StarTools

Ivo,

Thanks!

A link to the stack is below. This is ~2 hours, taken with an unmodified D5300 and an Optolong CLS-CCD filter, w/Flats and dark-flats, no bias, no darks.
DSS logs attached

https://www.amazon.ca/clouddrive/share/ ... xgPyr9fI3D

(I just split the topic from the feature request )

Thank you for uploading! I just had a look at the dataset.The main problem seems to be that no dithering has taken place between frames. When going for "advanced", faint objects like these, dithering - like flats - is not really optional. Otherwise you're just fighting the pattern noise (the faint "rain"/streak-like pattern in the background), trying to bring out faint nebulosity that is mired in this these patterns and streaks. This is causing the "mottle" you are referring to. Right now, it's hard to discern for algorithms (and humans) what is aberrant pattern noise and what is real faint detail, which is a shame because otherwise the data is deep enough to bring out lots of detail. Dithering usually also means you can dispense with darks/bias in a lot of cases, so it's really worth doing.

I'm not 100% sure what sort of rendition you are looking for, or how much nebulosity you wish to bring out, but I went with the following rendition;

--- Auto Develop
To see what we got. We can see a blue bias, oversampling, some noise, stacking artifacts
--- Bin
Parameter [Scale] set to [(scale/noise reduction 25.00%)/(1600.00%)/(+4.00 bits)]
Image size is 1498 x 997
--- Rotate
Parameter [Angle] set to [180.00]
--- Crop
Remove stacking artefacts.
Parameter [X1] set to [6 pixels]
Parameter [Y1] set to [5 pixels]
Parameter [X2] set to [1495 pixels (-3)]
Parameter [Y2] set to [994 pixels (-3)]
Image size is 1489 x 989
--- Wipe
It seems like the flats aren't fully correcting light fall-off/vignetting (or possibly over-correcting).
Vignetting preset with tweaked Corner Aggressiveness
Parameter [Dark Anomaly Filter] set to [3 pixels] to catch any dark pixels
Parameter [Corner Aggressiveness] set to [93 %]
--- Auto Develop
Final global stretch. I use a slice of the nebula as a good sample of what we want AutoDev to optimize for.
Parameter [Ignore Fine Detail <] set to [Off]
Parameter [RoI X1] set to [414 pixels]
Parameter [RoI Y1] set to [470 pixels]
Parameter [RoI X2] set to [935 pixels (-554)]
Parameter [RoI Y2] set to [537 pixels (-452)]
--- Color
Color balancing LP-filtered data is always tricky/impossible, as chunks of the spectrum are simply missing. I decided to just concentrate on getting the popular red/blue "HOO" rendering.
Parameter [Blue Bias Reduce] set to [1.25]
Parameter [Red Bias Reduce] set to [1.05]
Parameter [Cap Green] set to [100 %] (only use as last resort on non-LP-filtered data)
--- Life
Simple Isolate preset (which is the default as of 1.5). No mask!
Parameter [Inherit Brightness, Color] set to [Off] (this modifies the coloring too)
Parameter [Saturation] set to [200 %]
Parameter [Strength] set to [50 %]
--- Develop
Added a slight skyglow
Parameter [Skyglow] set to [3 %]
--- Wavelet De-Noise
Default.
Parameter [Grain Size] set to [4.5 pixels]

Hope this helps!

Thanks, Ivo!

Actually, I did dither. Just checked my guide log, and there were four dither events in that session.
So...if that's the source of my mottle, that leaves me wondering whether I needed more dithers, or more aggressive dithers, or both.
Any ideas?

- Bob

Hi,
First of all this image is really beautiful, my compliments to both of you!
Bob, which setup did you use to capture this image? I am using an unmodded Nikon and looking at this image encourages me a lot.

Ivo, saying “Color balancing LP-filtered data is always tricky/impossible” do you mean that it would have been better to avoid the filter? Sorry if it sounds stupid.

Gaetano

devonshire wrote:Thanks, Ivo!

Actually, I did dither. Just checked my guide log, and there were four dither events in that session.
So...if that's the source of my mottle, that leaves me wondering whether I needed more dithers, or more aggressive dithers, or both.
Any ideas?

- Bob

Ah, okay. That's great - that means you have dithering capability!
Yes, lots more dithering between every frame is highly recommended.
A spiral pattern is usually the most popular, as it keeps the deviations between frames close together, so you keep stacking artefacts to a minimum on all sides.

Hi,
First of all this image is really beautiful, my compliments to both of you!
Bob, which setup did you use to capture this image? I am using an unmodded Nikon and looking at this image encourages me a lot.

Ivo, saying “Color balancing LP-filtered data is always tricky/impossible” do you mean that it would have been better to avoid the filter? Sorry if it sounds stupid.

Gaetano

Well, avoiding a light pollution filter will allow you to image the whole visual spectrum. With the filter in place, some parts of the spectrum are filtered out (usually sodium lines). Typically this results in a distinct lack of yellow (you can see this in the Veil image above - you can't really see a good continuum from red->orange->yellow->white->blue in the stars) when you are trying to color balance; you can't balance something you never recorded. Leaving the light pollution in, you do get the entire spectrum and can thus achieve "correct coloring". You can even subtract the light pollution (using Wipe if you use StarTools). What you can't undo, however, is the noise that the light pollution signal introduces as well. This will remain in your image. This is one big reason why people want to use a light pollution filter; if you're not recording the light pollution at all to begin with, then you're also not recording the noise of the light pollution signal. The other reason is that light pollution makes your image overexpose quicker.

Also, no such thing as a stupid question

Ivo,

Ok - I had dialed back to RA only and not terribly aggressive. I'll go back to spiral and dither more frequently next session.
Thank you so much for your help and advice with this!


Gaetano,

What Ivo said

The image was taken with a Nikon D5300, a Hotech field flattener, an Optolong CLS-CCD filter, and a 102/621mm ED triplet refractor.

This was pretty much first light for the CLS-CCD filter. There are lots of different filters that favor nebula emissions and eliminate light pollution. Some highly specialized, some not. No magic in what I picked - it's a general purpose, dual-band filter that cuts out general light pollution, including the sharp blue from LED sources, and passes several kinds of nebula emissions. I had been using a UV/IR CUT filter to help reduce color fringing, and this particular filter's range provides equivalent functionality, so that was a factor, too. This doesn't increase the light you capture, it reduces the light that you don't want and basically improves contrast.

Here's a picture of the spectrum that this filter captures. Again - there are lots of different choices out there, depends on what you want to achieve.

- Bob