Please process this for me

[matt-c]

I've been doing this for a few years and i still can't get the hang of processing 

I managed to capture almost 6 hours on m45 but after processing it, it just looks unpleasing and has left a really sour taste in my mouth.

Would anyone take a crack at it to either prove im terrible at processing or my gear is rubbish.

The first image is 3 hours of data in photoshop

This next one is a printscreen off startools (6 hours) as i'm undecided if i want to buy it

 

All taken with heq5, canon 550d and a 200mm takumar lens. Subs 5-10mins.

i'm so close to getting rid of my gear as i just cant produce any decent images in the really limited clear sky i have and its not worth the stress at the minute.

I'll attach the tiff files incase anyone wants a go at it.

https://www.dropbox.com/sh/dyelvedqfzc7lqt/AACNO4cxDk9d039Bh9bRC1ELa?dl=0

Cheers!!!

 

[almcl]

Any chance you could post the unprocessed data?  Both the images in your dropbox have been stretched and so aren't really suited to further work.  The original  FITS file would be great.

[Knight of Clear Skies]

I can relate, I'm also struggling with my equipment and processing at the moment. With 5-10 minute subs I'm wondering if dynamic range could be a problem. Did you stop the lens down at all and what ISO did you use please?

[matt-c]

1 hour ago, almcl said:

Any chance you could post the unprocessed data?  Both the images in your dropbox have been stretched and so aren't really suited to further work.  The original  FITS file would be great.

Hey, thanks for the reply!

Those files are straight from deep sky stacker although i did set black point to zero, maybe that has done something?

[matt-c]

1 minute ago, Knight of Clear Skies said:

I can relate, I'm also struggling with my equipment and processing at the moment. With 5-10 minute subs I'm wondering if dynamic range could be a problem. Did you stop the lens down at all and what ISO did you use please?

i used iso 800 and the lens was stopped down to 5.6 after about half the subs had been taken 

I'm thinking maybe for something this bright i probably didnt need to go too the 10 min mark and it was a bit overkill?

[wimvb]

The longer exposures help bring out the faint nebula surrounding the stars. This is a very high dynamic range image, with bright stars. You did manage to reveal some faint nebulosity, but there is still quite some noise. The 3 and 6 hours you mentioned, what does that refer to? Total integration time?

I have downloaded your data and am doing some processing at the moment.

A few tips so far:

crop all stacking artefacts or set the DSS stacking parameters > Result to intersection mode

Don't do any processing in DSS, not even slider adjustments, if you have post processing software that can handle 16 bit tif images.

Make sure that you have a clean background before post processing (= no stacking residues, no gradients, neutralised background)

The first three steps I use on ALL images are:

cropping

gradient removal (DBE in PixInsight)

background neutralization & colour calibration (in PI)

I will post my results in this thread.

 

Cheers,

[almcl]

Matt

When I apply the Autodev function in StarTools to your image the result I get (below) shows the data has already been 'got at'.  As Wim says, switch off ALL processing in DSS.  StarTools likes completely unprocessed data - unfortunately not possible in DSS which insists on resetting the white balance and there's no way (that I know of) to stop it but everything else should be 'off'.

[Knight of Clear Skies]

1 hour ago, matt-c said:

i used iso 800 and the lens was stopped down to 5.6 after about half the subs had been taken 

I'm thinking maybe for something this bright i probably didnt need to go too the 10 min mark and it was a bit overkill?

Had a quick go at the 6 hours of data and got something very similar looking to your Photoshop effort. I think something has gone wrong here I'm afraid, I'd recommend turning off background calibration in DSS (under stacking settings/lights) and restacking. It might also be worth putting up a couple of your subs (maybe 5-minute and 10-minute examples) to see if there are any problems there.

[wimvb]

Here's my go at your 6 hr data, PixInisghted.

There is a red halo (chromatic abberation??) around some of the brighter stars, which was difficult to remove. Also, there is a hint of faint dust in the background, but it's too close to the noise level for me to get it out without stretching the noise. Your Startools image shows some of this already.

Merry Christmas,

[matt-c]

20 minutes ago, wimvb said:

Here's my go at your 6 hr data, PixInisghted.

There is a red halo (chromatic abberation??) around some of the brighter stars, which was difficult to remove. Also, there is a hint of faint dust in the background, but it's too close to the noise level for me to get it out without stretching the noise. Your Startools image shows some of this already.

Merry Christmas,

Massive improvement thanks a lot!

You're not wrong the image is plagued with chromatic aberration and what a great job you've done.

Merry Christmas to you too

[wimvb]

I came up with this basic recipe (in PixInsight)

split the image in its RGB components

Made a starmask (contour mask) for the red channel, trying to select only the brighter stars

Used this mask to reduce the stars in the red channel only, and recombined the RGB components.

Then stretched the image. First stretch: masked stretch (keeps the stars down and accentuates nebulosity)

After that selective colour saturation and desaturation (background). One of the desaturations was with the previous star mask. This reduced the remaining CA.

Now I'm trying to enhance the faint background dust without blowing up the noise. I'm not sure if this will work, but if I get decent results, I'll post them here.

Cheers,

[wimvb]

Here's an attempt to enhance the faint dust in the background. I turned down the colour saturation a few notches, because of the combined CA and colour noise. So also a less saturated reflection nebula.

As you can see, there is some variation in the background, but it's too close to the noise floor for me to pull out.

In order to improve this, once you know where the dust is and what is supposed to be background, you would have to redo the background neutralization process. Which is the very first processing step.

Cheers,

[wxsatuser]

IC 1995, IC354 and IC 341 are just showing.

The 'Claws' are just showing, they are hard to get right even with excellent data.

[matt-c]

i decided to re stack the image with no rgb background calib and no set black point.

https://dl.dropboxusercontent.com/u/50841363/re stacked 6 hours.tif

i havent had a chance to play with it yet but feel free 

 

[HenkSB]

Hello Matt

I processed your last TIF in DSS and WLPG and pretty much got the same result.  With 6 hours of data it ought to be much better if the seeing is reasonable, IMHO.  For instance, I tried M45 once with a cheap 300 mm Sigma lens with 17 x 56 second images at 6400 ISO and got a better result. 

To rule out the obvious, I presume you are stacking raw images, right?  I can imagine JPEGs washing out all the details.

Looking at the washed out results my first thought would be dew especially since the bright parts of the image are affected more than the rest.  Did you check for it?  Poor seeing can wash out the results too but that would be equal over the entire brightness range.  Not much you  can do about that.  Light pollution will also affect the image but it should not wash out the detail if enough data is available (such as 6 hours).

When turning up the contrast I noticed a cross hatch pattern throughout your image.  I don't now your location but if you live near London I can tell that the main lines are perpendicular to the RA direction according to Stellarium.  There is a slight elongation of the stars in that direction as well.  So the tracking is not perfect, did you autoguide?  If not it could be due to polar misalignment or periodic error.  For 5 to 10 minutes it is small though.  I can't think of any reason why that cross hatch pattern would occur since tracking issues should affect the image equally everywhere.  If it is a diffraction effect it has nothing to do with tracking errors.  I have no clue what could cause that.

The red halos around the stars are indeed probably due to CA from your zoom lens.

I don't think your problem is not so much image processing but the data content.  I wonder if artifacts would show up more clearly if you too shorter exposures at higher ISO.  Doing that should be OK if the camera can handle that ISO so long as the exposure is short enough to not wash out the stars.  It also depends on the DNR curve of your sensor.  May be worth a try, I don't think you need 6 hours to tell the difference.

Sorry I can't help much, hope some of these thoughts may help.

[wimvb]

I confirmed the crosshatch pattern in the background, mentioned by @HenkSB. It reminds me of a Moiré pattern, but I don't know what might cause this.

It will be (almost) impossible to remove this pattern during processing. The only way to subdue it is to darken the background to the point of clipping. This data will definitely not give results anywhere near the previous image.

There's also still a lot of noise in your image. For 36 subs (10 minutes per sub = 6 subs / hour, for 6 hours = 36 subs), I would have expected less noise, unless your camera suffers from a massive amount of hot pixels.

A remedy for this is to dither: move the mount 12 - 15 pixels in RA and/or DEC between exposures. Most astro camera control software will give you this opportunity.