This example shows the correlation of the trends in the temporal outlier graph to the white speckle of a damaged videotape in the panel. For instance, this image shows an existing QCTools graph, called TOUT (or temporal outlier), which gives an impression of the glitchiness of the frames and below the graph is a panelled image comprised of taking only the single column of pixels from each frame and showing them aligned on the same timeline. This allows a way to provide a visual impression of the content against the traditional QCTools graphs. The Panels are a series of visual images that are presented in the graph view of QCTools, where the width of the image, from left to right, represents the timeline of the image. These panels are added in QCTools version 1.2. In our development work, we created several types of Panels for QCTools, which are reviewed below. Whereas QCTools would typically present a numeric value per frame over a timeline to create a line graph, this approach would create a panelled single image that depicts a particular quality of the full timeline. In panels, we would use ffmpeg to create an image per frame that was a single-pixel wide and then stack these along a timeline. From this concept, we collaborated with the developers to create a concept that we called “Panels”. We foresaw adapting the timeline structure but filling it with visual information rather than simple numbers. While the graphs of QCTools provide information about the trends and events occuring on a timeline of audiovisual data, the simplicity of a plotted line graph has many limitations. We also requested some new options to how QCTools saves reports to help us scale our approach As QCTools is an open source application, the contribution of development here aids not only our work, but contributes to the preservation field as a whole. By defining quality control issues that we were diagnosing by hand, we collaborated with QCTools developers to plan out new features, such as present visualizations against a timeline in addition to graphs. With QCTools we found a suitable tool for this task, but identified several areas where expanded capabilities could aid in the efficiency or accuracy of our quality control work. That is why we were looking for a way to optimize and support the manual QC. This makes quality control very time-consuming, especially in larger-scale digitization projects with few resources on the QC side. ![]() Many things can only be checked in real time, i.e. An example for a digitization error, on the other hand, is playing and subsequently recording the film at the wrong frames per second speed. An example are sprocket holes from 16 mm film which – depending on decisions made for digitization – may be visible in the digital representation as a conscious decision. In addition manual quality control is conducted, which includes tasks such as checking the correct naming of the files, correct content or visual detection of analogue or digital artifacts.Įach analogue film format has its own specific characteristics which will subsequently be present in the digital representation as well, but must be distinguished from digitization artifacts. ![]() ![]() Our QC includes automated processes such as checksum checks and validation of the files against our policy. To give a brief overview of the workflow: every digital file we received from the external digitization service is subject to quality control (QC). While digitization is carried out by external vendors, we had to develop a workflow starting with the selection of the analogue film copy and ending with the ingest of the digital files into the digital archive. Digitization was started in 2017, since then approximately 2,500 film titles have been digitized another 700 film titles are planned. TIB’s film collection mainly consists of 16mm film with optical and magnetic sound as well as Digital Betacam. ![]() This includes overseeing digitization of the collection as well as the digital preservation of the objects created as a result. The Digital Preservation and Conservation teams of the TIB – Leibniz Information Centre for Science and Technology, Hannover, Germany are responsible for developing strategies for the conservation of physical film copies. A collaborative blogpost by Dave Rice (MediaArea), Merle Friedrich and Miriam Reiche (Technische Informationsbibliothek (TIB)) Introduction
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |