Noise reduction / pixel fixing with Heal

[Rowland]

Hi.

Before and after images show the removal of what I think are under sampled areas of background. Perhaps some read noise intermixed. Heal appears to be doing what it is supposed to do, very effectively, though the mask and heal parameters may require tweaking. The routine has been very effective raising the dark pixel values to represent what is probably an accurate representation of the area in general (M45). This followed the Color module with tracking off. I noticed the result was less effective prior to Colour, but Tracking must still be off.

M45 is often rendered with a dark background and blue reflection nebula. I have approximately 5 hours of data which reveals beautiful areas of red emerging at the periphery of the blue. The images here are about 3 hours worth at ISO1600 Canon 450D - coooled -10C, which may have something to do with the dark specks and may partly be a calibration issue.

Is there a way of doing this while retaining tracking. Any suggestions for Heal settings which may do a better job of producing a more uniform background.

before.jpeg (450.81 KiB) Viewed 7200 times

after.jpeg (402.44 KiB) Viewed 7200 times

[attachment=2]

[Rowland]

Due to the image limit I've posted this as an example of what I want to eliminate. Black spots all over the image.

EDIT: After a bit of tweaking and playing with masks and different size kernels I find it difficult to tag only the darkest pixels in a particular range of darkness (visual appearance). The dead pixel color/mono mask is the closest, but which mask settings should I use to tag a range of darker pixels from black (0) upwards.

Am I going about this the wrong way? From what I can see, Heal runs a type of winsorisation, with the mask defining the upper and lower values of tagged pixels?

Any ideas?

integrated_M45-2ST.jpg (373.93 KiB) Viewed 7199 times

[Guy]

Hi Rowland,

I expect you've tried these ideas already - but here is the sequence I would check before trying the Heal module.

- Problem with stacking - for example - are there any settings that affect the black point?
- Problem with Wipe - Dark Anomaly Filter setting too low - or if anomalies are too large in which case can you use the Mask
- AutoDev - Ignore Fine Detail set too low

Questions:
- Does it appear in first global stretch?
- At what point did you do the Heal?
- Can you post the workflow log?

Guy

[Rowland]

Thanks Guy. As follows.

Needs further green channel reduction. Very basic approach so as not to misrepresent the results.

I am further surprised by the effects of debayering. Colour artefacts are more of an issue than I first noticed.

The background didn't get the sampling I wanted of the image set, but it is clearly not a featureless blackness as seen in many images. The IR surveys of this area are quite dynamic. Just a matter of more time for the finer detail.

Rowland

Hi Rowland,

I expect you've tried these ideas already - but here is the sequence I would check before trying the Heal module.

- Problem with stacking - for example - are there any settings that affect the black point? - Yes. Thanks. Changed the preprocessing options to superpixel or four colors debayer - debayer was the issue.
- Problem with Wipe - Dark Anomaly Filter setting too low - or if anomalies are too large in which case can you use the Mask - Not so much of a problem after changing debayer options
- AutoDev - Ignore Fine Detail set too low - Same as above

Questions:
- Does it appear in first global stretch? - Yes, it did
- At what point did you do the Heal? [b - After color and before noise reduction tracking off - but I have since found that after noise reduction is also effective providing noise reduction is not overly smearing the background. Further noise reduction after heal is interesting as most noise is now averaged to neighbouring pixels[/b]
- Can you post the workflow log? - below

integrated_M45-2H.jpg (273.78 KiB) Viewed 7072 times

Code: Select all

StarTools 1.4.324
Sat Dec 17 10:49:29 2016
-----------------------------------------------------------
File loaded [/home/geoland/Astrophotography/M45-2/integrated_M45-2.tiff].
--- 
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [Outside ROI Influence] set to [15 %]
--- Crop
Parameter [X1] set to [39 pixels]
Parameter [Y1] set to [31 pixels]
Parameter [X2] set to [1899 pixels (-125)]
Parameter [Y2] set to [1271 pixels (-17)]
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [Outside ROI Influence] set to [15 %]
--- Wipe
Parameter [Mode] set to [Correct Color & Brightness]
Parameter [UNKNOWN] set to [No]
Parameter [Precision] set to [256 x 256 pixels]
Parameter [Dark Anomaly Filter] set to [1 pixels]
Parameter [Drop Off Point] set to [100 %]
Parameter [Corner Aggressiveness] set to [100 %]
Parameter [Aggressiveness] set to [85 %]
--- Auto Develop
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [Outside ROI Influence] set to [15 %]
--- HDR
Parameter [Small Detail Precision] set to [Max]
Parameter [Channels] set to [Brightness Only]
Parameter [Algorithm] set to [Optimize Soft]
Parameter [Dark/Bright Response] set to [Full]
Parameter [Detail Size Range] set to [21 pixels]
Parameter [Noise Suppression] set to [1.2]
--- Color
Parameter [Cap Green] set to [To Yellow]
Parameter [Bias Slider Mode] set to [Sliders Increase Color Bias]
Parameter [Style] set to [Scientific (Color Constancy)]
Parameter [LRGB Method Emulation] set to [Straight CIELab Luminance Retention]
Parameter [Dark Saturation] set to [2.00]
Parameter [Bright Saturation] set to [Full]
Parameter [Saturation Amount] set to [140 %]
Parameter [Blue Bias Increase] set to [1.22]
Parameter [Green Bias Increase] set to [1.00]
Parameter [Red Bias Increase] set to [1.00]
Parameter [Mask Fuzz] set to [1.0 pixels]
--- Wavelet De-Noise
Parameter [Scale 1] set to [90 %]
Parameter [Scale 2] set to [90 %]
Parameter [Scale 3] set to [90 %]
Parameter [Scale 4] set to [90 %]
Parameter [Scale 5] set to [0 %]
Parameter [Mask Fuzz] set to [1.0 pixels]
Parameter [Scale Correlation] set to [6]
Parameter [Color Detail Loss] set to [12 %]
Parameter [Brightness Detail Loss] set to [12 %]
Parameter [Grain Size] set to [4.5 pixels]
Parameter [Read Noise Compensation] set to [Off]
Parameter [Smoothness] set to [75 %]
--- 
Parameter [Quality] set to [Medium]
Parameter [New Darker Than Old] set to [Yes]
Parameter [Grow Mask] set to [0 pixels]
Parameter [Neighbourhood Samples] set to [0]
Parameter [New Must Be Darker Than] set to [Off]
Parameter [Neighbourhood Area] set to [200 pixels]
File saved [/home/geoland/Astrophotography/M45-2/integrated_M45-2H.tiff].

[Guy]

Glad you found the problem.
I like the image - it does show a lot more than we usually see.

Some thoughts:

• When using Superpixel mode - do you get coloured fringes as Michael Covington describes in his notes on Superpixel mode?
  - If you do have these problems then one other thing you could try (if you haven't already) is to go back to standard bilinear interpolation (with 'Set the black point to 0' unticked) and then Bin the image in StarTools.
• When trying to bring out the areas of red behind the blue have you tried the Life module Isolate preset - with (or without) a mask?

Guy

[Rowland]

Thanks Guy.

Glad you found the problem.
I like the image - it does show a lot more than we usually see.

Some thoughts:

• When using Superpixel mode - do you get coloured fringes as Michael Covington describes in his notes on Superpixel mode? It depends on the gradient between light and dark. Star rims for instance. I don't use DSS as I run Linux. PixInsight is my go to preprocessing program, but I am using my own script developed for DSLR (consumer digital camera) RAW and non-RAW preprocessing
  - If you do have these problems then one other thing you could try (if you haven't already) is to go back to standard bilinear interpolation (with 'Set the black point to 0' unticked) and then Bin the image in StarTools. Bilinear is a good old fall back, but I find four colors produces less artefacts as a rule. Partly it has to do with the use of my own processing script, which does not have a sigma clipping routine, as yet. Most times I can get away without it.
• When trying to bring out the areas of red behind the blue have you tried the Life module Isolate preset - with (or without) a mask? I often use Life, but not on this occasion
  Guy