Orion Nebula (M42)

[swannies1983]

The full moon has given me some needed time to process some of my data. Next up in the processing timeline was M42. I collected this data about a month ago but the processing scared me given the large dynamic range of the object. Here's my first attempt at processing.

Details
Scope: 8" Newt
Mount: EQ6
Camera: Unmodded Canon 400D
Exposures: 12x300secs @ ISO800, 10x120secs @ ISO800, 10x30secs @ ISO800 and 10x10secs @ ISO800 plus darks and flats.
Stacked in DSS and processed in Photoshop CS3

Picture here (http://www.flickr.com/photos/47268787@N05/6825254190/in/photostream)

I'm quite happy given that this is my first go. I had fun layer masking the core. Still need more data to increase SNR of fainter parts.

More than happy for anyone to provide any further processing advice or to have a go on the data yourself.

[Rigel003]

That sounds like a mammoth processing job with four different data sets. It's come out well but I think if anything you've overcompensated a little with the core masking. It needs to be the brightest part of the nebula by a long stretch. Maybe you've reduced the overall dynamic range a little too much.

[cventer]

Very nice indeed. You did a great job of getting the dynamic range of the core right.

My only comment is that its seems obvious where you have blended the core. There is almost a visible line. If you feather the selection more before belending the core you wont get that effect. You can then use curves on the feathered layer of the core to brighten or darken till it matches the surrounding nebulosity

[swannies1983]

Thanks for the comments and suggestions. I was struggling with the blending side of things. I followed these tutorials

http://atalas.businesscatalyst.com/Masks_for_Photoshop.html and

http://www.astropix.com/PFA/SAMPLE4A/SAMPLE4A.HTM

The first one mentions to guassian blur with a radius of about 20 pixels. The other link says:

"low resolution images will require a smaller pixel radius and high resolution images will need a larger radius. For a 1 megabyte image, try a blur of 10 pixels. For a 10 megabyte image, try a blur of 50 pixels."

I really did have to increase the radius up quite a bit (~70-80) otherwise the blending margin was quite obvious. Was I doing something wrong? I was also having issues deciding the area the go over with the brush. After the masking, I did use the lasso tool, feathered to 30pixels, on certain areas of the core to try and blend the brightness of the masking regions.

One big learning curve hey :lol:

[strongmanmike]

Overall a great first result Dan :thumbsup:

I agree with Graeme though the core and particularly that thick bit of the "wing" that cradles the trap, needs to stay bright.

The difficulty comes when you try to keep some of the cool detail inside the core and around the trap while at the same time still maintaining some good brightness in order to keep it looking real...and this is quite hard to get right.

Sounds like your approach is ok though and you probably just need to try some more variations :)

Mike

[swannies1983]

Ta Mike. I'm sure a repro is on the cards ;)

[swannies1983]

Question: do people tend to process each stack of exposures before layer masking OR do just a little bit of processing, then combine, then process hard?

[Ross G]

Great photo dan.

I love the colour and detail.

Ross.

[midnight]

Nice attempt Dan. You certainly have the concept of blending but the guys are right with the core and needing to be brighter.

I'm certainly no expert but on my goes, I processed each image and then blended last via the layers.

It's a good target to learn post processing skills and when I get back home to Australia, I want to give this another go with some old data.

Good luck!

Darrin...