How to process your first deep-sky image

This workflow is for multiple tracked exposures of the same deep-sky target from a colour camera or DSLR. For a wide Milky Way scene with a foreground, the moving stars and stationary land need different alignment; see our Milky Way guide for capture advice. For a single phone image, start with the phone astrophotography guide.

Keep an untouched copy of your original RAW files. Sort them into light frames (the target), darks, flats and bias or flat-dark frames if you took them. Check your camera and software instructions before choosing calibration frames: settings, exposure and temperature matching matter, and the correct combination depends on the camera and workflow. Siril's manual pre-processing guide explains the roles and order of these files.

Open a sample of the light frames at full size. Exclude obvious cloud, severe trailing, loss of focus and frames ruined by vibration. A satellite trail does not automatically mean an exposure is useless: a suitable rejection method may remove isolated trails when enough comparable frames are available. Check the stacked result rather than assuming it worked.

Calibration uses matching reference frames to address unwanted camera signal and uneven illumination. Flats can correct vignetting and dust shadows when captured with the same optical arrangement; darks address repeatable dark signal. They do not remove every kind of noise or rescue a blurred image. Align the stars, then stack the usable lights. Stacking increases the contribution of repeatable signal relative to random variation, but the improvement depends on the data and method; it is not a promise that all noise disappears.

For a free worked example, follow Siril's first-steps tutorial. It describes the folder names and a colour-camera preprocessing script. Read the requirements for the script you actually use rather than putting unmatched calibration files into folders just to make it run.

Save a separate copy of the stacked file before editing. Edges often contain uneven borders from alignment: crop those first so they do not influence later adjustments. Look for doubled stars, diagonal bands, dust rings or a strong gradient. If you see them, return to the relevant frames or calibration step instead of trying to hide the problem with aggressive sharpening. Our satellite trails guide explains how to compare affected frames and check rejection settings.

A linear stack normally looks very dark. A temporary screen preview can make it easier to inspect without changing the saved data. Siril's processing tutorial distinguishes its display stretch from the later edits applied to the image.

A gradient can come from sky brightness, moonlight or other uneven illumination. Background extraction estimates that uneven signal and subtracts it. Place samples on genuine background, avoiding bright stars, galaxies and extended nebulosity. If a target fills much of the frame, there may be little clean background to sample; compare the result with the original so that faint structure is not removed by mistake.

For a colour image, use a colour-calibration method supported by your software and data. Siril documents photometric calibration in its full workflow; it relies on a successfully identified star field. If that fails, check the image and metadata, or use a careful manual balance. Neither method establishes one universally correct colour for every processed image.

Stretching changes how the faint values in a linear stack are displayed, making the target visible on an ordinary screen. Work in small steps and watch the histogram as well as the image. Do not push the black point so far that the background becomes a featureless black; faint outer structure may disappear with it. Avoid clipping bright star cores, and undo a step if the image loses detail.

A useful first pass is restrained: reveal the target, leave the background slightly above black, and save a working copy. If faint detail is still indistinguishable from noise, stronger stretching will brighten that noise too. More suitable exposure time, darker skies or improved calibration may help on a future night.

Adjust colour saturation gently and use noise reduction or sharpening only where it improves the image. Zoom out as well as in: a smooth background at full size can look artificial if fine stars and nebulosity have vanished. Compare with the earlier stack and check for dark halos, coloured fringes and clipped stars. Keep a high-quality working file, then export a JPEG or PNG copy for sharing; retain your original and stacked files so you can reprocess them later.

For your next session, the budget astrophotography guide covers capture choices before spending on more equipment.

Reviewed 14 November 2026. This workflow draws on Siril's first steps, manual pre-processing and full image-processing tutorial. Menu names and scripts can change between versions. This is a starting workflow, not a claim of a hands-on comparison of software or camera models.