Getting Data...
Traditionally the word "Pipeline" is used to define:
“Software developments that automate within the tools an artist uses to remove manual or routine work without artistic value”
Automate the routine tasks that the artist does, so that they can carry out their creative work more effectively. A key example of this type of task is the choice of file names and the selection of folders where the generated material will be placed.
#Maya, #Nuke, #Fusion or #Houdini are some of the Software examples in which Pipeline automation is often used. For their organization and that of their material, these programs require a Naming Convention as well as a folder tree that will govern the dependencies of the files that call each other, likewise it is also necessary to control their modifications and dependencies.
To carry out this task, the #TDs (Technical Directors), using an API (application programming interface), program calls between systems to interconnect the tools. There are several types of APIs. Some based on a programming language, such as #Python or JavaScript, and on other occasions as simple calls, using REst or #GraphQL services.
Inside 3D Design tools.
To give an example; Let's choose the Layout task. It is the first task of the Production functional area and where the scene and cameras that make up the shot are defined. We will choose Maya as the tool for the example.
The layout artist opening Maya opens a connector tool with the production control platform, either PipelinePro or #Shotgrid. With your credentials or an encrypted token, you make the call to the api and it returns all the information so that the panel of the connector tool is fed.
If the task is starting, proceed to the construction of the scene, or Build. The platform sends you the list of content to be imported and the automation sets up the scene with all the necessary dependencies and settings. It is also recorded on disk so that it is ready to start your work. In many cases this is done automatically.
This collection of the assets is determined in the dependency tree defined for that specific project. It is the responsibility of the CG Supervisor to trace the dependencies and the formats that are going to be used in the files to optimize the performance and quality of the project. As an example, it can be determined that the model task draws two geometries, one high and the other low, each one will be used for different things and it will be determined that for layout the rig task will be imported that hooks the low one to be able to overtake her while modeling.
In PipelinePro there is a dependency diagramming app so that both the CG supervisor and the programmers can be clear about the process and the tools that must be programmed to keep everything connected.
The process and actions of the TDs code continue with the work in Maya, the checks before saving and the options to send for review or publish, but we'll talk about that in another post.
