CAGD 373 - Game Asset Production
I am the lead of Group 2, Team 3, and we replicated a game scene from the game Control. I believe that we were fairly successful in trying to do so and we achieved the kind of scene that we set out to make at the beginning of this project.
The work that I did over the course of this project was primarily on prop modeling. As you can see from the below images, I modeled 22 props in total.
Within the scene itself, the main feeling that we were trying to communicate was that of an abandoned, eerie and cold government facility in which there were no other signs of people being present there anymore. I wanted to be able to create a feeling of megalophobia through the central inverted pyramid and the high ceilings. The blue tone of the lighting and post-processing was pretty effective in setting the tone that we were trying to communicate. In addition, it was somewhat unintentional, but due to the small size of the FPSController, the scene felt very large and contributed to that feeling of a large scale building.
When it came to leading the team, I encountered many problems and I learned a lot more about being a lead. One of the problems that I quickly learned from at the end of this project was that my verification of models needed to be more in-depth. Our scene was very modular by nature of it being a regulated government building, but something that caused a lot of time loss was that the models created by my scene modeler did not always fit together. This was an issue of verification because I should have either 1. Got the Maya scene from him and checked whether the models fit together when we were early in production, or 2. Made him assemble the scene himself as he progressed so that he could quickly fix any issues with models that he encountered.
Another issue I encountered was the distribution of work could have been optimized. Although our set piece ended up being the large inverted pyramid in the center of our scene, I would have appreciated a high-poly model to go alongside it within the center of the scene. I assigned a high-poly model to my texturer to do while he waited for UV sets, but I assigned it far too late into the process (4th sprint) which made it difficult to focus on and we ended up scrapping it in favor of focusing on more pivotal work.
In the future, I plan on trying to consider the work habits of my groupmates more in order to optimize how much I can motivate them to do. The number one thing that I think would have made the progression of our project much smoother was understanding what would motivate my groupmates to do more. Chase needed little motivation as he was very self-motivated to try and make a mesh destruction system. Mo would have appreciated a single long-term high poly model very early in the process. Ryan would have been very effective if I assigned him 7-8 models each sprint rather than 5, because he would always complete the work that he was assigned each sprint but not ask for more.
Blog Post 5 (12/10/2024)
I am the lead designer on Group 2, which is working on attempting to recreate a scene from the 2019 game “Control”. This sprint, I focused on finishing all of the prop models that I had remaining in my backlog, as well as cleaning up any possible n-gons and creating UV sets so that my texturer, Mo, could do the final work on texturing our models. This sprint was far and away our most successful and productive sprint and I am very proud of the work that we were able to accomplish as a team. In the next couple of days, we will be able to put together the entirety of our scene and hopefully get it looking as close as possible to the original game scene.
The work that I personally did was, as mentioned, centered around finishing the rest of my work that I had in Maya. A large amount of time was allocated towards fixing the n-gons that were here and there on certain models. Something that wasted a large amount of time when I was fixing n-gons was that I ran into this weird bug that would delete certain faces when I multi-cut edges into faces in order to make them four-sided. Sometimes cutting more edges into the rest of the n-gons fixed the faces, and sometimes it would not respond whatsoever, forcing me to get creative with my solutions by either loading previous versions of the model or deleting the faces myself and using the “Fill Hole” tool.
Regardless of whether or not these missing faces were cooperating with me, the one thing that remained constant was that if I encountered that bug on any of the models, the UVs were destroyed beyond repair. I had to re-UV many of the models due to this issue. My workflow usually consists of modeling, UVing, and then fixing n-gons on a model, but this issue has single handedly caused me to reorganize my modeling workflow so that n-gons are one of my top priorities.
.png)
The models that I created were a recycling bin and trash can, a chip bag, an office chair, a node for the battery model to go into, and a simple fern using planes. I learned a lot during modeling the office chair and battery node. For the battery node model, I used the “Boolean” tool which allowed me to subtract models from each other in order to create gaps for the batteries to sit in. It took a very long time to cut in the proper edges so that the model consisted of quads, but once I did that, the model came out pretty well and I was satisfied with the result. The office chair was a fun model to make because it was my final model that looks somewhat complex but was actually on the easier side. The rest of the models were kept simple as their real world counterparts are far from complex.
Blog Post 4 (11/19/24)
I am the lead designer of Group 2, and we are recreating a scene from Control, a game that came out in 2019. Across the course of this sprint, I continued work on modeling the props of the scene, which were a much harder endeavor than I initially believed. However, overall, the models are coming together very well and the scene will be nicely populated with props as a result.
The models that I worked on over this sprint were an emergency light model, a large microphone model, a simple round planter model, a battery model and a table model. The most challenging of these models was by far the emergency light model. As previously mentioned in the previous blog post, the tripod that I created for the flash model was meant to be reused for two of the models that I completed this sprint. However, even with a large section of the model completed ahead of time, the remainder of the model took an extremely long time to model because of the intricacy in the details on the top section of the model. I created a gridiron to wrap around a portion of the model, and it took me far longer than I anticipated because of the number of extrusions and bevels that I had to do.
The next model I completed was the other model which utilized the tripod base, which was the microphone model. This model turned out to be much easier than the emergency light model, thankfully. The topper of the model was fairly similar to the flash model, but needed resizing and some additional extrusions and pieces.
After I completed those two models, I moved on to modeling a table so that the office rooms that are affixed to the main lobby within our game scene would actually look like offices. This model was very straightforward and consisted of many cylinders and cubes which did not need a lot of extrusions or beveling. The next sprint, I will need to model an office chair in order to complement the table.
I made a very quick round planter to decorate some of the hallways, and this needs little elaboration as it was a simple cylinder with some extrusions. The very last model that I made was the battery model. I will be creating a battery node model next sprint for the batteries to sit in. These two models are what is going to be used in order to implement the battery throwing system which my programmer, Chase, had programmed in the last sprint.
Blog Post 3 (11/12/24)
I am the lead for Group 2, which is trying to recreate the main hub scene from the third person action-story game called Control. This sprint, as is consistent with the previous sprints, was a sprint in which I primarily focused on completing props for the game scene so that the scene would have a lot of detail which the player could look closely at.
This sprint went well for me in terms of production. Although the prop models which I need to do are gradually picking up in complexity, I was able to complete five models and UV’d each of the models so that they are ready to be put into layouts and be textured once my texturer, Mo, is finished with the set modeling.
The models that I completed this sprint were a fire alarm sprinkler model, a security camera model, a toolbox model, a wall-mounted defibrillator model, and a tripod mounted flash/lighting model. The fire alarm sprinkler model was a model that I needed to do in order to give context within the scene for having the fire alarm lever which I modeled in the last sprint.
The security camera model was fairly straightforward to model, but very frustrating to UV due to the weird cuts that I had to make for the “hood” that the security camera has. I ran into a similar issue with the toolbox model which has overhangs on each side, but it was still simpler to UV because of the overall rectangular prism shape.
The wall-mounted defibrillator model was honestly very fun to make because it has a lot of simple pieces which come together to form the full model. I modeled the door of the model as a separate mesh so that when we build the scene, we can have the option to open the door or leave it closed.
The last model that I completed was a tripod mounted flash model. This model definitely took up the most of my time due to the fact that the tripod base was a type of model I was not familiar with making. However, it is important that I got this modeled early on, because there are a handful of other models which use the same exact tripod model in the game. Moving forward, I will be able to reuse the tripod model and change the top piece in order to create the other models from the game.
Blog Post 2 (11/06/24)
I am the lead for Group 2 in CAGD 373. Our group is trying to recreate a scene from Control, a third-person action story game that came out in 2019. Over the course of this sprint, we primarily put our heads down and focused on getting a lot of modeling done for this sprint so that we could start having our texturer, Mo, focus on texturing in the next sprint while we continue building a backlog of models for him to texture.
The work went pretty smoothly during this sprint, without any major complaints or issues. The models that I worked on this sprint were a mug, a fire alarm lever, a whiteboard, and a directory sign model. As a prop modeler, I wanted to focus on larger props that help create character within the scene. That is why the main models that I did for this sprint and the previous sprint were larger and give us the ability to customize what is being displayed. When the scene is being put together, we will be able to put what we want of our choosing onto the whiteboards or TV screen.
Something that I noticed from the references that we gathered for our game scene was that while the scene is a moderate size and does require a decent number of models, the models are all intentionally fairly mundane, meaning that the props that we do choose to highlight by modeling them need to make sure that they help to embody the idea of the scene being a normal government office with some oddly brutalist-futuristic architecture choices.
Lastly, besides the models and the behind the scenes card writing as a lead, I also created a scene blockout in our Unity project. Truth be told, this is probably something that I should have done in the previous sprint, but the idea had not occurred to me at all until I observed other groups doing the same thing. Moving forward, this blockout will make the rest of the scene construction much easier.
Blog Post 1 (10/30/24)
I am the lead for Group 2, Team 3. The game that we will be trying to replicate a scene from is Control. Overall, this sprint, I spent a lot of time focused on the planning phase of the future sprints. However, I do believe that I could have done more modeling and will push myself further in the coming weeks.
What I worked on over this sprint was a vending machine model and a TV model. For the tv model, it was a fairly straightforward process. When modeling it, the only hiccup which I encountered was that I needed to resize and scale the TVs dimensions appropriately. When we place the TVs into the scene in Unity, we plan on having videos idly running on the TVs, so making sure that the TVs fit the aspect ratio of 16x9 was important to ensure that we would be able to include that addition later on.
The vending machine posed a more of a challenge, because the vending machines within Control have a distinct interior with snacks and chips modeled. I modeled an interior to the vending machine, but decided that it would not be a productive use of my time to model the detailed interior when I could use some sort of decal projection or texturing instead.
For both of these models, I also completed UVs. However, I did not move into texturing with either of the models. The main reason why was because we wanted to have our texturer, Mo, do all of the texturing so that the textures are united. The other reason was because I have not figured out a distinct way to organize the UV layouts yet. For this project, our allowance of UV maps is 12 2k maps, or 48 1k maps. What will likely happen is that we will all focus on modeling (with the exception of our programmer), and then once we get to a stage where it is appropriate for us to texture, our texturer will work on that while the set modeler and I work on building out the scene in Unity.
Comments
Post a Comment