StarTools

Hello there oh wise ones.

I am having real trouble getting star tools to work, and I am almost certain I am either... doing something wrong, or my data is terrible.

Since I can't afford to buy a licence at this time, I am unable to save files, so screenshots will have to do, sorry about that.

So...

The stacked image file can be found here...
https://www.dropbox.com/s/emlcq5vyo1y6u ... 3.FTS?dl=0
It was produced using Deep sky stacker, using settings recommended on these forums in another post.
It consists of
20x 20sec lights @ 800iso
20x 20sec darks @ 800iso

I open this file using Star tools, and select the 2nd option from the left.

Once opened, I run autodev to see what I have.
https://www.dropbox.com/s/j0t4zgiln6jw8 ... 1.jpg?dl=0

As you can see, its a mess.

1st thing is to crop the edges away to get rid of the stacking artefacts.
https://www.dropbox.com/s/6j2s8u0h95ivn ... 1.jpg?dl=0

then.. I run wipe to try to get rid of the vignette and awful light pollution colouring
https://www.dropbox.com/s/xhb52ml45ihpk ... 1.jpg?dl=0

And I get the above.

I am clearly doing something wrong I think.

Would anyone be able to help with this? and if so... I think it would be most useful to try to explain to me the steps your taking, and more importantly... WHY you are taking them, and what they are doing to the image.

Hi and welcome to the ST forums!

To make a long story short; it's all good news! You are doing everything right. When launching Wipe (with Tracking on), it should have told you something to the effect of "...it is highly recommended to re-do your global stretch afterwards. This ensures the dynamic range that has been freed up can be put to good use."

The latter is exactly what is going on in your Wipe screenshot; the bias and gradients seem to have been successfully removed and you are now left with just the celestial signal we're interested in. Right now the dynamic range your monitor can display is not allocated very well to the signal we're left with; just a sliver of it is allocated to the signal (the rest used to be allocated to the gradient/bias). Therefore a re-stretch (suing AutoDev or Develop) is appropriate, this time for just the signal in all its glory without the yucky bias/gradient. Setting "temporary autodev" to "yes" in the Wipe module will give you an idea of what to expect in AutoDev.

A full process of your data could look something like this;

--- Auto Develop
To see what we got. We can see heavy light pollution, vignetting, oversampling, guiding/stacking errors, (small) stacking artifacts, coma and noise.
--- Bin
To make guiding error less apparent and reduce noise. You could also run the Lens module prior to
Parameter [Scale] set to [(scale/noise reduction 25.00%)/(1600.00%)/(+4.00 bits)]
--- Crop
To get rid of stacking artefacts You could also run the Lens module prior to to Cropping to ameliorate the coma.
Parameter [X1] set to [4 pixels]
Parameter [Y1] set to [4 pixels]
Parameter [X2] set to [1178 pixels (-4)]
Parameter [Y2] set to [783 pixels (-4)]
--- Wipe
Vignetting preset. The gradient is pretty severe and tapers off fast.
It is clear no flats are used (they are not optional!) and dust donuts/smudges on the CCD can be seen. Some minor banding is also visible (if becoming a problem, this could be corrected by dithering between frames, using bias frames, or using the Band module)
Parameter [Precision] set to [512 x 512 pixels] (needed if gradients taper off fast or undulate fast).
Parameter [Dark Anomaly Filter] set to [6 pixels] to catch the smaller dust specks (also needed if data is noisy) - bigger specks or donuts would need to be masked out.
Parameter [Corner Aggressiveness] set to [100 %]
Parameter [Aggressiveness] set to [82 %]
Parameter [Temporary AutroDev] set to [Yes], so we can see what an AutoDev result would roughly look like.
--- Auto Develop
RoI over M42 "wing".
Parameter [Ignore Fine Detail <] set to [4.8 pixels] to make AutoDev ignore the noise grain (otherwise it will start optimising the dynamic range allocation for this - what it thinks is - "detail").
--- Deconvolution
Bring out some detail in the core.
Parameter [Radius] set to [2.2 pixels]
--- Color
Default color balance looks good; you captured the core wonderfully (green O III emissions, mixed with blue of the trapezium stars) as well as the outer regions (mixture of blue reflection, dominant Ha and some Hb). Pay particular attention to the pink shockwave "serpentine" feature bordering the green core at 8 o'clock as it extends into the fainter nebula - great stuff. There is some cross-channel alignment trouble with the stars, causing color fringes, but in general a good "random" representation of all star colors/temperatures can be seen.
Parameter [Dark Saturation] set to [6.30] to introduce a bit more color in the darker regions.
Parameter [Bright Saturation] set to [3.90] to remove some color from the brighter regions (to counter the color fringing described above in the stars a little).
--- Wavelet De-Noise
Switching Tracking off.
Parameter [Color Detail Loss] set to [15 %]
Parameter [Brightness Detail Loss] set to [12 %]
Parameter [Grain Size] set to [9.2 pixels]
Parameter [Smoothness] set to [81 %]

And that's it!
All you need is more frames (it's never enough ), acquire some flat frames, dither between frames, check your guiding/alignment and you are seriously in business.

Hope this helps!

Many thanks for this!

I think I understand most of this, ad its helped me get my head around what ST is doing when I press buttons

I still have a problem though...

Trying to follow your steps for myself and the end result is nothing like the one you created.

I think the issue is the second Autodev after the wipe... I cant see to get a reasonable result from this... and it's likely that I aren't using the ROI properly as I dint really understand what it's doing.
I have tried selecting the whole bright are of M42, just a sliver, a cross through of "space" and M42... no joy, I just get a red saturated grainy mess

Sanguisaevum wrote: I still have a problem though...

Trying to follow your steps for myself and the end result is nothing like the one you created.

I think the issue is the second Autodev after the wipe... I cant see to get a reasonable result from this... and it's likely that I aren't using the ROI properly as I dint really understand what it's doing.

The Region of Interest is meant for giving AutoDev the local area it should analyse and optimise for. By the default the "local" area is the whole image (e.g. not so "local", but global instead).
Try giving it just M42's core and you will see it will make detail in the core perfectly visible, but other areas may not be very visible. Do the same thing for a selection of background and it will optimise just for that area and you will notice the midtones and highlights are "squashed" with not much detail visible in them. Give AutoDev a good representation of background and mid tones and highlights and AutoDev will do the best possible job.

I have tried selecting the whole bright are of M42, just a sliver, a cross through of "space" and M42... no joy, I just get a red saturated grainy mess

Did you use the [Ignore Fine Detail <] parameter so that AutoDev will ignore the noise grain?
Are you sure you applied Wipe beforehand? The saturated redness should have been remove by now?

Something like this should do the trick (though notice the heavy binning);

Thanks again.

That makes sense. Although I couldn't quite get the same results as you. I think I understand what was doing wrong now.

I actually ended up using the manual develop instead of a 2nd autodev.

And ended up with this...

https://www.dropbox.com/s/l6lpbr2vld66k ... n.bmp?dl=0

Not perfect, but happy from my first real go.

Thanks again!

Looking good!

A "manual" Develop is of course also a good way to stretch, though I personally almost always tend to put my faith in AutoDev (especially with data that is clean) as it is more objective.

FWIW, what you just produced is infinitely better than my (and I suspect many others') first go!