In-Line Spectrophotometer Banding Correction via FFT Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current In-Line-Spectrophotometer (ILS) systems in digital document reproduction devices are susceptible to banding defects, which affect the accuracy of color patch measurements and calibration, leading to undesirable artifacts like halftone banding due to varying photoconductive drum velocity and mechanical motion errors.
Innovation Solution
A method that involves analyzing the ILS data stream using Fast Fourier Transform to identify peak frequencies associated with structured noise components, determining a common banding frequency, and adjusting the distance between color patch repeats to synchronize with the banding wavelength, thereby reducing banding noise in measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the ILS measures color patches in the presence of banding defects, then the system can perform color calibration, but the measurement accuracy is degraded due to periodic density variations in the printed image
Solution Approach 1:
The patent identifies the banding frequency through FFT analysis and deliberately sets the patch repeat distance to be a non-integer multiple of the banding wavelength. This converts the harmful banding effect into a predictable periodic pattern that can be mathematically characterized and compensated for during calibration, transforming the measurement error source into a correctable parameter.
Solution Approach 2:
The system uses FFT analysis of ILS measurements to detect the banding frequency, then feeds this information back into the patch target design by adjusting the patch repeat distance. This closed-loop approach continuously adapts the measurement strategy to the actual banding conditions, improving measurement accuracy despite the presence of banding defects.
2Ease of operation
If the patch repeat distance is set to an integer multiple of the banding wavelength, then the measurements are easier to synchronize, but the banding noise reinforces and measurement accuracy deteriorates
Solution Approach 1:
The patent deliberately introduces asymmetry by setting the patch repeat distance to a non-integer multiple of the banding wavelength rather than a simple integer multiple. This asymmetric relationship prevents the banding noise from constructively interfering with all measurements, and the asymmetry is precisely calculated to distribute the error across different phases of the banding cycle.
Solution Approach 2:
The system dynamically adjusts the patch repeat distance parameter based on the detected banding frequency. Instead of using a fixed standard distance, the parameter is changed to optimize measurement conditions for the specific banding characteristics of the printing system, balancing synchronization ease with noise reduction.
3Measurement precision
If multiple patches are measured to average out noise, then random measurement errors are reduced, but the structured banding noise persists and limits improvement
Solution Approach 1:
The patent segments the measurement process by taking multiple individual patch measurements at different positions and phases relative to the banding pattern. By distributing measurements across different banding phases and combining them through averaging, the structured noise components cancel out while the true color values reinforce, improving both precision and repeatability.
Data Source
AI summary
What is disclosed is a novel system and method for detecting and correcting for In-Line-Spectrophotometer (ILS) measurements of constant value patches in the presence of banding in multi-function document reproduction systems. The present system analyzes the ILS data stream to identify structured noise components due to banding. An FFT is performed on each L*a*b* component in the ILS stream for a single test page. The peak frequencies from the FFT of the L* a* and b* channels are compared. Common frequencies in all 3 channels indicate a banding component. Once the banding frequencies and the banding wavelength are known, the color patch target can be adjusted to ensure the color patches are synchronized to the banding wavelength. By running a series of synchronized patches and averaging results, structured noise can be eliminated. In such a manner, a reduction of banding effects on color calibration can be effectuated.