Veralux HMS: What is behind the new stretching method for astrophotography?

Veralux HMS HyperMetric Stretch (HMS)

1. Introduction: The Algorithmic Revolution of Image Development

Astrophotography is at a pivotal turning point. While image processing has been a subjective art form for decades, dominated by manual curve adjustments and intuitive color correction, we are now witnessing the rise of physics-based algorithms. In this context, Veralux HyperMetric Stretch (HMS) is not just another tool in the PixInsight palette; it represents a fundamental shift in data handling philosophy: the transition from purely visual manipulation to photometric integrity.

The core problem of any astrophotographic processing is the discrepancy between the linear nature of the sensor and the logarithmic perception of the human eye. A CCD or CMOS sensor counts photons linearly; if the exposure time doubles, the signal doubles. The human eye, however, requires a non-linear compression of the dynamic range to perceive details in the shadows (the background of space) and the highlights (stars and galaxy cores) simultaneously. Traditional methods like Histogram Transformation (HT) solve this by boosting brightness, but often destroy the subtle relationships between color channels. The result is washed-out stars and color-distorted nebulae structures.

Veralux HMS aims to solve this dilemma through an approach termed “HyperMetric.” It promises to preserve the physical relationships of light waves—their vectors in color space—during aggressive brightness enhancement.¹ This report analyzes the technical architecture, practical implementation, and strategic advantages of this method. It is designed as a comprehensive guide leading the reader from the theoretical foundations of quantum efficiency to pixel-perfect application in the PixInsight workflow.

2. Theoretical Foundations: Physics of Light and Sensor Technology

To deeply understand how Veralux HMS works, it is essential to illuminate the physical deficits plaguing traditional stretching methods and how Veralux addresses them by integrating sensor data.

2.1 The Linearity Paradox

In digital astrophotography, we store data in FITS or XISF formats, which exist in a “linear state.” This means the pixel value (ADU – Analog Digital Unit) is directly proportional to the number of collected electrons.

• Black: Background values are often close to 0 (or a bias offset).
• White: Saturated stars are at the maximum value (e.g., 65,535 for 16-bit).

The problem arises during visualization. A linear image appears almost black on a monitor because the objects of interest (nebulae, galaxy arms) are only a few percent above the background noise. To make them visible, we must raise the curve extremely steeply (High-Contrast Stretch).

Traditional Histogram Transformation (HT):

HT applies a global transfer function to all pixels. However, if we look at a pixel consisting of Red (R), Green (G), and Blue (B), and we drastically raise all three values, they inevitably approach the white point (saturation).

• Example: A star is orange ($R=2000, G=1000, B=500$). After a strong HT, the values might be $R=60000, G=58000, B=55000$. The ratio has shifted; the star now looks almost white. The color information—the thermal signature of the star—is lost.²

2.2 The Veralux HMS HyperMetric Solution: Vector Preservation

Veralux HMS relies on mathematical work by Riccardo Paterniti and differs fundamentally from scalar methods.¹ While classical methods treat pixels as isolated brightness values, Veralux views each pixel as a vector in three-dimensional color space.

A vector has two main properties:

1. Length (Magnitude): Represents luminance or brightness.
2. Direction: Represents chromaticity or color.

The goal of Veralux is the decoupling of luminance geometry from chromatic vectors.¹ The algorithm applies non-linear hyperbolic stretching exclusively to the length of the vector while attempting to rigorously maintain its direction (the color). This prevents the typical “bleeding” of colors into white that occurs with aggressive stretching.

2.3 The Role of Quantum Efficiency (QE)

A revolutionary feature of Veralux is the inclusion of specific hardware characteristics. Sensors are not perfectly neutral; their sensitivity varies depending on the wavelength of light. This is described by the Quantum Efficiency (QE) curve.

• A sensor like the Sony IMX571 (used in cameras like the ZWO ASI2600MC) has a different spectral response than an older Kodak KAF-8300.
• Traditional methods often calculate luminance using the standard formula for human vision ($L = 0.21 R + 0.72 G + 0.07 B$). This is correct for daylight photography but ignores how astronomical sensors collect photons, particularly in the H-Alpha range (Red).

Veralux uses databases of QE weights (Quantum Efficiency Weights).¹ By having the user select their specific sensor, the script calibrates the luminance calculation to match the actual photon yield of the chip. This results in a “True Color” base that is scientifically more valid than generic luminance extractions.

3. Installation and Configuration Veralux HMS: A Technical Deep Dive

Integrating Veralux HMS into PixInsight requires an understanding of the software’s repository system. Unlike simple scripts that you copy into a folder, Veralux is installed via a managed update system, simplifying updates but making the initial setup more complex.

3.1 The Source of Truth: The Correct Repository

There is often confusion in the community about the script’s origin, as it is frequently discussed in videos by channels like “Hidden Light Photography” (HLP) or “Deep Space Astro.” However, it is essential to understand that Hidden Light Photography does not host the script in its own repository.⁴

The official porting project for PixInsight is managed on GitHub by user killerciao under the project name VeraLuxPorting.¹

• Original Author (Mathematics): Riccardo Paterniti (developed for Siril/Python).
• PixInsight Port: killerciao.

The Critical URL:

For installation, you need the repository URL. This is typically found on the VeraLuxPorting GitHub page. It usually looks like this:

https://raw.githubusercontent.com/killerciao/VeraLuxPorting/main

(Note: PixInsight often requires the direct path to updates.xri or the repository root. Always check the README of the GitHub project for the most current URL).

3.2 Installation Process: Step by Step

To correctly install Veralux HMS and avoid malfunctions, follow this detailed protocol:

1. Open Repository Management:
   Start PixInsight. Navigate in the menu bar to Resources > Updates > Manage Repositories. This is the command center for all external scripts.4
2. Add the Source:
   Click the Add button. An input field appears. Paste the exact URL mentioned above. Ensure there are no spaces at the beginning or end. Click OK. The list of repositories should now contain the new entry.
3. Check for Updates:
   Close the Manager window (OK). Now navigate to Resources > Updates > Check for Updates.
   PixInsight will connect to the servers. If the URL is correct, a package (often labeled “Veralux Script” or “VeraLuxPorting”) will appear in the list of available updates.
4. Execute Installation:
   Check the package and click Apply. PixInsight downloads the files.
   Important: After the download, a message appears stating that updates will be installed upon restart.
5. The Restart:
   Manually close PixInsight. Upon closing (or at the next start, depending on the OS), the update process begins. A small window shows the progress of script compilation.
6. Verification:
   After the restart, you will typically find the script under Script > Utilities > Veralux HMS or a dedicated category Veralux.
   Troubleshooting: If the script does not appear in the menu, go to Script > Feature Scripts. Click Regenerate. This forces PixInsight to rescan all script folders and register forgotten entries.5

3.3 Troubleshooting Network Issues

In some cases, especially in corporate networks or with strict firewalls, PixInsight cannot connect to GitHub (“raw.githubusercontent.com”).

• Symptom: Error message during “Check for Updates.”
• Solution: Download the script manually as a .zip from the GitHub page.⁷ Unzip it into a permanent folder (not the download folder). Then use Script > Feature Scripts > Add, navigate to this folder, and add the script manually. This bypasses the repository manager but means you will not receive automatic updates.

4. The User Interface: Parameters and Modes in Detail

The GUI (Graphical User Interface) of Veralux is refreshingly clean compared to monsters like GeneralizedHyperbolicStretch (GHS). However, complex mathematics lie behind the few controls. Understanding each parameter is crucial for fine-tuning.¹

4.1 Processing Mode

Here you decide on the philosophical alignment of the stretch.

Mode	Detailed Description	Use Case
Ready-to-Use	Optimized for aesthetic astrophotography (“Pretty Pictures”). It applies not only the hyperbolic stretch but combines it with a Star-Safe Expansion and a coupled MTF (Midtone Transfer Function). A key feature is Soft-Clipping: Instead of hard-cutting values exceeding the white point (1.0), they are rolled off gently. This preserves color in the brightest star cores.	Standard for 95% of users. Ideal for Instagram, Astrobin, and prints.
Scientific	Performs a pure, mathematically precise transformation strictly clipped at 1.0 (Hard Clipping). No aesthetic smoothing takes place. The data relation remains mathematically exact, but this often leads to visually harsh, blown-out stars.	Photometric analysis, searching for extremely faint structures (IFN, tidal tails) where star aesthetics are irrelevant.

4.2 Sensor Calibration

As explained in Section 2.3, Veralux uses QE weights.

• Dropdown Menu: Select your camera sensor here (e.g., “Sony IMX571”).
• The Meaning of “Rec.709”: If your sensor is not listed, Veralux often offers Rec.709 as a fallback. This is the standard color space for HD television. It is “safe” but does not utilize the full potential of spectral weighting.
• Strategy for Unlisted Sensors: Research which chip is inside your camera. A Nikon Z6, for example, uses a sensor related to the Sony A7III (IMX410 variants). Choosing a “relative” from the list is often better than the generic Rec.709 standard, as the basic architecture (Back-Illuminated, Bayer Matrix filters) is often similar.

4.3 Target Background

• Default Value: 0.20.
• Interpretation: This determines where the median of the background lies after stretching. A value of 0.20 means the dark sky sits at 20% gray.
• Critique & Adjustment: Many users find 0.20 too bright.⁸ Modern processing often aims for values between 0.07 and 0.12. A value too high makes the image look “washed out” and amplifies noise and gradient remnants.
• Recommendation: Start with 0.12 to 0.15. This leaves room for later contrast enhancement via curves.

4.4 Fine Tuning & Color Grip

These controls were added in later versions (v1.2.2+) to give the user more control over the result.¹

• Protect b: Controls highlight protection (stars).

• High Value: More aggressive protection. Stars are kept extremely small and compact (“tighter stars”). Danger: Can lead to dark rings (ringing artifacts) around stars if set too high.
• Low Value: Stars bloat more naturally.
• Color Grip: Perhaps the most powerful slider for visual impact. It determines the mix between “Science” and “Aesthetics.”

• 1.00: Pure Vector Preservation (“Vivid”). Colors are maximally intense, but star cores can look hard.
• Reduced Value (< 1.00): Blends in portions of a traditional scalar stretch. This softens the harsh color transitions in star cores and mimics the “soft” look of ArcsinhStretch without fully adopting its downsides. Lower this value if your stars look too “digital” or oversaturated.

5. The Veralux Workflow in PixInsight

A tool is only as good as its application within the overall process. Veralux is not a “one-click wonder” that saves a bad raw image. It is a precision instrument that demands perfect preparation. Placement in the Processing Pipeline is critical, as Veralux must be applied to linear data.¹

5.1 Phase 1: Linear Preparation (Pre-Processing)

Before you even open Veralux, the image must be impeccably prepared. Errors in this phase are mercilessly amplified by the HyperMetric Stretch.

1. Master Integration & Crop:
   After stacking, use DynamicCrop to remove all stacking artifacts at the edges. Black borders would distort Veralux statistics (“Auto-Calculate Log D”).
2. Gradient Removal (Essential):
   Since Veralux lifts the background (Target Background), any gradient (light pollution, moonlight) would become massively visible.

• Tool: Use GraXpert (AI-based) or DynamicBackgroundExtraction (DBE) in PixInsight. The image must have a perfectly flat background.⁹

3. Deconvolution & Sharpening (Linear):
   This is where BlurXTerminator (BXT) comes in. The golden rule of astrophotography states: Deconvolution belongs at the beginning, while data is linear and the Point Spread Function (PSF) of the stars is still intact.10

• Order: Apply BXT before Veralux. Veralux then stretches the already corrected, sharp stars. If you stretched first and then sharpened, BXT would struggle to correctly model the now non-linear star profiles.

4. Color Calibration (The Most Critical Step):
   Since Veralux relies on “True Color” and sensor data, white balance must be correct before stretching.

• Tool: SpectrophotometricColorCalibration (SPCC) is mandatory here. SPCC uses the Gaia catalog to calibrate star colors based on their spectral class and the filters used.¹³
• Warning: Do not use simple “Color Calibration” without photometric reference if you want to use Veralux’s full potential. Incorrect color balance is “frozen” by Veralux’s vector preservation and is harder to correct later than with traditional methods.

5. Noise Reduction (Optional but Recommended):
   NoiseXTerminator (NXT) can be applied linearly (“Denoise linear”). A noise-free background allows Veralux to stretch contrast more aggressively without amplifying “grain.”14

5.2 Phase 2: The Application (The Stretch)

Now that the image is perfectly prepared (linear, flat, color-calibrated, sharpened), Veralux is deployed.

1. Open the script.
2. Select your chip under Sensor Calibration.
3. Set Target Background conservatively to 0.12 to 0.15. (The default 0.20 is often too bright).
4. Click ⚡ Auto-Calculate Log D. The script analyzes the histogram and calculates the optimal stretch factor.
5. Select Ready-to-Use.
6. Use the preview (if implemented) or apply the script.

The Starless Workflow Trick:

Many pros separate stars and nebulae before stretching to treat them differently.2

• Step A: Apply StarXTerminator to the linear image (check “Generate Star Image”). You now have a starless image (Starless) and a star image (Stars), both linear.
• Step B (Starless): Apply Veralux to the starless image. You can be aggressive here (“Scientific Mode” or high stretch factor) to show the faintest nebulae without worrying about stars.
• Step C (Stars): Apply Veralux to the star image. Use “Ready-to-Use” mode here and potentially a lower “Color Grip” value to keep star colors soft and cores small.
• Step D: Recombine both images in the non-linear phase using PixelMath (Formula: ~(~Starless * ~Stars) for Screen Blending).

5.3 Phase 3: Post-Processing (Non-Linear)

After Veralux, the image is “stretched.” It now looks like a normal photo.

• Fine-tuning: The image will often look somewhat flat because Veralux is optimized for physical correctness, not “drama.”
• CurvesTransformation (CT): Use CT to apply an S-curve: Darken the background slightly (drag anchor point at Input 0.15 to Output 0.10) and lift the midtones. This brings back visual “pop.”⁸
• Saturation: Thanks to Veralux, saturation should already be excellent. If necessary, use CurvesTransformation in “Saturation” mode (S) for minimal adjustments.

6. Comparative Analysis: Veralux HMS vs. GHS vs. Histogram

The ultimate question in forums is: “Why should I use Veralux if I have Generalized Hyperbolic Stretch (GHS)?”²

To make a well-founded decision, we must compare the methods directly.

6.1 Histogram Transformation (HT)

• Method: Simple curve lifting.
• Advantage: Fast, standard in PixInsight, easy to understand.
• Disadvantage: Massively destroys colors in highlights. Stars turn white, nebulae lose subtle color differences.
• Verdict: Obsolete. Suitable only for quick previews or technical mask creation.

6.2 Generalized Hyperbolic Stretch (GHS)

• Method: Mathematically highly complex hyperbolic functions. Offers control over the “Symmetry Point” (where to stretch), “Stretch Factor” (how strong), and “Protect Highlights.”
• Advantage: Absolute control. An experienced user can shape virtually any curve with GHS and emphasize specific brightness ranges in isolation. There is also a “Color” mode for saturation preservation.
• Disadvantage: Extreme complexity. It often requires dozens of iterations and trial-and-error to achieve a perfect result. The learning curve is steep. It is a “manual” tool.
• Verdict: The scalpel for the surgeon. Unbeatable in the hands of an expert, but time-consuming and error-prone for beginners.

6.3 Veralux HyperMetric Stretch (HMS)

• Method: Automated, sensor-based vector preservation.
• Advantage: “Color Safety” is built-in, not optional. It delivers reproducible, physically plausible results with a few clicks. The separation of chrominance and luminance is inherent in the design. It is extremely fast in the workflow.
• Disadvantage: Fewer manual intervention options in curve shape compared to GHS. You “trust” the algorithm.
• Verdict: The modern standard. It’s like a car with an automatic transmission and launch control—it delivers consistently high performance without the risk of stalling the engine.

Detailed Comparison (Table):

Feature	Generalized Hyperbolic Stretch (GHS)	Veralux HyperMetric Stretch (HMS)
Philosophy	Manual control over mathematical function.	Automated physical correctness (Sensor-Aware).
Color Handling	Must be manually controlled (Color Blending Modes).	Automatic vector decoupling (Design Principle).
Learning Curve	Very Steep (Weeks to months for mastery).	Flat (Minutes to hours).
Reproducibility	Low (dependent on many manual parameters).	High (based on sensor data and statistics).
Time Investment	High (Many iterations needed).	Low (“One-Click” potential).
Star Protection	Manual via “Highlight Protection”.	Automatic via “Star-Safe” algorithm.

Conclusion: Veralux does not fully replace GHS for special cases where you want to manually carve out a very specific contrast range (e.g., in the core of the Orion Nebula). But for the initial transformation from linear to non-linear image, it is superior to GHS in terms of speed, color fidelity, and simplicity.²

7. Troubleshooting: Practical Problems and Solutions

Specific problems identified from forum analysis (Cloudy Nights, Reddit) are frequently reported by users. Here are the solutions.

Case Study 1: “The background is way too bright!”

• Symptom: After applying Veralux, the sky is not dark gray but light gray or milky.⁸
• Cause: The default value Target Background = 0.20 is too high for many aesthetic tastes. Additionally, Veralux interprets noise or gradients as signal and lifts them to the target level.
• Solution:

1. Lower Target Background in the script to 0.12.
2. Check your Flats and gradient removal (GraXpert). A clean background is mandatory.
3. Use CurvesTransformation after Veralux to pull the black point slightly to the right (clip histogram on the left without losing data).

Case Study 2: “My colors look oversaturated / neon-bright”

• Symptom: Stars look like “candies,” nebulae have unnatural colors.
• Cause: This is often a sign that vector preservation (“Vivid”) is meeting poorly calibrated colors. If the image had a color cast, Veralux preserves and amplifies this cast extremely efficiently.
• Solution:

1. Perform SPCC strictly before stretching.
2. Reduce the Color Grip parameter in the script from 1.00 to 0.80 or lower. This blends in some classic desaturation and makes the image look more natural.¹

Case Study 3: “The script cannot be found after installation”

• Cause: PixInsight has not yet indexed the new scripts.
• Solution: Go to Script > Feature Scripts and click Regenerate. Also, check under Script > Utilities, as it is often sorted there if no dedicated category was created.⁵

8. Conclusion and Outlook

Veralux HyperMetric Stretch marks a level of maturity in astrophotography software. It ends the era where “Stretching” was a destructive process that had to be painstakingly corrected (boosting saturation, fixing stars). By taking sensor physics and vector mathematics seriously, it delivers results “Out of the Box” that are often better than hours of manual work.

The advantages of the concept summarized:

1. Integrity: The image retains its physical meaningfulness (color temperature of stars).
2. Efficiency: The workflow is massively accelerated.
3. Accessibility: It democratizes high-end results. Even beginners can achieve perfect results without deep understanding of hyperbolic curves.

For the PixInsight user, this means: The combination of GraXpert (Background), BlurXTerminator (Sharpness), SPCC (Color), and Veralux (Light) forms the modern “Gold Standard” workflow of 2025. While GHS continues to have its place for forensic detail work, the future belongs to sensor-based, intelligent algorithms like Veralux.

Referenzen

1. killerciao/VeraLuxPorting: Initial release of VeraLux for … – GitHub, Zugriff am Dezember 12, 2025, https://github.com/killerciao/VeraLuxPorting
2. New Veralux HyperMetric Stretch v1.2.0 – YouTube, Zugriff am Dezember 12, 2025, https://www.youtube.com/watch?v=Yx6R6gDnzH8
3. VeraLux HyperMetric Stretch – an open discussion – Forums – AstroBin, Zugriff am Dezember 12, 2025, https://ssr.app.astrobin.com/forum/topic/206981?page=1
4. PixInsight Scripts – Hidden Light Photography, Zugriff am Dezember 12, 2025, https://hiddenlight-photography.com/pixinsight-scripts
5. I made my first script for Pixinsight – Page 2 – Cloudy Nights, Zugriff am Dezember 12, 2025, https://www.cloudynights.com/forums/topic/953135-i-made-my-first-script-for-pixinsight/page/2/
6. Add Script – PixInsight Resources, Zugriff am Dezember 12, 2025, https://pixinsight.com.ar/en/video/pixinsight-add-script-2.html
7. siril-scripts – GitLab, Zugriff am Dezember 12, 2025, https://gitlab.com/free-astro/siril-scripts/-/tree/main
8. New stretching script in Siril: Veralux – Beginning Deep Sky Imaging …, Zugriff am Dezember 12, 2025, https://www.cloudynights.com/forums/topic/987003-new-stretching-script-in-siril-veralux/
9. When to use BlurXterminator and NoiseXterminator? – Galactic Hunter, Zugriff am Dezember 12, 2025, https://www.galactic-hunter.com/forum/technical-questions/when-to-use-blurxterminator-and-noisexterminator
10. At what step do the deconvulsion in the image processing – Cloudy Nights, Zugriff am Dezember 12, 2025, https://www.cloudynights.com/forums/topic/954549-at-what-step-do-the-deconvulsion-in-the-image-processing/
11. When to apply Deconvolution to an image – Stargazers Lounge, Zugriff am Dezember 12, 2025, https://stargazerslounge.com/topic/251816-when-to-apply-deconvolution-to-an-image/
12. BlurXTerminator – A Breakthrough in Deconvolution? – Cosgrove’s Cosmos, Zugriff am Dezember 12, 2025, https://cosgrovescosmos.com/tips-n-techniques/blurxtermintor-a-breakthrough-for-decon
13. L/RGB workflow with NoiseXTerminator and StarXTerminator – Experienced Deep Sky Imaging – Cloudy Nights, Zugriff am Dezember 12, 2025, https://www.cloudynights.com/forums/topic/868888-lrgb-workflow-with-noisexterminator-and-starxterminator/
14. How to Install BlurXTerminator, NoiseXTerminator & other RC Astro PixInsight Plugins!, Zugriff am Dezember 12, 2025, https://www.youtube.com/watch?v=1clZ7lH5WFA
15. Step-by-Step: Activating PixInsight’s Power Trio – BlurXterminator StarXterminator NoiseXterminator – YouTube, Zugriff am Dezember 12, 2025, https://www.youtube.com/watch?v=dVRih5p-5oA
16. Siril’s Generalized Hyperbolic Stretch: A Powerful Tool for Astrophotographers – YouTube, Zugriff am Dezember 12, 2025, https://www.youtube.com/watch?v=LCUjQCBPNcY
17. Generalized Hyperbolic Stretch – RawTherapee – discuss.pixls.us, Zugriff am Dezember 12, 2025, https://discuss.pixls.us/t/generalized-hyperbolic-stretch/45827