Horsehead and flame - 4 panel mosaic

[ONIKKINEN]

The Horsehead and Flame nebulae shot with an 8'' newtonian, a Rising Cam IMX571 OSC camera and an Antlia Triband RGB Ultra filter:

42min per panel to a total of 2h 48min over 3 nights, first of which was in December 2022 and the 2 others in the past few months with one of them resulting in just 6 minutes per panel due to 12m/s winds throwing the scope around - this is a tricky target to image from 60 degrees of latitude so data trickles in at a pace of an hour per year it seems. Be lucky to reach my 10h target by the end of the decade! Binned x4 and further reduced by resampling to 80% close to the end of the process.

Mosaicing by PixInsight with BXT ran in correct only mode before (works flawlessly, absolutely pixel perfect mosaic!), the rest of the process mostly in Photoshop.

Comments, critique welcome.

-Oskari

[WolfieGlos]

1 hour ago, ONIKKINEN said:

42min per panel to a total of 2h 48min over 3 nights

Wow, where has the noise gone?!

Nice image and processing.

[symmetal]

Excellent, and very impressive considering the altitude. 😊 My only query is the edge of the area of sharpening on the flame nebula looks a little hard.

Alan

[ONIKKINEN]

2 hours ago, WolfieGlos said:

Wow, where has the noise gone?!

Nice image and processing.

Thanks, Binning and resampling to 3.6''/px helps a lot and NoiseXterminator took care of the rest. I apply NXT with masks and layers to where i think its needed most at several different points in the process including before, between, and after stretching.

1 hour ago, symmetal said:

Excellent, and very impressive considering the altitude. 😊 My only query is the edge of the area of sharpening on the flame nebula looks a little hard.

Alan

Thank you, i think i see what you mean. I often push sharpening too hard and only spot it a day or two later when my eyes are "used to" looking at the new image.

I softened some parts a bit to take the edge off.  Doesn't look worse at least to my eyes so probably an improvement.

[symmetal]

Yes, your last image looks more natural, to me anyway. 🙂

Alan

[Roy Foreman]

Nice rendition.  This area is tough enough at my 52° location, so you are doing extremely well at 60°

Great work.

[ollypenrice]

It's a great result. The sharpening, though hard, didn't offend me and my only thought concerns the colour of the flame, which is rather reddder than in most renditions. My OSC cameras have always given it a kind of yellow-orange colour but I guess this is the filter at work?

Olly

[ONIKKINEN]

6 hours ago, ollypenrice said:

It's a great result. The sharpening, though hard, didn't offend me and my only thought concerns the colour of the flame, which is rather reddder than in most renditions. My OSC cameras have always given it a kind of yellow-orange colour but I guess this is the filter at work?

Olly

Thanks, its the filter + SPCC in PixInsight that leads to the red flame, at least i think it is. Didn't do anything specific to the red channel that would give a red look to things that aren't supposed to look red (like a red curve etc).

Interestingly, the flame is more yellow before colour calibration takes place:

So i think one could get the "usual" looking palette with a more manual approach to colour calibration instead of SPCC.

[ollypenrice]

1 hour ago, ONIKKINEN said:

 

So i think one could get the "usual" looking palette with a more manual approach to colour calibration instead of SPCC.

I firmly believe this, but it's easier for me because my colours are more or less right from the stack and rarely need more than ABE or DBE and a bit of SCNR green. I wonder what the uncalibrated image would look like with a big dose of SCNR green because, apart from the blues looking green, the colour looks good.

Olly

[ONIKKINEN]

52 minutes ago, ollypenrice said:

I firmly believe this, but it's easier for me because my colours are more or less right from the stack and rarely need more than ABE or DBE and a bit of SCNR green. I wonder what the uncalibrated image would look like with a big dose of SCNR green because, apart from the blues looking green, the colour looks good.

Olly

It would look like the usual flame with a manual colour calibration + SCNR, i know because i did that a year ago when i didnt yet have PI. Will post a version sans SPCC later today or tomorrow.

But i think the lack of yellow is "real" (whatever real means with narrowband imaging) because the filter does not really pass any yellow. The bandpasses are roughly 35nm in the deep blue, OIII, and Ha/SII so no yellow or orange is passed.

[ONIKKINEN]

Another version, with a manual colour calibration this time:

Its wasn't quite as straight forward as simply SCNR the green away and call it a day (skill issue, most likely) to get the yellow flame but here it is anyway.

 

[geeklee]

On 21/01/2024 at 18:38, ONIKKINEN said:

42min per panel to a total of 2h 48min over 3 nights, first of which was in December 2022 and the 2 others in the past few months with one of them resulting in just 6 minutes per panel due to 12m/s winds throwing the scope around - this is a tricky target to image from 60 degrees of latitude so data trickles in at a pace of an hour per year it seems.

That looks great Oskari (all the revisions shown too).  I was bemoaning the altitude of this area the other day but you've definitely got an even greater challenge a few degrees higher.  The binning/resampling has worked a treat to generate a dusty, Ha filled image that's still full of detail.