduced images using MaximDL, then loaded in SBIG’s CCDSharp. CCDSharp - LR deconvolution - 2 iterations (recommend downloading this free tool from SBIG) – I prefer its dconvolution process over MaximDL’s. Very simply to use and has the ability to sample more stars for improve accuracy. Just make sure that the stars you select are not saturated/white clipped.

MaximDL DDP on luminance layer with custom background and mid-levels. I think this is where things went a little wrong as I don’t feel I sampled the mid-levels well enough. Still learning how to use DDP effectively. It’s a powerful non-linear stretch tool if you know how it operates.

In Photoshop:
Chrominance non-linear stretch using PS shadow/highlights tool (not curves or hue/saturation adjustments). Really nice feature of PS is the shadow/highlight tool. This resulted in a richer colour.

Cleaned luminance and chrominance layer with smudge and healing tool – slight star shaping performed – but kept it real – not a big fan of over manipulating.

Used the colour range tool to provide a granular selection, then inversed and applied a noise reduction filter to remove noise from dim areas. This has significantly produced a cleaner image, but I wasn’t careful with the selection criteria and as such some faint nebulosity was also noise reduced (need to adjust feathering values). All details within the main nebula are maintained as this was not selected

Final curves to luminance and raise saturation to taste on chrominance layer

Flattened image

Duplicated flattened layer and produced a high pass filter mask with a 3.5 pixel radius. Using the mask, I ran over the highlights and shadows of the image I wanted to bring out more detail in. This is an alternative to sharpening which inherently introduces noise (easily seen in the original). I then changed the layer mask to 60% opacity. When at 100%, incredible details were present, but the image took on a plastic appearance.
The final result isn’t as sharp as the original considering I really over did the sharpening in the original (IMO). The open cluster stars have remained reasonably tight and well resolved, but still appear sharp and saturated.

The above processing routine is one of many. Colour balance is still right, but saturation possibly a little too high. All depends on what you like. I will contemplate the two images (original and new) for a while. I will re-evaluate them in a few days with fresh eyes again.

<===>
Okay - now for the fun part...

For those that are interested in trying to process this image yourself, I have publicly provided the LRGB FITS files (in IEEE float format) for you to download and try. Those that have never worked with FITS files, here’s your chance (not significantly different to 16-bit TIFF files), but you'll need to have the right software (or photoshop plugins). I didn't want to convert them into TIFF files as some of you may have processing routines that require larger values than 65,500 scales. Depending on your routine, 16-bit integer format by result in clipping. Each file is a crop of the large wide field frame and is 7.5Mb size (so 7.5x4 is 30Mb to download the lot). The files have already been reduced (dark/flat) and registered/combined, though you may need to re-register the L to the RGB in your favourite tool. Note: These will be available for a short time only - when my hosting providers sends a nasty email to me regarding bandwidth consumption I'll need to remove them!

Luminance (http://www.cosmicphotos.com/reprocessed/M8-Lum70sub10.fit), Red (http://www.cosmicphotos.com/reprocessed/M8-Red30sub5.fit), Green (http://www.cosmicphotos.com/reprocessed/M8-Green30sub5.fit) and Blue (http://www.cosmicphotos.com/reprocessed/M8-Blue30sub5.fit)
Luminance is 70min using 10min subs
RGB is 30min each using 5 min subs

(Note; as mentioned these are reasonably large files)

There are two things that I ask;

Firstly - you're free to use your own processed version on your website or other publication for non-profit purposes – however I ask you to please provide credit “Image acquisition by Jason Jennings – www.cosmicphotos.com”.

Secondly – post a pic of it here and explain in detail the processing steps you used for the benefit of all. I am only putting my data that I’ve collected out there to view the processing routines of others and glean from how you processed the image. Don’t be shy, give it a go if you feel you can push the processing limits.

Above all have fun and enjoy the learning curve.:)

[RB]

Wonderful images Jase as usual.
I love looking at your work and reading the details of your images.
I've just downloaded the FITs plugin for PS (never processed FITs files before).
Looking forward to having a play, thank you for the opportunity to have a go.

:thumbsup: :thumbsup:

[jase]

Thanks for your comments Andrew. :thumbsup: If I can provide an insight into how I process particular data, then I hope others will experiment with similar techniques. I note that quite a few people have downloaded the M8 FITS I have publicly provided (many are not in AU - I determined this from the source IP address ranges hitting my website). I look forward to seeing results and processing techniques of other people. :) Unless of course, its all too hard. :lol:;)

[sjastro]

Very nice image Jase.

And thanks for your comments on the Allira article.

Clear skies

Steven
http://users.westconnect.com.au/~sjastro/small/