Virtual Heritage Project Best Practices

Introduction

This guide outlines the best practices for pursuing a virtual heritage 3D modeling project built by student teams with faculty supervision. After a brief explanation of virtual heritage and the types of digital resources 3D modeling can facilitate, it provides tips and resources for organizing the research, modeling, and interactive design that goes into creating a virtual heritage project.

What is virtual heritage?

Virtual heritage is a growing field that uses both image- and text-based historical evidence to recreate historical scenes, cityscapes, and/or objects in a virtual space. A finished virtual heritage project allows a user to immerse themselves in a historical time and place that is otherwise inaccessible. Technologies like 3D modeling, sound design, game engines, and virtual reality can enable this immersion, and that technology is itself enabled by rigorous use of historical evidence and interpretation. Virtual heritage projects can be small scale, like our semester-long Crafting Yokohama project that recreated the architecture and atmosphere of two streets in 19th century Yokohama, or large scale years-long projects like Virtual Angkor, capable of sophisticated simulations of what a day in the life of a city of thousands was like over 700 years ago. What scale of project you can successfully achieve is determined by your technological resources, the time limit of the project, and the size of your team.

Virtual Heritage Project Workflow

There are three basic types of work involved in a virtual heritage project: research, modeling, and interactive design. Research is the work that goes into finding, gathering and interpreting historical evidence. Modeling is the creation of 3D models, and interactive design incorporates those models into an interactive experience using a game engine, such as Unity or the Unreal engine. There can be some overlap across any of those categories, but generally speaking research must be undertaken first to provide parameters for the modelers, and the models must be finished before they can be imported into a game engine. These different types of work require different types of technical and methodological skills. While it’s a good idea for a project leader to be familiar with the basics of each type of work, it’s advisable to recruit team members as specialists who can concentrate on developing skills in these specific areas.

The size of your team will likely depend on the ambitions of your project. Unlike some digital humanities methodologies, virtual heritage is very difficult to complete as an individual researcher. The sheer scale of creation involved in modeling means that projects are often best served by a team structure. If using the JapanLab model of a single faculty member overseeing a team of students, a reasonable configuration for a first time or smaller scale project might be 1-2 researchers, 3-4 modelers, and 2-3 interactive design specialists.

Virtual Heritage Best Practices

This list of best practices is organized along the same general workflow as a digital heritage project. It is written with the assumption that an individual faculty member or project leader will organize a team of students in the role of specialists, but this information should still be useful for projects with different configurations.

Team Organization and Logistics

1.     Know who is making what decisions

a.     Decide early in the project who will be making decisions about scope, visual style, level of detail, narrative, historical interpretation, etc. It is completely reasonable to have student teams lead in the decision making, but best establish early on how those decisions will be made cohesively.

b.     Make sure you have a schedule that includes deadlines for when bigger decisions about scale, scope, and visual style need to be locked-in, and allow some buffer for
complications, should they arise.

2.     Create a file-sharing system

a.     A virtual heritage project involves many files, few of which are automatically shareable. Set up a file sharing system early that makes it clear which files are the authoritative versions, and make sure that all team members have adequate access.

b.     You can use Google Drive, Github (2 GB maximum per project), Box, or whatever shared drive system your institution uses.

3.     Give team members defined roles

a.     Assign student team members roles as researchers, modelers, sound designers, or game engine specialists. While there can be shared responsibilities, distinctive roles allow students to take the lead in their specializations with confidence and proactively develop their skills in those areas.

b.     It is also a good idea to give basic training in all the methodologies being used early in the project, so that the researchers understand the work of modeling or interactive design, and the more tech-focused specialists understand the historical perspective of the project.

4.     Create a regular schedule for full and partial team meetings

a.     Researching, modeling, and designing can be isolating and individualistic work. The best way to ensure a cohesive project and a collaborative team is to set up regular meetings to make sure smaller specialists groups can give updates, provide feedback, and make decisions together.

Research

1.     Create a database of visual and written sources

a.     Before the project begins, create a database of any historical sources that might be useful for the research and modeling teams. It can be added to throughout the project, but putting together this database will also help the project leader judge how feasible a virtual heritage project on any given topic is.

b.     Seek out multiple forms of visualizations and description, including maps, photographs, paintings, illustrations, tax records, written descriptions, etc.

c.      The database can be a sophisticated interactive SQL database or a simple list of citations and links in a shared excel spreadsheet.

2.     Make a map of visual sources

a.     Using Google My Maps or another mapping program of your choice, make a map of the visual sources you’ve gathered for the database. This can be an early task for the research team and will help the team get a better sense of the space you want them to model.

b.     The map does not have to be beautiful; its main purpose is to serve as a visualization of historical evidence and geographic placement.

3.     Develop a guiding philosophy around historical evidence and interpretation

a.     Historical evidence will always leave gaps that 3D models will have to fill in. Only three sides of a building may be pictured in photographs, but all four must be modeled. As a team, determine what your priorities are for interpreting evidence. What are the most important aspects of this historical time and place that you want to emphasize? What are reasonable inferences to make with the evidence that you have, and what are not? What level of fidelity or detail does your project need to have to be successful?

Modeling

1.     Determine the foundational visual style of the project

a.     Modeling is an art form and requires artistic choices. To have a cohesive visual look, it is important that your modelers know whether they should be making make low-poly (fewer polygons, less detail) models, high-poly (more polygons, more detail) ones, or if the project requires realistic texturing, a specific color palate, etc. The more photo-realistic you try to make your models, the more time and resources will need to be invested in those models. Additionally, you may want to determine the angle of view for your modelling scenes, such as whether they are to be seen from street level or an elevated view; each has advantages, but modelling both options will likely double the time you need to devote to the project.

2.     Make a list of modeling priorities

a.     This is a task that is best done as a full team. Have your research team report to the rest of the team about what they’re finding in the visual and textual evidence. Using that evidence, make a list of models that the modeling team will have to make to populate your desired scene. Mundane, everyday object that feature often and repeatedly are often the most worthwhile. This can include buildings, architectural features, smaller objects, plants, vehicles, streets, gutters, and even animals and people, depending on the experience of your modelers. This list will be helpful in determining priorities for the modeling team and keeping track of progress that’s being made. It will also keep the amount of repeated or redundant work to a minimum.

3.     Reuse models

a.     It’s a good idea to keep an eye out for models that can be reused instead of recreated every time. For example, wooden sliding screens are a common feature of Japanese architecture. Instead of building a new screen for every building, modelers can make one screen, put it in the shared file directory, and use that screen file in whatever model they need.

4.     Keep an eye out for complicated models you can purchase

a.     There are large marketplaces of 3D models that may be relevant to your project and affordable to purchase (see the resources section for specific sites). When it comes to more complicated models like vegetation, animals, vehicles, etc., it may be worth exploring what has already been made. Many of these models cost less than $20 and can save significant time on a project.

5.     Plan lots of time for modeling

a.     Modeling will take up the most time of any other part of the project, so plan accordingly. The beginning of this process is a steep curve, as the modelers are essentially creating something from nothing. However, as the number of reusable models and the modelers’ familiarity with the subject matter grows, modeling speed will increase.

6.     Seek out online resources and tutorials

a.     There is no right or wrong way to model, but there are a lot of different approaches you can take while modeling. Luckily, there are a lot of (often free) resources available to help your modelers pick up new skills or different techniques of modeling. Youtube is a great resource for finding tutorials and process videos, as is Reddit.

b.     In addition to these online resources, plan time for your modeling team to work with and consult each other—it’s useful for problem solving and sharing any methodologies that team members develop on their own.

Interactive Design (Game engine)

7.     Practice with the engine and build prototypes while waiting for the models

a.     Since the importation of 3D models into the game engine is the last stage, the interactive design specialists may feel like they are spinning their wheels waiting for their part of the project to start. However, they can make good use of this time by building prototypes with placeholder models, becoming familiar with the game engine, testing atmospheric sounds, assessing whether or not the overall goals of the project are executable, and exploring postprocessing effects that will benefit the final project. Make sure to keep them involved in regular meetings so they can see how things are developing.

8.     Seek out online resources and tutorials

a.     The needs of a virtual heritage project are often somewhat different than the video games or interactive design projects that your team members may already have training in or experience building. There are a lot of resources out there to familiarize your interactive design specialists with the game engine you’re using and bridge any gaps in their previous knowledge. Again, Youtube and Reddit are useful places to start.

9.     Have a minimum viable product and modular goals

a.     A minimum viable product is the smallest version of the project that could be considered functional. It has none of the bells and whistles, but accomplishes the primary goals of the project. Make sure your whole team, but especially your game engine specialists, understand what the minimum viable product is and how to achieve it.

b.     Any goals beyond the minimum viable product should be modular. Perhaps it is adding non-moving figures to a city street, creating a 3D soundscape, applying post-processing effects to the model, VR compatibility, introducing a new kind of building model, etc. These should be contained so that some could be implemented or, as is likely, others can be cut without jeopardizing the minimum viable product.

10.  Incorporate sound to breathe life into a virtual space

a.     Sound design is an incredibly effective aspect of immersion for virtual heritage projects. A neighborhood that sounds of clanking streetcars and crowds will create a very different atmosphere than one that sounds of crowing chickens and running water. It can also add life to projects where adding animated human figures is prohibitive.

Technologies and Resources for Virtual Heritage Projects

3-D Modeling Programs

Autodesk Maya: The industry-standard 3-D modeling software. Students and faculty can access it for free using an educational license, and it must be renewed every year. It is a resource-intensive program and requires equipment with a dedicated graphics card.

Blender: An open-source 3-D Modeling software that is an alternative to Maya and a sophisticated program in its own right. It is free to use and has a wide array of plugins developed for an even wider array of functions. It is less resource intensive than Maya but can also be less stable and less regularly updated.  

Game Engines

Unity: An industry-standard game engine that requires relatively little coding to operate at a basic level (The language that it uses is C#). There is also a huge asset store that can give you access to user-created 3D models, game design features, and more, sometimes for free. Unity has a wealth of official educational resources and many additional tutorials. The engine is free to use unless your organization has made more than $100,000 in revenue from Unity projects over the last 12 months.

Unreal: The Unreal engine is increasingly popular, and like Unity can be operated with relatively little coding (the coding language is C++). There is also an asset marketplace that is not quite as robust as Unity’s, but it is growing. There are also not quite as many educational resources as Unity, but that is quickly changing as it becomes more popular as a development engine. A 5% charge on royalties only applies if the project makes more than $1 million.

Resources for buying/licensing 3D models

https://www.turbosquid.com/

https://www.cgtrader.com/3d-models

https://www.artstation.com/marketplace/game-dev/resources/3d-models

Unity Asset Store (can be proprietary for Unity game engine)

Unreal Engine Marketplace (can be proprietary for Unreal game engine)

Sound mixing:

https://www.audacityteam.org/

Sourcing free sounds:

https://freesound.org/