The images you see on this site are not single, simple photographs. Because deep-space objects are incredibly faint, often invisible to the naked eye, capturing them requires an "accumulation" of light over long periods. My Seestar S30 Pro telescope utilises a high-sensitivity Sony IMX585 sensor to record these distant photons.

Instead of taking one long photo (which would be ruined by satellites, wind, or tracking errors), I capture dozens of short exposures, typically between 10 and 60 seconds each. During this time, the telescope's on-board computer accurately tracks the object's movement across the sky, ensuring that the light hits the same pixels on the sensor for every frame. This phase of the project is a patient game of gathering as much raw data as possible.

To follow the stars across the sky without distortion, the telescope must be perfectly synchronized with the Earth’s rotation. I operate my equipment in Equatorial Mode, which means the telescope rotates around a single axis that points directly at the North Star (Polaris). For this to work accurately, the foundation must be a "perfect" starting point.

Before I begin imaging, I use a specialised Astro Essentials Leveling Base and a ZWO TH10 Fluid Head to ensure the mount is perfectly level relative to the horizon. Even a tiny physical tilt of just one or two degrees at the base will cause the telescope’s tracking arc to miss the stars' path, resulting in "trailed" or elongated stars. A perfectly level base means the telescope's motors can track with sub-pixel precision, keeping the distant object sharp and the stars perfectly round.

Once the observation is complete, I am left with dozens of individual raw files. At this stage, each photo is quite "noisy" and filled with digital grain and random artifacts generated by the camera sensor. At this stage I switch to a astro-photography image processing suite called 'PixInsight' and my first task is to 'stack' the raw images together.

The software mathematically aligns every star in every photo and then calculates the average value for every single pixel. Because the digital noise is random but the light from the object is constant, the noise cancels itself out during the averaging. This results in a single "Master" image that is remarkably smooth and contains a much higher level of detail than any single frame could provide.

A number of options within PixInsight are used to clean up the image.

Background extraction is used, to remove artificial light gradients and atmospheric haze, creating a neutral "floor" for the signal. The image is then color-calibrated using spectrophotometric analysis, which cross-references the stars in the field against a database of known star colors to achieve scientific color accuracy. To further refine this, any residual green cast, a common byproduct of the sensor used in the scope, is neutralised to balance the colour channels.

Finally, the stars themselves are analyzed and corrected for optical distortions or tracking eccentricities, pulling them back into perfect circles while the data is still linear.

After stacking and some image cleaning is complete, the resulting master image appears almost entirely black to the human eye. This is because the camera records the light in a "linear" format meaning the data is mathematically accurate, but it is all compressed into the deepest shadows of the digital file.

To reveal the hidden detail, I perform a 'Multiscale Adaptive Stretch' in PixInsight. This process expands the faint, low-signal information (a nebula, for example) into a visible range while ensuring the brightest areas (the star cores) don't become overexposed. This is the critical moment where the true colours and complex structures of the deep-space object are revealed.

One of the most complex challenges in processing an object (a nebula, for example) is the presence of thousands of stars. Because stars are significantly more luminous than the surrounding gas, any attempt to sharpen or brighten the object would cause the stars to "bloat" and dominate the frame.

To prevent this, I use a mathematical process to separate the image into two distinct data layers. By identifying the specific Point Spread Function (the shape and brightness) of the stars, the software can algorithmically subtract them from the image. This leaves me with a "Starless" layer containing key detail of the object and a "Stars-Only" layer. By isolating the data in this way, I can refine the delicate textures of the object without affecting the stars, or vice versa.

Once the initial cleanup is done, I use curves to carefully "stretch" the light and reveal the hidden details. By creating an S-curve, I can darken the background to a deep, natural black while simultaneously pulling the faint detail of the object's shadows. During this process, I also adjust the luminosity to add a sense of 3D depth to the dust lanes and boost the saturation to bring out the natural colours. The goal is to create a high-contrast look where the object pops vividly against a velvet-dark sky without losing the delicate outer gas.

The final step is the Recombination of the 'stars' and 'starless' processed layers. This merges the two datasets into a single, cohesive image where stars are sharp and the object's intricate details are fully visible.

The final image is then exported into a web-ready format. I save a lossless master for my records and a high-resolution version for the gallery.