M 101

[trobbet]

Thanks Ivo!

Now I get it...

Somehow i was getting the idea that this was something new.

I am more than willing (in fact, I'd love to) to send the fore mentioned .fit file, but again I can't find your e-mail address. If you need mine, its trobbet at aol.

Thanks again for all the help,
Tom

[admin]

Thanks Ivo!

Now I get it...

Somehow i was getting the idea that this was something new.

I am more than willing (in fact, I'd love to) to send the fore mentioned .fit file, but again I can't find your e-mail address. If you need mine, its trobbet at aol.

Thanks again for all the help,
Tom

I PM'ed you with my e-mail a couple of hours ago but will send one to yours as well

[admin]

Now I get it.

Awesome

Thanks for sharing the M101 data with me.
This is what I came up with;

M 101 39X170.jpg (96.15 KiB) Viewed 8477 times

Workflow was the following;

--- Auto Develop
Default parameters to see what we got.
We can see big red bias, stacking artifacts, fair bit of noise.
--- Bin
Parameter [Scale] set to [(scale/noise reduction 50.00%)/(400.00%)/(+2.00 bits)]
To undo debayering interpolation.
--- Crop
To frame M101 better and get rid of stacking artifacts.
Parameter [X1] set to [684 pixels]
Parameter [Y1] set to [766 pixels]
Parameter [X2] set to [2183 pixels (-685)]
Parameter [Y2] set to [1752 pixels (-129)]
--- Wipe
To get rid of bias.
Default parameters.
Parameter [Dark Anomaly Filter] set to [5 pixels]
Whoops! Noticed I left in a staking artifact (lower left corner).
Cancel.
--- Crop
Cropping out stacking artifact.
Parameter [X1] set to [158 pixels]
Parameter [Y1] set to [43 pixels]
Parameter [X2] set to [1343 pixels (-156)]
Parameter [Y2] set to [938 pixels (-48)]
--- Wipe
Parameter [Dark Anomaly Filter] set to [5 pixels]
--- Auto Develop
For final stretch. ROI over M101.
Parameter [Ignore Fine Detail <] set to [4.8 pixels]
Parameter [Outside ROI Influence] set to [4 %]
--- Deconvolution
Auto-generate mask.
Parameter [Radius] set to [1.7 pixels]
--- HDR
Taming core of M101 a little. (Tame preset)
Parameter [Detail Size Range] set to [242 pixels]
--- Wavelet Sharpen
Using same mask that Decon created.
Parameter [Amount] set to [173 %]
Parameter [Small Detail Bias] set to [93 %]
--- Life
Isolate preset (set full mask)
--- Color
Parameter [Cap Green] set to [To Yellow]
Parameter [Dark Saturation] set to [3.80]
Parameter [Green Bias Reduce] set to [1.15]
--- Wavelet De-Noise
Parameter [Scale 5] set to [52 %]
Parameter [Grain Size] set to [28.6 pixels]
Parameter [Read Noise Compensation] set to [1.23 %]

And that's it!

So, comparing to your image, my interpretation is a bit more subtle. The ringing around the stars in your image seems to stem from a radius for Decon that is too high (e.g. doesn't correspond to the blur that we're trying to undo). The noise is a bit better under control in my image, but other than that, everything else is a matter of taste!

[trobbet]

Ivo

This is tremendously helpful!
You improved my image a lot, and I'm trying to use this example to learn StarTools better.

Things you did differently than me which obviously helped:

Auto Dev.
You set Ignore Fine Detail at 4.8; I used 3.2.
You set Outside RIO Influence 4%; I left mine on the default 15%.

Decon
You set Parameter radius at 1.7; I used 3.1.

HDR
You used Taming Core preset; I used Reveal.
you used Parameter 242; I used default 64 (for Reveal)

Color
You raised Dark Saturation to 3.80; I left mine on default.

De-Noise
You raised Parameter Scale 5 to 52%; I let mine at default 0%.
You raised Parameter (Grain size) to 28.6 pixels; I was somewhere around 5.5!

Some of these differences I am beginning to get; others I have questions about.

One, I struggle with how to set the Parameter Radius in Decon. How do I determine "the size of the blur" I want to have removed. Do I just try different radius's, and see how it looks, or do you do something like zoom in and estimate the pixel amount; or what?

I also struggle with where to set "Ignore Fine Detail" in Auto Dev., and Grain size in De-noise; similar issues. How does one begin to determine these parameters?

I have some other confusions, but there will be time for those down the road.

Any pointers here would be much appreciated.

CS, Tom

PS, Again, This is absolutely amazing software, and you have been very generous with your help!!!

[trobbet]

Oh!

One other thing I really meant to ask.
In Wipe, did you not mask out M 101? I have been in the habit of always masking out galaxies, so as not to lose fine detail in the galaxy. I tried it both ways just now, and see only a slight difference overall, but I think I like the masked out one better.

Cheers, Tom

[admin]

Oh!

One other thing I really meant to ask.
In Wipe, did you not mask out M 101? I have been in the habit of always masking out galaxies, so as not to lose fine detail in the galaxy. I tried it both ways just now, and see only a slight difference overall, but I think I like the masked out one better.

Cheers, Tom

Good point Tom!
Actually you caught me being lazy...
You will indeed get a better result (retention of its outer halo) if you mask out the galaxy.

[admin]

Auto Dev.
You set Ignore Fine Detail at 4.8; I used 3.2.
You set Outside RIO Influence 4%; I left mine on the default 15%.

I also struggle with where to set "Ignore Fine Detail" in Auto Dev., and Grain size in De-noise; similar issues. How does one begin to determine these parameters?

The Ignore Fine detail setting makes sure that AutoDev won't optimise for fine detail under that threshold (which basically helps it ignore noise grain), otherwise it'll try to bring out that fine detail. Bring it up until the noise that AutoDev brings out isn't reduced any further.
The Outside ROI Influence I lowered because I wanted AutoDev to concentrate more on the ROI and less on the data outside the ROI (which was mostly background with noise in it)I. It's usually good to leave a little bit optimization for the region outside the ROI. If you completely set it to 0, you may start getting clipping.

Grain Size in Denoise is very much related to the Ignore Fine detail setting; they both specify a threshold (detail size) beyond which noise grain becomes negligible. For Denoise this threshold is visual (e.g. what is perceived by humans, which is less forgiving), for AutoDev this threshold is analytical (e.g. what is perceived by AutoDev's algorithm which is a bit more forgiving).

Decon
You set Parameter radius at 1.7; I used 3.1.

One, I struggle with how to set the Parameter Radius in Decon. How do I determine "the size of the blur" I want to have removed. Do I just try different radius's, and see how it looks, or do you do something like zoom in and estimate the pixel amount; or what?

Yes, that's pretty much what you do.
For a number of reasons, Decon in StarTools is solely focused on reversing the blur that atmospheric turbulence (aka 'seeing') caused. Depending on how severe this blur is and what the angular resolution is of your setup, detail may be 'smeared out' over multiple pixels. The point spread function (PSF) that is used to model this blur has the form of a Gaussian profile. The 'radius' parameter specifies the width of the profile. This will be vary from night to night.

In practice in StarTools, you slowly bring up the radius parameter and monitor any stars that are not protected by a mask. As soon as they start showing signs of ringing (a ring that is darker than their background starts forming), you know the radius is too high to accurately model the blur. This is when you back off a little until no such artifacts are introduced anymore.

Color
You raised Dark Saturation to 3.80; I left mine on default.

It introduces a little bit more color in the darker parts of the image (not much difference in this instance though).

De-Noise
You raised Parameter Scale 5 to 52%; I let mine at default 0%.
You raised Parameter (Grain size) to 28.6 pixels; I was somewhere around 5.5!

Activating the larger scale helps getting rid of very large noise blotches/grain. It's effective when you can't otherwise smoothen the background well enough.
Regarding Grain Size, see above.

PS, Again, This is absolutely amazing software, and you have been very generous with your help!!!

Thanks Tom! Happy to hear ST makes a difference.

[trobbet]

Ivo,

Your explanations and the level of detail is very very helpful!

Much Thanks, Tom

[Cheman]

Che