Quality Assurance and Quality Control for Digital Products Created on the Zeutschel Camera Systems
NARA is implementing a structured QC protocol for preservation master digital objects produced on the Zeutschel camera systems. Below is a outline of our current processes. This workflow is implemented on a project-by-project basis and is not yet consistently implemented for all preservation master products although it is our eventual goal to do so. We continue to adjust our benchmarks and workflows as needed.
Scanning Technician scans the Universal Test Target (UTT) twice daily, once at the beginning of the shift and again at the end of the shift. Using the Golden Thread software which evaluates the UTT Target, the Scanning Technician receives feedback on the Zeutschel camera performance. Specifically, NARA is looking to determine values on the following objective performance measures. See Glossary of Terms for definitions.
- Tone response curve
- Delta E (or color encoding error)
- Noise (ISO 15739)
NARA is refining its aimpoints for these objective performance measures to align itself with the FADGI Star Rating system in which equipment must not fall below the 2 Star minimum. UTT Target values for equipment reading at 2 Stars or below in these performance areas generate error messages; UTT Target values for equipment reading at 3 Stars and below generate warning messages. The exact aim point values are still in the testing phase. Once these values are confirmed, they will be added to this site.
If the UTT Target values are within NARA's predefined range for the project goals, image capture proceeds. If the values are outside the predefined range, corrective action is initiated. Image capture only occurs on equipment that is verified to be performing within target ranges.
Embedded metadata: NARA is developing embedded metadata guidelines for reformatted still images.
Technical metadata: NARA is developing a schema for reformatted still images based on its reVTMD schema for video technical metadata.
NARA is developing an in house job tracking order system to better coordinate work at the Branch level.
Item Level Evaluation: Manual Visual Analysis
NARA uses a custom tool to determine an subset of files to visually inspect. This subset is determined by an algorithm that determines a subset of files based on the total number of files to be inspected as a group and the predetermined Acceptable Quality Level (AQL). NARA currently uses the Zero-Based Acceptable Sampling Plan from MIL-STD-105E. MIL-STD-105E has been withdrawn but the sampling method is still valid. The sample set is compared to the original source object to ensure order fulfillment correctness. The following are manually compared against the original document: correct dimensions, spatial resolution, orientation proportion/distortion, image skew, cropping, and scale reference.
Item Level Evaluation: Automated File Characteristics Analysis
NARA uses its File Analyzer tool to assure uniformity in file name structure, to assure the correct file pairs are created