Welcome back to my studio development blog series.
In this instalment I would like to talk about the research I have been doing into furthering my texturing and material skills, the results of my character sculpt and my first look into using the Quixel 2 Suite.
Firstly I have begun to look into Physically Based Rendering (PBR) as that is the current standard in the industry. From my research I have gathered that with the advances in technology we can now do away with assumptions we made to cut corners for the sake of our technological limitation. Currently however now that we have more processing power at our fingertips we can more accurately model and reproduce the phenomena that help to give more “believability” to our art and industry.
I say believability as yes, it does allow for much more realistic materials and thus more realistic assets, sometimes animation needs to be larger than life and so we can make an object that behaves in a believable manner that still feels real within its world.
PBR is meant to be a more intuitive and physically accurate way to create materials. It mostly allows us to more accurately determine the parameters by which light interacts with our object. This is of utmost importance as light is the one central factor that determines how the rest of the world is seen.
I plan on investigating PBR further as this blog series continues but as of now here are a list of the resources I have been using.
With that said I have been working on my sculpt pipeline and have created a character model that I could try my first experience with the Quixel Suite on. I learned a lot from that process including ways I can streamline my sculpting pipeline and some ways to prepare my model for texturing.
So Above you can see on the left is the topology I achieved very quickly, I learned that you can simply retopologise within Autodesk’s Mudbox allowing me to take my base mesh sculpt it to my heart’s content and then just automatically retopologise it to specified parameters. The second picture shows the difference between the low and high poly models and the third shows what it was like with some basic materials assigned to it in Quixel Suite (including baked normal and ambient occlusion maps).
The process of creating and texturing this model taught me a lot, particularly once I took the mesh into Quixel. I learned very quickly that I need to be more efficient with my use of UVW space and that I need to find ways to reduce the number of seams. The placement and size distributions of my UV’s need to be more consistent. I like to give more “real-estate” to areas with more detail like faces however that results in not enough space for other areas leading to disasters like the texture on the bindings and pants.
I also learned that I need to learn how to consolidate my texture maps. Throughout the process of preparing the model for texturing I taught myself how to use multi-sub-object materials and multiple UV co-ordinates to bake out separate Normal and AO maps for separate parts of the mesh. However I then realized that I had no idea how to recombine them into a single map resulting in my having to redo the UV’s and bake out a 4K map.
I want to learn to consolidate my texture maps and to more efficiently break up my model for a more streamlined workflow.
Truth be told, I almost didn’t want to talk about this model as it most certainly didn’t turn out the way I wanted it to, it is actually a little saddening how it turned out, however it is important to learn from one’s failures as well as one’s successes.
While I am not happy about the product I have created I do think the project has been a most valuable tool in helping me to learn the limitations of my current skills and practices and more importantly the things I have yet to learn that will help me overcome and push past my current capabilities.
So on that note, there’s always more to be done.
Till next time…