Summary of "SASpro Beginner Tutorial: Free Data to Download and Get some Basic Processing Under Your Belt"
Summary — main ideas and lessons
This is a beginner walkthrough for getting started with an astrophotography processor (referred to in the video as “SEI Astro / [S]uite Pro”) using a single master FITS file (NGC 1333) provided as practice data.
The video demonstrates a practical, step-by-step linear → nonlinear processing workflow aimed at beginners:
- Open the data and fix basic calibration/visual issues.
- Remove gradients and perform color calibration (background + stars).
- Correct star aberrations, selectively sharpen/denoise.
- Remove stars for stretching, stretch and enhance the nebula, then recombine stars.
- Clean blemishes, apply local structure/HDR-like adjustments, and save final images.
Key teaching points and philosophy:
- Start small: use simple datasets for learning rather than huge, multi-filter stacks.
- Work iteratively and conservatively: make small adjustments and avoid over-sharpening or over-denoising in a single pass.
- Preserve data: avoid clipping blacks (do not make the sky “pure black” — keep minimums above zero).
- Use separate steps for background neutralization and star white-balance (color calibration).
- Keep originals and versions (history/duplicate views) so you can revert and compare.
Detailed step-by-step methodology
The following ordered steps reflect the practical workflow shown in the video, with tips and recommended parameter hints.
-
Download / open
- Get the latest program release from the project’s GitHub or the mirror Google Drive (links shown in the video).
- Use the included beginner practice data: a single master FITS of NGC 1333.
- Open the master FITS in the program (double-click the canvas or use File → Open).
-
Initial display and cropping
- Use automated display stretch (hotkey “A”) for visualization. For one-shot color cameras, prefer an unlinear display to avoid a teal-dominant preview.
- Crop away stacking artifacts with the Crop tool.
-
Inspect histograms and baseline alignment
- Open the histogram (toggle log axis) to inspect channel minimums and medians.
- If channel minimums differ, remove the minimum pedestal to align zeros on each channel:
- Use “Pedestal removal” (button or Functions → Pedestal removal).
- This aligns channels better without performing median matching/linear fits yet.
-
Remove background gradients
- Two options shown:
- AB (adaptive/background sampling):
- Turn Auto-stretch on, draw exclusion zones around nebulous/dense features so sampling uses true sky background, and apply.
- Pros: you control exclusion regions to avoid sampling bright nebulosity or galaxy cores.
- Grappert / Graexert remove-gradient:
- Simpler: run remove-gradient with an appropriate smoothing parameter (set smoothing high enough to avoid removing real small features).
- Optionally denoise as part of this step.
- AB (adaptive/background sampling):
- Duplicate the document before testing different methods so you can compare results.
- Two options shown:
-
Color calibration (two-stage)
- Background Neutralization:
- Use an eyedropper or “Find background” to sample a representative sky patch and apply background neutralization.
- This aligns the histogram for the background.
- Star White Balance / Color Calibration:
- Run the star-based white balance (SPCC / SFCC in advanced modes). The tool detects stars and neutralizes their colors.
- Use an adequate sample count (e.g., 300+ stars); the program reports the count when applied.
- Verify visually and with a linked display stretch to confirm a neutral background and correct star colors.
- Background Neutralization:
-
Star-shape correction
- Run Ricardo’s aberration remover to correct elongated or misshapen stars and field-wide aberrations.
- Recommended even if stars look fairly good; the model is being updated over time.
-
Sharpening and initial denoising
- Use a tool like “Cosmic Clarity” to combine sharpening and denoise:
- Modes: sharpen only, denoise only, or both. The video uses both.
- Conservative example settings: stellar sharpening ~0.35, non-stellar ~0.5.
- Denoising: start modestly; avoid heavy color denoise until you gauge noise levels.
- Note: without a capable GPU, these operations may be slow.
- Use a tool like “Cosmic Clarity” to combine sharpening and denoise:
-
Remove stars before nonlinear stretch
- Use StarNet (preferred) or DarkStar to create a starless image and a stars-only image.
- Confirm the image is still linear (answer “yes” only if you haven’t stretched yet).
- Keep both layers/views so you can process the starless data separately and recompose later.
-
Initial linear → nonlinear stretch (statistical stretch)
- Use Statistical Stretch on linked channels (since the image is color-calibrated).
- Choose a conservative stretch amount (example: 0.13 rather than the full 0.25) for objects with limited nebulosity.
- Stretch stars separately using a Star Stretch tool if available.
-
Mid-process contrast and shaping - Use Curves for localized contrast control: - Place anchors to protect deep darks (avoid raising them to zero). - Apply small, incremental curve adjustments. - Use Hyperbolic Stretch for finer control: - Control-click an area to set a symmetry point (targeted contrast control). - Parameters: A (S-curve strength), B (slope around symmetry), HP (highlight protection) to avoid blowing cores. - Apply in small steps.
-
Additional denoising (after some non-linear processing) - Run another denoise pass (Cosmic Clarity or Grappert) after some stretch to clean remaining noise. - Option: process RGB channels separately for better color denoising (slower) or process as a single-color pass (faster). - Example: Cosmic Clarity denoise strength ~0.8 (adjust per image).
-
Local detail enhancement - Wave-scale dark structure enhancement: - Enhances darker structures (a “dark HDR” effect). Adjust mask gamma to control what the mask selects. - Useful to show faint wispy structures. - Wave-scale HDR: - Compresses bright range and reveals detail in bright regions. Defaults are a good starting point. - Use these tools conservatively — they can reveal a lot quickly.
-
Blemish removal and handling stellar halos - Use Blemish Blaster (spot removal) for bright blemishes and star artifacts: - Adjust brush radius and feathering to match blemish size. - Be careful removing star halos — halos carry color; over-removal can look unnatural. Consider lowering opacity on removals. - After blemish edits, click “Apply to document” to commit changes.
-
Recompose stars back into image - Recombine the starless image with the stars-only image: - Blend mode: Screen (typical). - Preview and inspect for problematic areas. - After recomposing, blemish-blast any remaining problematic star artifacts (tone down halos rather than removing them entirely).
-
Final local tweaks and export - Optional: use color masks to enhance specific colors (e.g., bring out reds), add local contrast enhancements, or additional structure boosting. - Avoid clipping: check pixel readouts and histogram minimums — don’t let any channel hit zero. - Suggested targets: - Minimums: ~0.02–0.03 (keeps sky dark without clipping). - Median: personal preference; ~0.1 as a starting target. - Save your work: - Save as FITS if you need master/scientific preservation. - Save flattened versions for sharing as JPEG or PNG.
Practical tips & warnings
- Start with simpler/smaller datasets when learning (avoid massive multi-filter stacks).
- Use exclusion zones when sampling backgrounds for gradient removal to avoid measuring nebulosity as background.
- Don’t over-sharpen or over-denoise early — iterative mild passes work better.
- Don’t aim for a pure-black sky; clipping to zero destroys data.
- If GPU is not available, several operations (Cosmic Clarity, StarNet) will be much slower.
- Keep a history / snapshots so you can revert and compare steps.
Tools, plugins and features mentioned
- SEI Astro (the program demonstrated)
- Download locations: project GitHub and Google Drive mirror
- Practice data: NGC 1333 (single master FITS)
- Core features/tools referenced:
- Display Stretch (linked / unlinear)
- Crop tool
- Histogram (with log axis)
- Pedestal removal (minimum subtraction)
- AB gradient removal (adaptive/background sampling)
- Grappert (remove-gradient and denoise)
- Background Neutralization (eyedropper / Find background)
- Star White Balance / Color Calibration (SPCC / SFCC)
- Ricardo’s aberration remover
- Cosmic Clarity (sharpening + denoise)
- StarNet (star removal)
- DarkStar (alternative star removal)
- Statistical Stretch (linear → nonlinear stretch)
- Star Stretch
- Curves and Hyperbolic Stretch (symmetry point, A/B, HP highlight protection)
- History Explorer (version/step browser)
- Blemish Blaster (spot removal)
- Wave-scale dark structure enhancing (dark HDR-like)
- Wave-scale HDR
- Plus-star recomposition (blend mode: Screen)
- Export: FITS, JPEG, PNG
Speakers / sources featured
- The video presenter / narrator (host of the tutorial; introduces “Welcome to SEI Astro” and demonstrates the workflow).
- Tool/authors referenced:
- Ricardo (author of the aberration remover)
- StarNet (star removal tool)
- DarkStar (alternative star removal)
- Cosmic Clarity (presenter indicates they developed it)
- Grappert (gradient removal / denoise)
(End — no further conversation.)
Category
Educational
Share this summary
Is the summary off?
If you think the summary is inaccurate, you can reprocess it with the latest model.