«The General Animation Seminar »Anitsukuri 2021 (あ に つ く 2021)It was held from September 18 to September 20, 2021. As one of the sessions, a seminar was held behind the scenes ““Zombieland Saga: Revenge“. They talked about the “peculiarities” of CG production, such as the live performance scenes that are a feature of this work. “Zombieland Saga” is an anime comic about girls resurrected as zombies and struggling to hide their identity as local ideals in Saga Prefecture.».
«Producer ‘Koutarou Tatsumi', played by Mamoru Miyano, is a very tense character, the story is strange and the members of ‘Franchouchou', who are doubted as ideals, are brave. At the “Anitsukuri 2021 (あ に つ く 2021),” a three-day comprehensive anime event beginning September 18, 2021, a seminar was held as one of the sessions to disagree behind the scenes of a show of this work. As in the other sessions, the production team discussed the techniques and attributes of 3DCG and digital production for animation fans and the next generation of creators.».
The purpose of the sequence is to “improve the expression with the same model as the previous work” and “raise the average quality of the live presentation sections”
«In the conference entitled “I work in 3DCG to capture the characters and expressions in Zombieland Saga: Revenge“, The scene show was first explained on stage. Work on the stage scenes (ie concert scenes) began approximately six months prior to broadcast. There are two dance performance patterns, one drawn and the other 3DCG. The 12th, last episode was the most expensive episode. The second half of the story, Part B, is mostly live performances, and creating an audience of around 30,000 was a daunting task.».
«Ai Kuroiwa, 3DCG animation director, spoke about how they approached this point: “Before we started the sequence, we had to review the workflow and how to organize the audience. The “Zombieland Saga: Revenge” objective had two purposes: to see how far we could go in terms of performance using the same models as in the previous project, and to use the knowledge of the previous project to average quality. raise the scenes on stage. To achieve this, we tried to show broken images as much as possible“».
«Let's look at some examples of the work being done on this project. The first step was to create the line drawings and 3D character models based on them. In order to use the 3D model of the previous work and improve the expression, the movement mechanism of the 3D model, known as “rig”, was changed in this work in many points. (Creates bone structure and digital controls within geometry).
«First, for the so-called “oscillating” parts, such as the hair, the number of sections was increased to be able to use the “SpringMagic” script. For the dance scene, where flexibility is required, the arms have been reinforced to reduce unnatural movement. The skirt is also programmed with a “Dash controller” for movement in conjunction with the movement of the legs, reducing the time to correct the movement of the skirt.».
«Now we will focus on the troubled episode 12. For Episode 12, they had to create the dance scene in 3D, so the background art in 3DBG was also created. For the 3DBG model used in episode 12, the modeling work began with a “blender” at an early stage in the background venture. After the “art board” was completed, 3D stage textures were created. A motion to give a character at the time of choreography and dance performances was included. However, the movements of the soul cannot be perfectly reproduced, so in the end the animators reclassify the recorded data».
«Once the 3D models, 3DBG, and motion capture are ready, it's time to start creating your screens. When the storyboard is ready, we use it as a basis for deciding which parts should be CG and which parts should be drawn, based on the schedule. We consider it to be the most effective method. Not only are captures used to create CG animations, but animators are also provided to guide the movements of the drawing.».
«Small models and virtual cameras were used at the “CG meeting”. The virtual camera (the movement of the screen within the three-dimensional phase) was used for discussions such as “How do we show the large living room?” The virtual camera was also used to communicate the director’s vision to the staff in a remote location. After the CG meeting, it's time for the CG animators to start their work. Layout, primary and secondary are the three stages of the process. The “layout” is primarily used to determine the location of the camera. In the primary phase facial expressions, movements, positions and timing are set, and in the secondary phase the image is completed».
«Character movements are created from motion capture. The motion capture is rough, so you have to change the angles of the arms to make them look nicer, and if the choreography covers the person, you have to adjust it to make it look better. In primary schools we take care of these things and make the set look good».
«Of course, we do not deny all gross motion from motion capture, and there are instances where small gestures are left intact and difficult to reproduce by hand. The most important thing for the production team was the facial expressions of the characters. Because of the smooth movements created by motion capture, facial expressions also had to be specified, otherwise they would look too CG. We had to pay attention to these features to highlight the charm of the characters».
«An important feature of the show was the “squint” pattern. In this post we have placed a narrow eye pattern close to the design. Because the shape of the closed eye is different from the designed narrow eye shape 50% of the time, without this pattern the average expression cannot be drawn well when the eyes are closed or the expression flapping while you are laughing».
«The grace of the stage at live shows is not just an ideal. Viewers are also throwing lanterns in their seats as part of the show. In the twelfth episode of this play, however, we plan to put in about 30,000 people, and it is not realistic to work on all of them. That is why we have created a more efficient and effective work environment, with greater automation and greater mobility throughout the process. Rather than creating all viewer models, we reduced the amount of data by preparing two types of models: one for close relatives and a simplified version for distant scenes. In a previous production, the attention colors had to be changed manually, but in this version we use a plugin that assigns the colors at random, so that the proportion of colors used can be adjusted to match the color of the character image. In the middle».
«So, a dynamic live scene of episode 12 was completed. The live scene is now available on the official Avex Pictures YouTube channel, so why not watch it again in light of the fact that you now know all the work that went into it?».
Fountain: Anime! Anime! | Zombieland Saga