Project Brief:
After the horrifying fire of Kyoto Animation, July 2019, I started to conduct research on the storage of anime images. During my search, I was very surprised to learn that many Japanese animation industries are still lacking a cloud storage system. In this project, I designed and developed a relational database on MySQL Workbench. This project accents of MySQL language skill and mindset of relationship clarifying, which hopes to become the background knowledge in future research & design on database related projects.
Introduction:
Commonly, animations consist of rapid succession of sequential images called frames. GENGA is referred to the original images or the copies of the drawing of important scenes made by senior animators, and is used as a model by other animators to draw other DOUGA frames. DOUGAs are the usual images that bridge between GENGAs to able frames become anime cuts. The concept of these images is similar to animation cel in the western animation industry. A 24-minute anime episode usually contains more than 6,000 frames. To manage these images, either make sure they could hit the deadline and supervised, revised by senior animators and for future archive uses, I’d like to design a database to track and manage these images. The database is designed for anime producers and the assistant anime producers, who are responsible to make sure the anime can be successfully produced and broadcast on cable TV or online streaming platform on time while producing animes. They have to assign and track the progress of thousands of GENGAs and DOUGAs, make sure the frames are qualified and are being corrected, colored and consistent. They might also have to archive the frames by cuts in case of future using, like story review or product usage. To successfully get access to needed images, metadata of each works, episodes, cuts, and images have to be systemically stored. Besides title and number of cuts, the date of production, broadcast, and the date of related DVD/Blu-ray releases, descriptive scene title and the subtitles of the cut would also help the user of the database find their targets. For each image, importance level, credits of the contributors, version history, and the need for image correcting could help identify the issues which should be eyed on and assist tracking macroscopic working progress of the animations. The producers take benefit of the concrete database system rather than remember every single part of the works. The database should be able to solve the chaos during the production of animes and reduce the possibilities of late submissions. The contributors can also track their own progress by sharing limited access to the database and schedule their working plans more efficiently. For the database, information about images contributors, and the timeline of each part of the works should be stored correctly. Sequential images from cuts. The descriptive tags and lines which could illustrate the usage and content of each cut are also needed. The cuts are managed by zero to more episodes since there might be deleted cuts and re-used cuts. Cuts which share the same storyline should use consistent tags to keep relationship with each other. If the animators received specific instructions from the directors, it should also be stored by the system. An anime episode contains many cuts, it belonged to an anime series, published by one or more anime companies, directed by one or more directors. The required information can be attained during production. Once an anime finishes completely, the database can change the status into “closed” which indicates the information became archives.
According to my plan, the main user group of this database would be the product manager of anime films.
Insert data into the database systematically
Using the database to create plans
Access of archives so that they can also help directors or anyone who need frames in groups of anime, episode, cuts, or any specific condition.
Retrieve the information related to frames easily: including information of contributors, deadlines, locations, and the stages they are currently in so that they can manage the schedules of works.
Challenges:
This image describes the conceptual relationship I tried to put in my database design in the early stage. With efforts and advice from my professor, I revised it and turn it into an ERD logical data model, then input the data frame and test data in MySQL workbench to develop a database. Compared to the first proposal of this project, the design of the database is more focused on the relationship between the frames. With the recognition of the importance of contributors, I separate 3 different levels of these roles. Which are: officers that manage the schedule of anime working processes, the crucial roles who participate in the whole serial (such as, character designs, scene designs, etc), and contributors who work on every single frame to compose animations. While the anime is being made, the officer who is in charge of the anime can keep an eye on the workflow to make sure everything working on the track. After the production, the database provides information about derivative products and physical location for those frames worth preservation.
After iterated the design of the database and implant test data in MySQL Workbench, I used the reverse engineer to create this ERD graph. In addition, to create the structure, I also design some SQL queries to support data inputting and auto-updating. For example, the producer can generate multiple empty DOUGA frames based on existed GENKA frame with trigger DOUGAGENERATOR.
After the DOUGA frames being generated, users can create records of assignments for tracking and notification by simply update the information on the DOUGA frames.
Similar working flow on creating moving records of the original paper-based frames. If a cut of frames being sent from 1 studio to another studio, the log would automatically be created by the trigger.
With the data in the database, the producer -- or any users who need to check on the status of the works, would be able to get a list of incomplete works, unassigned works, and staff with specific skill or skill level.
