Blacklevel Compensation
Last updated
Last updated
When using multiple overlapping projectors, residual light is inevitably emitted even when a completely black image or no signal is being displayed. This phenomenon is known as the black level. Additionally, most projectors emit light in regions outside their actual image area—known as the margin—which cannot be directly controlled by the projector.
The complexity of black level compensation arises due to several factors:
The brightness of the margin correlates with the brightness of the projected image.
Margins are subject to the same perspective distortions as the image area.
Projectors behave non-linearly at low brightness levels, and this behavior varies by manufacturer and model.
Overlapping projector margins accumulate light, leading to noticeable bright spots.
Cameras used for calibration have limited dynamic range, making it hard to capture true black images.
Human vision is highly sensitive to small differences in brightness, especially in dark conditions.
As a result, black level compensation is achieved by slightly increasing the overall projection brightness to homogenize the overlapping regions. This always reduces contrast and makes true black impossible to reproduce.
Goal: Apply the minimum possible increase in brightness to visually mask overlapping areas, balancing accuracy and perceived image quality.
Before embarking on the actual implementation of a black level uplift, it is important to be fully aware of the prerequisites and limiting factors. We will therefore now take a closer look. Only by understanding what can be compensated for and to what extent can the tools for black level uplift be used sensibly.
A projector cannot compensate for its own margin—it can only be adjusted by overlapping neighboring projectors.
Projector pixels are the finest spatial unit and cannot be partially controlled. See above diagram: compensating is limited by the largest pixel size (gray line)
Spatial resolution varies across the projection surface due to geometry and projector placement.
Overlapping projectors usually differ in pixel orientation and resolution, leading to only approximate edge matching.
For example, in a setup using 10 UHD projectors calibrated with a 5 MP camera, each camera pixel may cover up to ~26 projector pixels, limiting the effective calibration resolution.
The limited number of black level brightness steps (e.g., 20), nonlinear light behavior, and perceptual sensitivity restrict the use of smoothing or dithering techniques.
Conclusion: The effective accuracy of black level compensation is determined by the spatial resolution of projectors and their measurement quality in relation to their neighbors.
Even if projector edges are approximated with pixel accuracy, visible differences may still occur—especially at short viewing distances.
There are four approaches in VIOSO 6 available. For a quick test, you can start with texture based blacklevel compensation. As soon as you are looking at production-grade results, go for Region based blending compensation.
Manual blacklevel region generation is also available, but recommended only for setups with low complexity.
This compensation of the black level is based on existing blending textures that are used for blending the projectors. It's the most straight forward approach, but is limited by the resolution of the camera that is used to scan the projectors.
Advantages:
No dedicated calculation required
No explicit margin handling needed
Supports automatic recalibration
Simple slider-based configuration (yellow slider)
Disadvantages:
Resulting black level is relatively bright (~20–30% projector brightness).
Significantly reduces contrast for actual content.
Activation: After blending is completed via the Color & Blending button.
Option: Enable blended blacklevel.
Adjustments:
Set all region brightness sliders to 100% (blue rectangle).
Use dark/bright value maintain to adjust contrast (green arrow)
Use blacklevel slider to adjust the effect of the blacklevel uplift (yellow arrow)
This method requires a dedicated calculation to create overlap regions with higher accuracy than the texture-based approach. The regions are calculated in the camera space, which reduces the computation time but provides less accurate regions.
Advantages:
Lower black level brightness (~8% projector brightness).
Fine-tuning possible based on human perception.
Disadvantages:
Limited accuracy in edge detection due to camera space processing.
Requires consistent camera and projector orientation for margin customization.
Automatic recalibration depends on prior conversion task.
Preprocess: Activate during or after camera-based calibration. A standard size is used for initial calibration with regard to the size of the margin.
Perform conversion task from the menu Calibration > Conversion Tasks > recalculate blacklevel: This will perform a conversion that creates the blacklevel regions based on the camera image.
Margin Settings: adjust left/right/top/bottom (blue rectangle) to whatever fits to your projector setup
Option: use all in one blacklevel method controls how the brightness of the margin can be adjusted
Enabled: Margin is treated like any other overlapping region. In setups with a high degree of overlap, overlapping margin regions may potentially be too bright, as they cannot be distinguished from normal overlap regions
Disabled: Separate compensation for margins and overlaps.
Perform the conversion by clicking Perform
Adjustments
Within the Color & Blending dialog the impact of the blacklevel uplift can be adjusted:
Important: Disable blended blacklevel (red arrow)
Adjustments:
Set all region brightness sliders to 100% (blue rectangle).
Use dark/bright value maintain to adjust contrast (green arrow)
Use blacklevel slider to adjust the effect of the blacklevel uplift (yellow arrow)
Ensure
blended blacklevelis disabled.
This method requires a dedicated calculation to create overlap regions with higher accuracy than the texture-based approach. The regions are calculated in the projector space, which resembles the highest accuracy for the blacklevel compensation regions.
Advantages:
Higher accuracy due to projector space computation.
Independent margin configuration regardless of orientation.
Lower black level brightness (~8%).
Viewer-adjustable compensation tailored to perception.
Disadvantages:
Requires preprocessing.
Takes long time to compute.
Preprocess: Activate during or after camera-based calibration. A standard size is used for initial calibration with regard to the size of the margin.
Perform conversion task from the menu Calibration > Conversion Tasks > Projector Space blacklevel:
Adjustments: see above
With this method, the overlapping regions can be adjusted or created manually, otherwise this method corresponds to "Region based".
Advantages:
Full manual control of overlap zones.
Based on accurate projector space data.
Disadvantages:
Requires manual setup and understanding of projection geometry.
How to use:
To avoid having to start from scratch, deactivate “auto blacklevel compensation” option in Figure 3 and run the conversion task Projector Space overlap:
Adjustments
Within the Adjust projector dialog each projector can be adjusted via the blacklevel button
Each overlap region can be adjusted, new overlaps can be created and their shape adjusted with mouse and keyboard interaction:
