Help With Processing M42

[russellhq]

I posted this image earlier of M42 but I want to try and improve on it:

So I started again and re-stacked the 2 exposures, picking out the best subs, the stacking info is:

Exposure 1

23 subs, ISO800 150s

20 darks, ISO800 150s

40 Lights ISO800

Exposure 2

18 subs, ISO800 30s

20 Darks, ISO800 30s

40 Lights ISO800

Both esposures were stacked using Median Kappa Sigma Clipping.

Next, I changed the levels in photoshop and used the following tutorial to layer the 2 exposures:

http://www.astropix....GIT/LAYMASK.HTM

This got me to the following stage:

From here, I'm kinda stuck trying to increase the contrast and make the colours more vivid.

[Steve Ward]

This tute of Doug's any help ? . . . . .

Steve.

[russellhq]

Thanks steve, the tutorial was great, but what I can't get is the images to really stand out. They look washed out if you know what I mean.

Here's my last attempt after too many hours on Photoshop!

[Spikey]

Just had 5 mins on your image Russell.

It's a quick up of the saturation and then a high pass filter -> blend overlay -> touch of noise reduction -> +3 lightness -> flatten layers.

It's a fairly subtle effect but you can see the difference it makes.

[russellhq]

Thanks spikey, I see the difference and I like it. You've managed to remove the "murkiness" and introduced more contrast in the central region. I'll have to give those filters a try.

Would it help if I posted the original stacked images for trying out? And what's the best format to post them in since the TIFs are quite large!

[Spikey]

It's a subtle thing but you really see it when the images are side by side. Make a layer and see what effect the slider has on the filter. It does add some noise and artefacts around brighter areas but you can see the effect it's creating.

If you want to share the original stacked tiff then dropbox or simnilar is your best bet.

The .png's seem to be a bit better than the jpegs for displaying on the forum and are easier to use if it's just demonstrating a technique.

[russellhq]

OK, thanks. I'll have a look at it tonight and copy the files into dropbox.

[squeaky]

More than one way to skin a cat...

For this I did:-

Create a separate layer, equalise, mode=softlight, fill=50%

Flatten

Adjust >> saturation >> 25%

Flatten

High pass sharpen (200+)

Flatten

To get....

[russellhq]

I've uploaded the stacked TIFs below.

https://www.dropbox....acked30s_HQ.rar

https://www.dropbox....cked150s_HQ.rar

[ollypenrice]

Assuming you have good short and long exposures (I used 11 seconds, 50 seconds and 5 mins) then this is the method to use to combine them. It is by far the best way to do it.

http://www.astropix.com/HTML/J_DIGIT/LAYMASK.HTM

I followed this method and have alswys stuck to it.

http://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-R4v6bFN/0/X3/M42-WIDE-2FLs-X3.jpg

http://ollypenrice.smugmug.com/Other/Nebulae-and-clusters/i-QdsxNMJ/0/X3/M42CCBOV2010-X3.jpg

Olly

[russellhq]

Thanks Olly, that's the method I used to combine my 2 stacks (30s & 2min 30s). I think where I'm struggling is stretching the stacked images.

[jezhughes]

i can't remember what i did as i'm half asleep, but if you really want to know i could make a good job at redoing it...

[russellhq]

Thanks Jez, that looks great.

Looking at the image there are also a couple things I'm not quite sure how to fix. It still looks like there's vignetting at the corners even after applying the flats. And there also appears to be a gradient from left to right with the left edge showing a lot more noise than the rest of the image.

Any ideas what's causing this and maybe how to fix it?

[ollypenrice]

To fix gradients after flat fielding the top choice is certainly DBE in Pixinsight. The cheaper, and very good, alternative is Gradient Xterminator by Russ Croman.

Olly

[harry page]

Hi

Thanks for the Data I used Pixinsight DBE- hdr composition-stretch-hdr wavlets-colour sat- LHE( contrast tool )

Pushed as far as I dare

Harry

[russellhq]

I don't know what you've done but I like it!

I've uploaded some stacks with more data this time. 40 off @ 30 and 36 off @ 150s if you want to try pushing it a bit further.

https://www.dropbox....cked30s_40F.rar

https://www.dropbox....ked150s_36F.rar

[russellhq]

Also, I wanted to ask what you think of the stars and if I need to make any adjustments to my tracking, I thought they looked a little bloated?

[harry page]

Hi

Will have another go , as for your stars a little soft ( focus ) but no obvious sighn of bad tracking

Harry

[squeaky]

OK...

There are obviously lots of different software options, but a fairly comprehensive and cheap set of actions can be found here:-

http://www.eprisephoto.com/astro-actions

At 15 dollars (ten quid?) they're not bad. They include a gradient tool, vignetting, and star size reduction.

There are also a number of good tutorials here:- http://www.eprisephoto.com/basics-new

These include (in the advanced section) how to create a star layer so that you can process stars and nebula separately. Stars end up bloated as you lift the fainter nebula because you also lift the fainter areas around stars - unless you separate them.

HTH

[russellhq]

Thanks, Harry. I was concerned about the tracking because my PHD graph wasn't as flat as some others I had seen. But since the image is fine I can assume my graphs are fine!

Thanks for the link squeaky, I'll have a look and as you say it's only a tenner! I've also heard of Noel's Actions, I might look those up too.

[ollypenrice]

When you look at a PHD graph you have to allow for the focal length of the guidescope and the size of the guide camera pixels. In Bin 2 with an ST80 you have a 400mm FL and effectively very big pixels so you can hope for a near flat guide trace. If you use an OAG through a scope of 2.5 metres FL and an unbinned guide camera (small pixels) you can expect the trace to look Himalayan.

Olly

[russellhq]

Thanks olly, I've still to get my head around this part of guiding!

I think the focal length of my finder is 185mm (well that's the distance from the sensor to the lens), the lens is 50mm and the sensor in the finder has 6um pixels.

The focal length of the scope is 1200mm, the aperture is 150mm and the camera pixels are 5.2um.

http://www.wilmslowastro.com/software/formulae.htm#ARCSEC/PIXEL

Using the calculator in the link above, the guider sensor sees 6.69 arcsec/pixel and the imaging camera sees 0.89 arcsec/pixel. But now I'm not sure what to do with this info??

[russellhq]

Bump...

[ollypenrice]

Thanks olly, I've still to get my head around this part of guiding!

I think the focal length of my finder is 185mm (well that's the distance from the sensor to the lens), the lens is 50mm and the sensor in the finder has 6um pixels.

The focal length of the scope is 1200mm, the aperture is 150mm and the camera pixels are 5.2um.

http://www.wilmslowa...tm#ARCSEC/PIXEL

Using the calculator in the link above, the guider sensor sees 6.69 arcsec/pixel and the imaging camera sees 0.89 arcsec/pixel. But now I'm not sure what to do with this info??

This kind of thing is not my forte but here goes;

In my opinion the mismatch in sampling rates between the finder guider and the imaging scope is worryingly large. You want your imaging scope to be tracking sub pixel so that errors won't show. Using your figures for sampling rate we see that you have about 7.5 x more arcseconds per pixel in the guide camera so you have to divide its pixel size by 7.5 in order to get equivalence. However, the situation isn't really that bad because the seeing won't allow you to resolve at anything like 0.89 so you have more tolerance. Your seeing might be more like 2 arcseconds, I don't know. So I'd guess that to image sub pixel at 2 arcsecs your average guide error would need to be at around a quarter of a pixel. Someone else needs to run over this, let me stress! My mount can certainly do this, as can an NEQ6 or HEQ5. Below is a good night with the Tak mount in AA5;

Harry reckoned your stars looked soft but not trailed. That might be because the guiding was generally not quite as tight as the imaging scope needed to perform to it's limit but the errors were random and gave round stars. That can certainly happen. We see it a lot with the long focal length setup which is actually trying to out resolve the seeing as well. This might argue the case for binning 2X2. You would gain signal, perhaps, and not lose resolution because the seeing sets the real limit. Unfortunately for us the camera won't bin successfully.

Back on the Layer Masking video, it wouldn't play to the end for me but when Doug first layered up the two images he hadn't balanced their colours to any significant degree. I would do this before combining them, just by eye. One was far more magenta and one far more red. Get them close before you start. (Of course he may have done it later.) Once you have the balanced image showing its two layers combined, do some colour balancing to get them to look seamless. Also, very important, check the saturation of the short exposures. When part of an image is very bright and highly exposed it tends to look thin on colour. Lift the saturation big time to make it seamless with the long exposure image. You see lots of layer masked M42s with the Trapezium well blended in terms of brightness but very washed out in terms of colour.

Olly

[russellhq]

This kind of thing is not my forte but here goes;

In my opinion the mismatch in sampling rates between the finder guider and the imaging scope is worryingly large. You want your imaging scope to be tracking sub pixel so that errors won't show. Using your figures for sampling rate we see that you have about 7.5 x more arcseconds per pixel in the guide camera so you have to divide its pixel size by 7.5 in order to get equivalence. However, the situation isn't really that bad because the seeing won't allow you to resolve at anything like 0.89 so you have more tolerance. Your seeing might be more like 2 arcseconds, I don't know. So I'd guess that to image sub pixel at 2 arcsecs your average guide error would need to be at around a quarter of a pixel.

This is kinda what I was thinking, a small pixel movement on the guide scope would lead to a large pixel movement on the imaging scope. But as you say, seeing is on my side (probably the only time!) so I might get away with keeping the guiding under 0.25 of a pixel.

If I was struggling to maintain guiding error to within 0.25 of a pixel I guess I could add a 2x barlow which would double the guiding range to 0.50 pixels but I'd lose half my FOV. Would that make sense?