Scalable Black Level Management Recommendations Guide

1. Understanding the Black Level Problem

The first thing to understand about black level management is this: the ideal solution is optical, not digital. The problem is rooted in physics. Projectors cannot produce true black, they emit residual light even when displaying a fully black image. In a multi-projector display, this residual light stacks in overlap regions, creating visible brightness bands and seams that are most distracting in dark content and dark environments. Even with a perfectly blended digital signal, the light overlap is noticeable to the human eye and can be distracting if it’s not considered ahead of time.

Optical blend plates - filters mounted at the projector lens, physically remove this excess light before it reaches the screen. A correctly blended display with optical correction requires no software compensation and delivers true black levels under all content and ambient lighting conditions. An optical solution makes no trade-offs for image contrast or the projectors dynamic range.

1.1 Why software correction has limits

All software black level correction operates on the same fundamental constraint: it cannot make any area of the image darker. The projector is already at its minimum light output when displaying black. The only available operation is to raise the brightness of darker regions upward to match lighter ones.

This has two practical consequences. First, the corrected black level across the entire display will be at least as bright as the brightest uncorrected overlap region — heavier overlap geometries produce a brighter effective black floor. Second, the uplift applied to achieve uniformity directly reduces the contrast ratio visible to the viewer.

Understanding this constraint is important when evaluating system performance: a well-corrected display may still appear "washed out" compared to a display with optical blending, particularly in cinema or simulation environments where dark scene fidelity is critical.

1.2 The “halo” effect

A second issue affects virtually all DLP projectors and must be addressed separately from the basic uplift. The Halo Effect describes light that leaks beyond the edge of the Digital Micromirror Device (DMD) - the physical boundary of the image-forming element. This light falls outside the registered projector boundary and cannot be controlled by the basic black level uplift, which only operates within the registered image area.

In practice, halo correction is frequently required in real-world deployments. The width of the halo varies by projector model but is typically several pixels at the screen edge. Without halo correction, even a well-tuned basic uplift will leave visible and mismatched bright stripes at projector boundaries.

Scalable's software provides a dedicated halo correction tooling, described in Section 3. The beta automatic calibration feature (Section 4) addresses halos automatically as part of its camera-based measurement process.

2. Basic Black Level Correction

Basic black level correction is available as part of the Color & Brightness licensing package. It is accessed via the Black Level panel in the Scalable Display Manager interface.

The panel identifies regions on the display where intensity changes as content transitions between projectors — the overlap zones where residual light accumulates. For each region, an uplift value raises the brightness of the darker projector(s) until the region matches neighboring areas.

2.1 Basic correction procedure

2.2 Advanced correction tab

The Advanced tab provides finer control over corrections at intensities slightly above pure black which is useful for scenes that are dark grey rather than fully black, where the basic correction may appear inconsistent.

To use it, raise the Image Intensity slider from 0 (pure black) and observe the correction quality at a representative dark grey level, around 25 is a useful starting point. The Scale sliders adjust the correction independently for regions covered by 1, 2, 3, or 4 projectors.

3. Halo Correction

For most projector deployments, basic black level uplift alone is insufficient. The halo effect creates bright edges at projector seams that the basic correction cannot reach. Halo correction is a second-tier adjustment that targets this specific artifact.

Halo correction is not enabled by default. It is activated via Advanced Options, after which a dedicated Halo Correction tab appears in the Black Level panel.

3.1 Enabling halo correction

Open the Advanced Options dialog via one of these methods:

• Scalable drop-down (top left) > Support > Advanced Options
• Keyboard shortcut: Ctrl + O

Add the following option:

Click Finish and restart the software. The Halo Correction tab will appear at the top of the Black Level panel.

3.2 Halo correction procedure

3.3 Troubleshooting: verifying changes took effect

If halo corrections appear to have no visible effect on the projected image, verify the changes were applied by comparing the calibration output file directly:

1. Navigate to: C:\Program Files\SystemData\DefaultSystem\LastCalibration
2. Locate ScalableBeta.bmp and create a copy of it.
3. In the software, navigate to the Black Level > Halo Correction panel.
4. Move the Size or Uplift slider to a large value.
5. Click Update Calibration.
6. Open the updated ScalableBeta.bmp and compare it to the copy. A visible change in the file confirms the tool is working. If the file is unchanged, verify the advanced option is set correctly and the software was restarted.

4. Beta: Automatic Camera-Based Black Level Calibration

4.1 What it does

Automatic black level calibration uses calibration cameras to measure actual projected light levels across the screen and compute a per-pixel uplift that achieves uniform black levels across all projectors. It replaces the manual region-selection and halo-correction workflow with a single calibration run.

The algorithm works by capturing images of 100% black on all projectors, calculating a target intensity based on the brightest measured area, and computing the per-pixel adjustment needed to make every part of every projector's output match that target. The process runs iteratively across all cameras until convergence.

Key capabilities:

• Works with any number of projector overlaps
• Automatically corrects projector halos — no separate halo correction step required
• Corrects for curvature, keystone, and hotspot-based intensity variations
• Corrects for color (RGB) bias across projectors

Key limitations:

• Increases the effective black floor, heavier overlaps produce a brighter result. Contrast is still reduced.
• Very sensitive to stray light. The room must be as dark as possible during calibration.
• High-gain or rear-projection screens with viewing-angle-dependent hotspots may not look optimal from positions far from the camera eyepoint.
• Does not work well with systems using ROI (region of interest) files - automatic edge mask detection can fail.

4.2 Camera and environment setup

Before running automatic calibration, the following preparation is required:

• Minimize all ambient light. Any stray light that reaches the screen will be incorporated into the correction as if it were projector output, producing artifacts in the calibration.
• Set camera exposure significantly higher than normal, bright enough to clearly distinguish between a black and a dark grey projection. Note that this exposure setting may be too bright for standard geometry calibrations on some systems.
• Focus cameras well. Sharp camera images improve boundary matching at black level region edges.
• Ensure cameras are positioned near the viewer eyepoint. The algorithm corrects for what the cameras see if cameras are at a different angle from viewers, viewing-angle-dependent screen effects may degrade results.

4.3 Enabling automatic calibration

Run geometry calibration first, then enable the feature via Advanced Options:

This adds an Automatic Calibration tab to the Black Level panel. The feature does not replace the existing Basic and Advanced tabs — it operates alongside them.

For Scalable Desktop configurations, also set:

4.4 Running calibration

4.5 Supported output formats

Automatic black level calibration results are applied to the following output formats:

• Display Client test images
• SDK: OpenGL, DX12, Vulkan, DX11 (requires 12.0 SDK DLL)
• MPCDI
• NVIDIA warp (Scalable Desktop and SDM Remote NVIDIA warp)

4.6 Settings reference

The following settings are available in the Automatic Calibration tab. Defaults are suitable for most systems — only adjust if you have a specific reason.

4.7 Supported cameras

Validated camera models include Basler GigE and Canon. Multiple cameras are supported simultaneously - the algorithm combines measurements using standard camera weighting. Recommended not mixed camera types.

5. Choosing the Right Approach

The three correction methods can be used independently or in combination. The table below summarizes when each applies.