LHaRGB of Tarantula

[Paul Haese]

Last night I drove down to the observatory and removed the flattener, recollimated (it needs more work) and imaged the Tarantula Nebula in LHaRGB. It needs a bit more time as I got 70, 70 45, 40 40 minutes worth per channel.

I am happy with the star colours but not all the shapes are to my liking. Guiding was good as I went back to using the guide scope for the interim, in the subs there was no evidentary problem with flexure, just what looks like a collimation issue near the bottom third of the image.

Not sure about this blend as I just stacked the Ha and red separately as they were both different binning and then I summed them both to get my Ha R blend. The colour was initially over powering but I did some curves adjustments to get this colouring.

More data needed but I am much happier with this data set than my efforts on NGC1365.

Click Here for image (http://paulhaese.net/tarantulaqsi.html)

Feel free to comment.

[jase]

Lovely work Paul. In fact, I'd say its the best image you've produced with this instrument. Stars resolve remarkably well and as you say, they've got good colour to them. The Ha blend works. The neb itself isn't a pure Ha red anyway given the high OIII in the area. You should be pretty happy with this effort. Not a huge amount of data, but a rewarding imaging has been the result. Well done.

[Paul Haese]

Thanks Jase coming from you that means an aweful lot. I will try to get more data on it in the next week. No rest for the wicked.:)

[bmitchell82]

blinder of a shot, vibrant reds, and nice stars.! congrats seems your hard work and patience is paying off

[Paul Haese]

Thanks Brendan, as they say only good things come from hard work. I have to say it again, this monochrome DSO imaging is harder than I thought but I must say seeing a nice Ha come out of the camera is pretty special late at night. More to learn and everything to master yet. must not rest on my laurels.:);)

[multiweb]

Yeah very nice and sharp. Top shot. :thumbsup:

[AlexN]

I'd be really happy with that!

Looks great Paul, the colour, the stars seem very well resolved, that 0.7" per pixel resolution is really showing itself there! :)

The only issue is that tilt at the imaging plane.. What does CCDIS say about it? collimation or tilt or a bit of both?

Not withstanding that, it looks great!

[Paul Haese]

Alex, it says collimation, very little tilt or tip; those numbers are in the decimal area. Getting the collimation right is a pain. Yes the oversampling does help and seeing was quite good down at the house last night. In fact there is a run of it for the next couple of day. High pressure with very little jet stream over head. That usually makes for nice seeing, but it does get very dewy though. Primary and secondary got dewed up again last night.

[AlexN]

Heaters Paul.. Heaters... I'd wrap a heater around the secondary baffle, yes it will create a little turbulence in the optical path, but lets face it, people take stunning images with RCOS scopes, they have a heater built into the secondary.. As for the primary, if you added a dew shield as well, you could safely assume that the heat rising out of the dew shield would stop the ingress of dew on to the primary... Failing that, perhaps fans? Not that you particularly want to be pulling it all apart again...

Yeah.. I really fluked the collimation on my one when I had it.. took all of 5 minutes, did it all by eye and the shots came out razor sharp...

In any case, you have to be pretty damn close to capture the detail you have in this shot.

[bmitchell82]

its a bit like that, i remember doing my helix shot (althogh not a great shot) but it had round stars on 10 min subs. :) i was over the new moon about it!

[Paul Haese]

Using a dew shield now, so that does not work so well.

I did try using a dew heater wrapped around the secondary but it did terrible things to my stars. If you remember my image of 4945?

Like you said the collimaition is very close, just a smidge out really but enough to be a pain. It will get worked out eventually. I have jagged the collimation a couple of times now. When I did eagle and 5128 it was sweet but with the QSI on the back it tends to make it a little harder. No live view to work with, which is not great.

I am pondering upon what I will do with the dew problems. The heater really needs to be inside the assembly and that could be problematic given how fussy the rotational alignment is with these scopes.:confuse3:

More to work on though.

[AlexN]

Yeah your collimation was brilliant with your Eta Carinae, M16 and 5128..

Do you have the option of downloading a subframe on the QSI583? If you download a small sub frame, maybe 50x50px, your download speed should be quick enough to use it for collimation... Otherwise you could perform collimation using the QHY5... you can take .2sec subs.. just use a fairly bright star at say, mag 4 and that should easily provide enough speed to adjust collimation at near enough to real time...

That dew problem sounds like a real bummer... I remember telling you I'd had issues with the secondary fogging up and you'd not seen it... I thought the heater was the best option to solve it, although I do remember your 4945 shot having some weird star action going on... Maybe fans in the rear cell would do the trick, but as you say, the rotational alignment is critical and I wouldn't want to risk doing anything silly...

[TrevorW]

I'd be happy with that

nice one Paul

[gbeal]

Aye, T'is a lovely shot Paul, as we have come to expect from you. The corners suffering slightly? Being picky I know.
What is the dew issue? Is it the secondary? If so, perhaps look at a specific dew heater, like I got from the SDM man, it goes inside the secondary holder and just has a wire protruding.
Gary

[Hagar]

Looks real good Paul. Detail,detail, detail. I think I miss the focal length you have in this scope.

Well done.

[gregbradley]

You're on a roll Paul.

Very nice.

Now for some serious exposure time.

Greg.

[desler]

Yes Paul, Can get lost in the detail! Stunning.



Darren

[Bassnut]

A good effort Paul, but. This is a difficult one to not look messy, more to do with processing than capture. Just IMO, uniqely, Trant sometimes presents better with less overwhelming star action, and less neb extention to preserve the 3d effect the narly bits can provide. Its an artistc interpretation of course, your rendition is natural, but to stand out, sometimes extra thought on processing is handy.

A Sid in this regard, as a second choice might be interesting?

[AlexN]

Agreed... If you upped the strength of the Ha in the data it should dull out the stars a bit and you could numb down the stars and bring in more of the sharp detail that Ha provides... You could be ultra daring, and remove the stars all together :) Dare to be different! :D

[strongmanmike]

Excellent Paul

You are improving at a rate of knotts with that new setup :thumbsup:

If you are after contructive critique the only thing I would say is it does look a bit monochromatic...?..not sure why but could be due to the ha blending :shrug: . The sharpening looks to get noticably stronger towards the centre and imparts a flattened look to the image but this is perhaps just the undercorrected optics..?

Only minor feedback above though, this is great work.

Mike

[telecasterguru]

Paul,

I think you have done a very good job of pulling it all together.

Frank

[Paul Haese]

Thanks all for the comments and suggestions.


Gary, the corners have problems due to it not being flat fielded and collimation needs more attention. A perfectly valid observation for sure.
Dew is not only the secondary but also the primary. I will take a look at the SDM option, it might work quite well and remove this particular problem. Thanks.



Yeah my thoughts too. Any suggestions? I was thinking 10 hours. 4.4 is not quite enough.




Fred, totally agree about this area, even guiding just slight off like this effort can make it look wrong. I would welcome any suggestions you have about lessening the stars and increasing the 3D effect.

Pardon my ignorance but what is Sid? Can you explain how to do this?



Alex how do I up the Ha strength? How do I numb down the stars? I have not explored this yet.



Mike, with the blend I just stacked one lot and then the red and then summed them together. Do you have a suggestion for how I should blend it? Did I do the wrong thing?

I did a mask and sharpened the image that way using the one I learnt on Ken Crawford's site. Did you have a suggestion for making it look better?

Thanks once again for everyones comments. :)

[AlexN]

How did you go about doing the LHaRGB layering Paul?

I generally use clip masks to combine RGB channels from the 3 registered images, then take the red channel into a separate window in PS, layer in the Ha registered frame, set it to screen blending mode, and set the opacity to anywhere between 45~75%, higher sometimes for predominantly red targets.. Blending the Ha with the red will ensure better saturation of reds in the final image... The stronger you make the Ha (higher % opacity) the less stars will show through in that channel. If you were to blend the Ha into your L channel the same way as mentioned before, but say, 30~40% opacity, that would numb out the stars a fair bit..

You could remove them all together by iterative use of the dust and scratches tool in photoshop... Set radius to 12 and set Threshold to 50-120, by looking at the enlarged image, set the threshold so that the center of the brighter stars disappears but image stays sharp.
Next, lower radius two points to 10 and set threshold to 30-80. Follow the image quality and details when setting your threshold, too high, you won't properly remove the star, too low and you'll remove detail from the image highlights. Repeat several iterations with incrementally lower values, the more iterations the better. In the last iteration, radius should be 1 and the Threshold between 3-10. If there are leftovers from brighter stars, use the clone tool or spot healing brush to clean them up.

Sometimes this can leave a funny looking affect in the darker zones of the image, by creating a duplicate layer of the starless image, and applying a 4px gaussin blur, or a slight median filter to the duplicate, use a hide all mask to only blend in the smoothed background, but not the blurred details in the nebula, you can end up with a very smooth, inky black background with no stars and retain a very sharp, star free image of the nebula...

[Octane]

Paul,

That's quite a beauty. Well done. :)

P.S. A "Sid" is a Sidonio. That is, a reprocess. ;)

Regards,
Humayun

[Phil Hart]

that's impressive depth and detail for tarantula. i'm a long way short of that with my focal lengths.

i'd be interested to see the result if you blended some Ha into the Luminance and whether you can maintain the nice rich colour balance. i find it so attractive to use the contrast of the Ha in the luminance (and reducing star sizes) but so hard to maintain a natural colour balance?

[Paul Haese]

Thanks Alex for the blending technique. The image link below is a Ha Luminance blend at 65%. I really like the way it looks but it clearly needs more data.

Thanks Humayun for clearing that up.

Phil what do you think? I like this myself.

Click Here (http://paulhaese.net/tarantulaHaLRGB.html)

[bloodhound31]

Not keen on the color, but the detail is astounding Paul. Awesome work. I always enjoy looking at your posts.

Baz.

[strongmanmike]

Here ya go Paul, I just added a few hours of data for you :D

http://www.pbase.com/strongmanmike2002/image/119392913/original

It's a 50/50 blend of our data sets

Wadda ya recon? :P

Your latest version is a marked improvement on the first rendition by the way :thumbsup:

Mike

[Paul Haese]

Hey thanks Mike, that is super. 50% more data looks the go. Will have to do a narrow band mix with it though.

Yes I like it too, amazing what a little change in blend can produce.

[strongmanmike]

Cool huh?

Just Sidonio'd it again though so have another look - blink it with your original it's even sharper now :thumbsup:

The first round was a straight 50/50 mix nothing else, now I have reduced the star ray bottom left, increased contrast and saturation a tad and due to all the extra data managed to sharpen it up further, I recon it looks great.

Mike

[Phil Hart]

I think that's a good result.. very subjective but I would choose your 'natural' colour balance over Mike's, although obviously the extra data is good. just blending in HA is hard enough, mixing in tricolour narrowband HA-OIII-SII is certainly getting adventurous! :)