Your Final Journey Postmortem
Post Mortem: Studio Duat Spring 2018
■ Introduction and overview of the project
○ Studio Duat
○ Platforms: CAVE and Oculus
○ Engine: Unity
○ Art: Maya, ZBrush, Substance Designer, Substance Painter
■ What went well (or what went right)
○ While our project was overscoped for the course of one semester (15 weeks), we made choices that streamlined our process and allowed each of the core teammates to focus on contributing their strengths.
■ Frequent stand-up meetings helped everyone know who was doing what, and prevented people from working on the same task at the same time.
■ Rapid prototyping at the start helped create a unified vision when provided with a weak preproduction package.
● When designing layouts based on target emotions, figuring out what elements to incorporate to evoke those emotions made the process fairly straightforward and easier to iterate.
■ Parallel workflows and development among each team member’s discipline to balance project focus helped four very different environments come together with a unified feel.
■ Even we don’t have any sound designer, VFX artist, nor concept artist in our team, we still tried our best to contribute as much as we can, from editing the dialogue, creating our own particle effects, and simple but effective concept to make sure all the team members share the same vision.
■ Outsourcing work to make up for skills that team members lacked was effective.
● Concept art
● Voice acting
■ Efficient communication between artists, and combined input at designing the environment,to create consistent environments from models,textures and lighting.
○ We approached technical challenges and artistic prototyping in similar ways in order to increase our iteration time and improve our workflow.
■ Creating models to scale (real world proportions) helped to make assets in the scene feel like they were the right size in scenes that used regular life size proportions.
■ Procedurally generated PBR material, makes the iteration of texture more efficient and effective, and applicable for many different surfaces.
■ Right choices in softwares used for creating our art assets, the combination of detail sculpture using Zbrush with the hard surface modeling in Maya, and terrain creation in unity.
■ Dare to learn and apply new skills and technology in the project, from using nano mesh, fabric mesh of Zbrush to model stalities and wheats, retopology of high-poly character in Maya, and PBR material creation using Substance Designer.
■ The back and forth of assets between artists to improve their quality and optimisation, including more efficient uv unwrapping, reducing of polycounts, and reducing of material numbers.
○ We got very lucky that our team worked consistently and productively together.
■ Members had good work ethic and passion for the project at hand.
■ Members had a good understanding of their own pipeline and the role it played in the project as a whole, which helped when it came to overall project efficiency.
■ All team members welcome iterations and non stop polishing of our assets
● We have over 5 versions of judgement chamber models, and over 10 versions for the stone floor material, over 5 different layout for the land of day and land of night, over 5 passes of lighting, and the multiple pasees of animations, improvement of camera path system and adjustment of volumetric fog.
■ What could have been better (or what went wrong)
○ Communication could have been better when considering team morale
■ Some decisions are often made without the consent of all team members - makes some team members feel like their input is not valuable or useful.
■ Lack of comfort in working environment - reduced efficiency and morale of some team members.
■ Lots of ideas thrown away without considering the possibilities or input of all members. Could have improved project or at least have been considered.
■ Credit not given where credit is due.
○ Different factors posed some members feeling like they needed to put in overtime and even sleep at the ETC - project exacerbated health issues.
○ Lack of breaks in scheduled core hours caused feelings of guilt when personal matters needed to be taken care of or if a break was needed.
○ Don’t overscope.
■ Lessons learned and conclusion
○ Technical Lessons
■ Avoid creating terrain using Zbrush, which creates undesirable and difficult to change uv map, and limits the resolution in UV, try using the terrain creation tool in engine, then export to Maya, and apply tileable textures to ensure higher resolution
■ For the sake of optimization, keep the material count for each model under 3
■ It is a challenge for creating open environment in VR, due to the scale of assets is based on the actual human size, every asset needs to be scale up many times, textures are blown up many times as well, resulted in the poor resolution of the textures. We couldn’t find the best solution to this problem, due to the limited resolution in VR, but procedural modeling, and tilleble can ease the resolution problem to a large extent.
■ When modeling and creating materials for open environment that is 10 times or more of the character size, artists need to picture the size of models and materials in engine, and adjust the details accordingly. For example, the amount of grout between bricks may looks alright for pavement, but when the material is actually for huge stones in a temple, the amount of grout should be significantly smaller. The chips and cracks on building in Maya may looks suitable, but when it is scaled up in engine, the cracks look much more visible then they should be.
○ Team Management Lessons
■ All members should study or learn at least some part of each subset in the entire pipeline. This is super important as it helps them understand the time constraint and challenges that come with each task.
● This will improve understanding between one another and most likely make everyone feel more comfortable in the end.
■ Reaffirmed belief that when everyone is driven to a goal and passionate about a subject matter, that great things can happen.
■ It is important to have timely and effective communication between artists and programmers, to let artists understand the right way of using the tools created by programmers to help the production, and also let both the artists and programmers have better understanding of each other's’ work.