Tag Archives: 7NM

Creating a Hubble Palette Image from OSC Dual Band Data

Many people like myself have transitioned from a MONO camera to a One Shot Colour (OSC) for whatever reason, for me it was all about not being able to get the required amount of time due to weather conditions here in the UK. When I first considered moving to an OSC camera, it dawned on me that I would not be able to produce the vibrant Hubble Palette images that I could produce by imaging with specific filters on my MONO camera, specifically Hydrogen Alpha (Ha), Oxygen 3 (OIII) and Sulphur Dioxide 2 (SII) which would then be mapped to the appropriate colour channels when creating the final image stack.

Now along came Dual and Tri band narrowband filters for OSC cameras which peaked my attention, the Dual Band filters allow Ha and OIII data to pass, the Tri Band filters allow Ha, Hb (Hydrogen Beta) and OIII to pass but at a high Nm value. I reached out to my friends at Optolong who had two filters, the L-eNhance and the L-eXtreme, the L-eNhance is a Tri Band filter, but after speaking with Optolong it would not work well for me at F2.8, so I went with the L-eXtreme Dual Band filter which has both the Ha and OIII at 7nm.

After receinving my ASI6200MC Pro, I decided to start acquiring data on a 1/2 to 2/3 moonlit nights on the North America Nebula, and so far when writing this post I had acquired a total of 60 frames of 300 seconds each at a gain value of 100, I processed the image my normal way in PixInsight and below is the result of the image:

North America Nebula, 60x300S @Gain100, Darks, Flats and BIAS frames applied with the ASI6200MC Pro using the Optolong L-eXtreme Dual Band 2″ Filter

I thought that my data looks good enough to work with and experiment with trying to build an SHO (Hubble Palette) image with, and I have spoken with Shawn Nielsen on this exact subject a few times so he gave me some hints and tips especially with the blending of the channels. So off I went to try and produce an SHO image.

Before we start, there are some requirements:

  • This tutorial uses PixInsight, I am not sure how you would acomplish this with Photoshop since I have not used PhotoShop for Astro Image Processing for a number of years
  • Data captured with a One Shot Color (OSC) camera using a Dual or Tri Band Narrowband filter
  • Image is non-linear…so fully processed

Step 1 – Split the Channels

In order to re-assign the channels, you have to split the normal image into Red, Green and Blue channels, I found this to work better on a fully processed “Non-Linear” image as above, once this was done, I renamed the images in PixInsight to “Ha” – Red Channel, “OIII” – Blue Channel and “SII” – Green Channel, this makes it easier for Pixelmath in PixInsight to work with the image names. Once this was done, I used PixelMath to create a new image stack with the channels assigned, and this is how PixelMath was configured

Red Channel = SII
Green Channel = 0.8*Ha + 0.2*OIII
Blue Channel = OIII

Once applied this produced the following image stack (do not close the Ha, OIII or SII images, you will need these later on):

SHO Combined image from PixelMath

Step 2 – Reduce Magenta saturation

As you can see from the above image, some of the brighter stars have a magenta hue around them, so to reduce this, I use the ColorMask plugin in PixInsight (You will need to download this), and selected Magenta

ColorMask tool with Magenta selected

When you click on OK, it will create the Magenta Mask which would look something like this:

Once the mask has been applied to the image, I then use Curves Transformation to reduce the saturation which will reduce the Magenta in the image


The result in reducing the magenta can be seen in this image, you will notice there is now no longer a hue around the brighter stars

Result after Magenta Saturation reduced using Magenta ColorMask and Curves Transformation

Step 4 – ColorMask – Green

Again using the Color Mask tool, I want to select the green channel, as we will want to manipulate most of the green here to red, so again ColorMask:

This then produced a mask that looks like the following:

Step 5 – Manipulate the Green Data

Once the Green Mask has been applied to the image, since most of the data in the image is green, we are looking to manipulate that data to turn it golden yellow, so for this we use the Curves Transformation again

The above Curves transformation was applied to the image three times whilst the the green mask was still im place, and this resulted in the following image changes:

Resulting image after green data manipulated in the red channel using Curves Transformation

So as you can see we are starting to see the vibrant colours associated with Hubble Palette images

Step 6 – Create a Starless version of the OIII Data

Now remember I said not to close out the separated channel images, this is because we are going to want ot bring out the blue in the image without affecting the stars, so for this we will turn the OIII image into a starless version by using the StarNet tool in PixInsight

Here’s the OIII Image before we apply StarNet star removal:

Default settings used in the StarNet process

This resulted in the following OIII image with no stars:

OIII Data with stars removed using StarNet Process

Step 7 – Range Selection on OIII Data

Because we do not want to affect the whole image, we will use the range selection tool on the starless OIII image to select areas we wish to manipulate, now we have to be careful that the changes we make are not too “Sharp” that they cause blotchy areas, so within the range selection tool, not only do we change the upper limit to suit the range we want to create the mask for, but we also need to change the fuzziness and smoothness settings to make it more blended, these are the setings I used:

Which resulted in the following range mask

Rangemask created using the RangeMask process

Step 8 – Bring out the Blue with Curves Transformation

We apply the Range Mask to the SHO Image so that we can bring out the Blue in the section of the nebula where the OIII resides, with the range mask applied we will use the Curves Transformation Process again as follows:

Curves transformation process to increase blue, reduce red and increase saturation of image with rangemask applied

The result of which is:

Result after first curves transformation with RangeMask applied

As you can see we have started to bring out the blue data, but we are not quite there yet, with the range mask still applied, we will go again with the curves transformation only this time, just reducing the red element:


The result of the 2nd curves transformation with the Range Mask is as follows:

Resulting image after 2nd pass with Curves Transformation to remove the red elemtn in the range mask

Step 9 – Apply Saturation against a luminance mask

On the above image, we extract out the luminance and apply as a mask to the image, and we then use the Curves Transformation for the final time to boost the saturation to the luminance

Luminance Mask to be applied to image
Curves Transformation with Luminance Mask applied

Final Image

I repeated the same process on my Elephant’s Trunk Nebula that I acquired the data when testing out the ASI2400MC Pro and this was the resulting image:

Elephant’s Trunk Nebula, 19x300S at Gain 26 on the ASI2400MC Pro with the Optolong L-eXtreme Filter using the workflow in this article

I hope this tutorial helps in producing your SHO images from your OSC Narrowband images, I know many of my followers have been waiting for me to write this up, so enjoy and share.

The source to the halo around bright stars

When I moved to the Sky-Watcher Quattro telescope I noticed some bizzare halo’s around bright stars in my images, this was evident in both my Atik 383L+ CCD Camera as well as my QHY183M ColdMOS Camera when using the Quattro 8-CF at F4, if you browse my galleries you will see what I mean, and it was more noticable in my Narrowband images. Below is one of my recent images where you can see the halo around Magnitude 3.9 star 15 Mon in the Christmas Tree Cluster / NGC2264.

I contacted Baader back in February 2019 since all of my filters were Baader, and I noticed that the Halo was present in all of my filters but significantly less in Red, but more prevalent in Narrowband filters, so the logical cause would be the filters. Baader immediately dismissed this to be the fault of their filters and suggested that my Coma Corrector be the root cause.

Not convinced that the Coma Corrector was causing the issue, I did some research online and came across a brilliant page on the Astronomik website where they claim to have resolved the majority of the Halo issue, and after reading the following line from the page I was convinced the filters were my issue:

In recent years very fast optical systems have become popular for imaging. The energy in a filter induced halo grows exponentially as the f-ratio decreases. Additional to this, the smaller the FWHM band pass of the filter, the stronger the halo.

The above line described my issue perfectly so I mentioned this to Baader who again dismissed the possibility of it being their filters and again put the blame firmly to my optical train. Again not happy, I contacted Astronomik and Eric emailed me back very promptly and offered to send me out one of their 6nm Ha filters to test. A few days ago the filter arrived and I was able to perform some testing against the Baader filter also for comparison on the same star.

Since the star in my image above was of magnitude 3.9, I wanted to find something similar, so I found star Alhaud VI and proceeded to obtain 15x300S Exposures for each filter, and here are the results:

Astronomik 6nm HA filter, 15x300S with Darks and Flats applied
Baader 7nm Ha filter, 15x300S with Darks and Flats applied

So as you can see the Baader filter shows a high amount of Halo around the bright star and the Astronomik filter does not, now if this was something to do with the rest of the optical train there would be evidence in the Astronomik filter also.

Now I agree there will be some reflection in the optical train, all that glass in the coma corrector, the glass on the camera etc, so I thought I would have a look at both images in a bit more detail, zoomed in on the stars there is what appears to be a slight halo in the same place on both images:

Astronomik 6nm Ha Filter
Baader 7nm Ha Filter

So both filters show the Inner Halo which in my opinion would not be visible in an image, but again clearly the Baader filter has some reflection issues happening as you can clearly see two additional Halos. The interesting thing about all three Halos is that the central one visible in both filters has no relationship to the distances between the other two in the Baader, however the two outer Halos on the baader are the same distance apart as the middle halo is from the star, so clearly this is some sort of reflection.

Conclusion:
Astronomik have done a fantastic job at eliminating Halo artifacts around bright stars, clearly the Baader filters are causing major Halo artifacts because if this was the optical train then it would be evident in the Astronimik filters also, I suspect that the Baader filters are not optimised for faster focal ratio imaging systems. I have provided this information to Baader and await a response from them.

Good job Astronomik Filters

NGC281 – Pacman Nebula in Hubble Palette Narrowband

NGC281 – Pacman Nebula

NGC 281 is located in the constellation of Cassiopeia and part of the Perseus Spiral Arm. It includes the open cluster IC 1590, the multiple star HD 5005, and several Bok globules. Colloquially, NGC 281 is also known as the Pacman Nebula for its resemblance to the video game character.

The Image consists of
21x600S in SII – Mapped to Red
21x600S in HA – Mapped to Green
21x600S in OIII – Mapped to Blue
25 Darks and Flats

The HA Channel was also used as a Luminosity layer

Equipment Used
Mount: Sky-Watcher EQ8 Pro
Imaging Camera: Atik Cameras 383L+ Mono CCD Cooled to -20C
Imaging Scope: Sky-Watcher Quattro 8-CF 8″ F4 Newtonian
Coma Corrector: Sky-Watcher Aplanatic Coma Corrector
Guide Camera: Qhyccd QHY5L-II
Guide Scope: Celestron Telescopes C80ED Refractor
Filter Wheel: Starlight Xpress Ltd 7x36mm EFW
Filters: Baader Planetarium 7nm HA, OIII and SII 36mm unmounted

Software Used:
Image Acquisition: Main Sequence Software Sequence Generator Pro
Guiding: PHD2
Stacking and Pre-Processing: Maxim-DL
Post Processing: Photoshop, Noise Ninja