Reply To: COLLAB2021 – The Spice Must Flow (TSMF) Blog
Spice Baron Interviews
[Note: I have redesigned this forum thread by eliminating the individual team project status sections. They were placeholders to give status reports. As we are finishing up this project, I will provide an aggregate summary below. These placeholders will remain as future articles can be placed there so that they are in the beginning of the forum and not dispersed throughout. This should make them easier to find.]
I had the privilege to interview several of the Spice Barons during the past few weeks. I asked them the same questions, and documented their responses. Here is a summary of these interviews.
What is the most special thing on your team that you are excited about?
The overwhelming response was: “the people.” The participants of this project were very engaged and helpful to each other. Many brought reference images to study. There were collaborations to problem solve issues. People pushed each other to make better assets. Especially those where more than one person was assigned a specific asset.
One perspective shared was that the first time he had viewed the Spice Vendor’s House that he thought it would be a piece of cake to build 3D models. Once the project started, the perception changed. The project was much bigger and it presented more challenges than originally expected. The conclusion was that this would be a good learning experience for all of us.
Several were excited about the creek and how the water will interact with the watermill in animation. The challenge being: how to get the water to fall off of the paddles of the watermill as the wheel is turning. Also, how will we get the water to show ripples, especially as it flows into an obstacle like a rock?
What are some of the challenges or obstacles that your team have faced in Blender?
Responses were clearly divided between team leaders that had normal 3D modeling and those responsible in leading teams to create Nature assets.
On the 3D modeling side, there was often apprehension coming into a week where, for some, a new technique was introduced. Confidence through practice had to be built when tackling each phase of the project. The types of issues were:
- What is the best way to streamline the making of normals?
- For sculpting, what brushes should be used and what are the techniques in using them?
- How do you bake a normal? For that matter, some beginners at first didn’t understand what a normal is.
- There was some confusion in what the high poly mesh was going to do to the lo-poly model.
Certainly, these types of challenges are part of the learning process. With a combination of Kent’s streamed instruction, assistance with the team, and perseverance, these issues were overcome – the proof can be found in the finished assets.
The Nature groups had different challenges. At the top of the list was: how to keep the poly count down? It was estimated that there is over 400 square meters of area that needs to be filled with mountains, trees, bushes, ivy, grass, flowers, rocks and a creek. The size of that space is not trivial, and challenges the efficiency of Eevee.
Let’s take a tree for example. It has a trunk, multiple branches and many leaves. The number of vertices in each leaf gets multiplied many times over as they are placed throughout the branches. The poly count adds up quickly. So the challenge is to find the right balance between making a stylistic tree with a low poly count.
During the first few weeks, a lot of time was spent with trial-and-error experimentation to find the right balance between technique and minimizing vertices. Aesthetics was also the focus of testing to find the best way to texture the assets. Hand-painted textures belong to this style, especially for foreground assets that will have the scrutiny of a closer camera. Background assets can get away with procedural textures.
Clouds also presented a challenge in this project. Volumetrics are simply too slow. Primarily because the physics refreshes on every movement of the camera. Again, a lot of experimentation was required to land on the best method.
A general challenge that all teams faced was meeting the objective of stylized art. Simplifying the shapes, lines, features and color down from realism. It is a balance between too much and too little detail. Once this hurdle was overcome, the results truly look fantastic.
What modeling or technique do you think needs to be highlighted within your team?
In general, the responses to this were brief and aligned with the types of assets the team had responsibility. Each team and individuals found the best methods to make wood, tile, stone / rock, trunks / branches, leaves, mountains, etc. Certainly, not without a lot of hard work with experimentation. In all cases, Spice Barons praised their team’s accomplishments. A few quotes include: “knocked it out of the park.” “They succeeded in keeping the poly count low.” And, “taking shape in glorious ways.”
One area that deserves a little discussion is Geometry Nodes. This is a new feature in Blender that offers a significant amount of power and versatility. Especially for a project like this. Several people in the Nature group used geometry nodes; some for the first time. The reason for their use is that there are some things that you cannot do with a Physics Particle System. Let’s use the randomization of leaves as an example. In a Particle System you can only randomize the scale of a leaf. Using Geometry nodes, you can randomize the leaf’s length, width and rotation.
A novel technique was used for background trees. The tree image is placed on a plane. Using Geometry nodes, the plane rotates to be perpendicular to the camera. So, the tree always faces the camera, regardless of the camera’s position. This technique eliminates the need for a 3D tree filled with branches and leaves in the background, which would significantly add to the project’s poly count. Simply brilliant.
Are there any basic tips that you have that would benefit a beginner to Blender?
Universally, the prominent response is to get a solid foundational understanding of Blender. Start with the fundamentals and don’t try to do too much at once. Divide the learning into smaller chunks and learn step-by-step. And, don’t expect to create perfect artwork quickly. It will take time, practice and persistence.
CG Cookie offers in-depth courses on the fundamentals of modeling, sculpting, shading and texturing. If it works for you, go through those courses and take detailed notes that you can quickly refer back to. It is important to know where to find information.
If you encounter a problem in Blender, try to research the solution on your own. A lot of what we do is problem solving; the more self-sufficient a Blender user is, the better in the long run. HOWEVER, before you get too frustrated – ask a question on the forum. There are many people in this community that sincerely want to help. Take advantage of it. The important thing is to hang in there and don’t give up. It is a journey to learn 3D art.
Give feedback (when requested) and also be willing to openly receive critiques. A beginner may feel that they have nothing to offer in providing critical feedback due to their inexperience. Providing feedback forces the evaluation of the art. Perhaps such exercises will result in the beginner to grow faster. Being involved like this will also help in learning the jargon (or technical language) of Blender. When it comes time for you to request feedback, perhaps people would be more receptive to giving you a critique. And we all need a second opinion to grow.
Joining a collaborative project is also encouraged. Following a tutorial is great to build what the author is teaching. In a collaboration, you are challenged by building your own unique asset. Pick an asset to build that is not too hard but will stretch you. Problems will pop up, which is always a great learning experience as you go through the process of learning how to solve them. There is also the benefit of working with others, where you can learn from each other.
What’s different about a team project compared to an individual project?
The consensus is that being on a team you can tackle larger projects that may not have easily been accomplished as an individual.
There are several benefits to a team:
- Camaraderie – Collaborating with others and focused on a common goal. You feel like you are a part of something bigger. In fact, a part of the community.
- Support – Whether you are struggling or simply looking for a second opinion, there is feedback and encouragement in a team environment. You are not alone.
- Growth – Team interaction is often motivating and encouraging. There is the opportunity to learn from others. There is also accountability, which pushes you to do your best work.
An advantage of an individual project is consistency. You should not be able to see that assets in the scene were made by different people, which is more likely in a team setting. Working on a project alone allows for uniformity in the modeling, shading and texturing of assets.
What are the challenges of having a remote team project? What has it taught us?
There are two main challenges:
- Response Times – This project has participants that are literally living in all corners of the world. And hence, they are in different time zones. Responses to inquiries can range from immediate to take a day or two.
- Text-based communication – You can lose context with an all-text communication platform. Sometimes it is hard to describe complex issues in a brief written statement. For some members, English is a second language. While this may not be debilitating, it should be recognized that some things may be lost in translation.
Of course, it is acknowledged that our collaboration is based upon a volunteer structure. But, in a corporate environment, project meetings are conducted either face-to-face or via video conferencing. If a project member has a question, it is easy to pick up a phone to get an answer. Responses are immediate and less likely to be misinterpreted – even working on different continents. Stronger personal connections among team members are also more likely to be developed.
On COLLAB2021, Discord was setup as a free VoIP, instant messaging and digital distribution platform. Early tests proved that Discord did not perform very well with video conferencing. But it has proven effective in collaborating through chat and audio, where computer screens can be shared. This platform definitely helped in the collaboration, improving communication in real-time and sharing of ideas. But the issue that participants are still in multiple time zones has not changed. And, not everyone on the project utilized this platform.
[Editor’s Note: My personal observation is that it was Kent, Jonathan and the Spice Barons that set the tone of this project. At every step they were engaging and encouraging. They made this an enjoyable and educational experience. The results are truly fantastic!]
End Spice Baron Interviews
