Monochrome astro images taken with narrowband filters give us superb contrasts and details.
To colourise them, we assign each filter’s data to one or more of the main colour channels, red (R), green (G) and blue (B).
While there are three main narrowband types, hydrogen-alpha (Ha), oxygen (OIII) and sulphur (SII), we don’t need all three for beautiful results.
If we map Ha data to the R channel and OIII data to the green and blue channels, we make what's called a HOO-palette image.
For more advice read our guide to astrophotography image processing.
Using Siril
Siril is a free image-processing program that, with its tools StarNet and Pixel Math, provides an end-to-end imaging workflow for this process, including stacking, removing stars and improving background gradients.
Pixel Math is particularly interesting as it allows you to carry out mathematical operations on each image pixel.
This means that, among other things, you can add, subtract, multiply or divide the images created by your narrowband filters to control the strength of each in your RGB channels.
Here we’ll take you through the steps we followed using Pixel Math in Siril to combine and colourise narrowband data, to produce a HOO image of the North America Nebula.
Getting started
Our start images were two stacks of narrowband data: one set of Ha and the other of OIII (see Before, above).
We opened Siril and assigned our working directory (the folder where our stacked data was located) by clicking the home symbol (highlighted, image above) and choosing our folder location.
We opened both stacked images by selecting Conversion, and added our files by clicking the + button.
We selected our two images and clicked Add.
Our images appeared and could be viewed separately by clicking the icon on the bottom-right of the screen (highlighted, image above).
To align our two images, we clicked on Registration, selecting Global Star Alignment under Registration method and then Go register (highlighted, image below).
We checked that the registration and alignment process was successful by clicking Console, which read Registration successful at the bottom. At this point we had aligned but not combined our two stacks.
If the alignment process leaves one of your images with black borders, now is the time to crop these out.
Click and drag to highlight the area to keep, then right-click and select Crop Sequence. The crop applies to both Ha and OIII images.
Combining the images
We could now use Pixel Math to combine our aligned images and map our filters to the RGB channels.
We clicked Image Processing and selected Pixel Math. The Pixel Math window opened (highlighted, image below).
Our two images appeared under the RGB channels.
Each was assigned a default variable, in our case ‘l1’ and ‘l2’.
We changed these to ‘Ha’ and ‘OIII’ by clicking on the box and typing over (highlighted, image above).
It’s important to type these variables exactly as you assign them, in the RGB boxes above.
Unticking Use single RGB/K expression let us modify each box.
The image above shows a basic HOO image; we simply typed ‘Ha’ into the R box, and ‘OIII’ into G and B.
Clicking Apply showed the result of our RGB combination. Note, though, that there are many other combinations to try.
For example, you could multiply your Ha with your OIII data (Ha*OIII) or even double the Ha (Ha*2).
Simply type in your required changes exactly like a mathematical formula.
Our final image (shown at the top of the page) was then ready for further processing in Siril.
3 quick tips
- If Global Star Alignment doesn't successfully align your images, try the two/three-star registration selection method.
- Search online for narrowband combinations for formulae to paste into Pixel Math.
- Check the Function/Operator box (highlighted, screenshot 3) for the correct syntax for formulae.
Are you an astrophotographer? Send us your images to contactus@skyatnightmagazine.com.
This guide appeared in the June 2024 issue of BBC Sky at Night Magazine.
