I first saw and admired those amazing astrophotography images in the Hubble Palette colours about two years ago and since then I have been unable to find a way to easily create this effect with my many images. Generally, I take images in either narrowband or broadband RGB. They both end up with many shades of red. I have achieved some great pictures this way but I finally decided to give it a serious go and convert my images to the Hubble Pallete. The result of my first image has blown my mind!
Here is my very first image:
The deep sky object I chose to image was the Tadpole Nebula (IC 410). I took the photos in January 2024 over two nights. The images are composed of 400 x 2 minutes guided exposures using my CEM 26 equatorial mount and a Celestron 130mm Newtonian reflector OTA. The camera I used was the ZWO ASI533 MCPRO (OSC – one shot colour camera set to 0 degrees C cooling).
The focal length is 650mm and the sky conditions were stable and Bortle 5 so not too much light pollution. The moon was only 10-20% illuminated.
How did I convert my image to Hubble Palette Colours?
After a lot of experimenting and some sleepless nights trying to get to grips with this I found the easiest method was as follows:
- Load the original RGB stacked photo into Siril
- Split the image into three channels Red, Green and Blue.
- Recompose the image using Pixel math to create a simulated Hubble Palette.
- Process the image adjusting the colours to my taste in Photoshop.
This process with practice should only take about 10 minutes I think. I tried several other methods but they didn’t work for me or took too long. I have to thank ChatGPT for assisting me in my research and for the suggestions how to do this. AI can only guide us the final trial and error was up to me and this workflow worked liked magic!
See in detail exactly how to do this for yourself. Watch my YouTube video:
Maybe you can try it with your own images and let me know how it goes for you.
Which image do you prefer? Share your thoughts below in the comments section – I’d love to know what you think!
How did I create the Hubble Palette images?
I used the free software called Siril to stack the images. The image was then split into three channels, Red, Green and Blue. The image was then recomposed using Pixel Math.
Pixel Math is a tool in Siril that enables you to create an image and mix different proportions of each colour into the three RGB channels. Changing the formaulae you use changes the proportions of one colour channel to another and therefore creates different colour palettes.
Ideally, for the Hubble Palette you should collect your data using a narrowband filter. I used a ZWO Duo-band filter that collects two bands of light, one in the Hydrogen Alpha and the other in Oxygen III. You can also collect light using a broadband RGB filter but the first method is better because there will be less light pollution due to the narrower bandpass and HA and OIII will be separated into the RED and Blue/ Green channels.
Normal processing of narrowband or broadband data often gives very red images, especially for emission nebulae which are composed of mainly HA gases. Pixel Math lets us rearrange these colours and map them to the signal our filters detect in the different parts of the light spectrum. Here is the formula I used for each channel in Pixel Math:
Red channel = 0.8 x Red
Green Channel = Red
Blue Channel= 0.5 Blue + 0.5 Green
This creates the basic hubble colours in the image which you will then process in your image editing program. I use Photoshop.
The Basic Composition of SHO Hubble Images
The three channels Red, Green and Blue are representative of three different gases in the deep sky object (nebula or galaxy). These gases produce different frequencies of light which the filter used splits up into the different colour channels in an RGB image. Here is the standard RGB composition compared to Hubble Pallette:
| Standard RGB Palette | Hubble (SHO) Palette |
| Red (HA – hydrogen alpha) | Red (SIII – sulphur) |
| Green | Green (HA) |
| Blue (OIII – oxygen) | Blue OIII |
The best way to create a true hubble palette image is to use three filters for HA, SII and OIII and combine the data into the correct channels. There are triband filters that pass only these three light frequencies or dual band filters that only pass HA and OIII (like the ZWO filter I used). When using a duo-band filter it will be necessary to create a simulated sulphur channel and Pixel Math is one way to do this.
Other Images I have converted into Hubble Palette
Please understand that these are not actual SHO images because I have simulated the three channels using Pixel Math in Siril. I will upload more images here as I create them and explain more about the processes I used to develop the colours and improve the details. I have tried different Pixel Math formulae out and some work really well.
Other Pixel math Formulae to Try
It’s clear to see hear the art that is involved in astrophotgraphy because the images I have created are all very different. Many online images I see with this colour scheme are very orange and blue and so I will now create a few images to look similar to that.
Here are a few different formulae to try and experiment with. Which is the best depends on the look you are aiming for:
| Channels | Standard Hubble | Enhanced OIII | Balanced | Creative |
| R | 0.8 * R + 0.2 * B | 0.6 * R + 0.4 * B | 0.5 * R + 0.5 * B | R |
| G | R | R | R | 0.5 * R + 0.5 * B |
| B | B | B | B | B |
Leave a Reply