Having recently rekindled my strive for human modelling and animation again just for practice, it was quite a struggle to get back into best practice for optimised models. So I had researched and found a few things that could also be of help to anyone starting out in 3D again.
Looking back at my previous models (keeping in mind that the last time I modelled human beings was back in my uni years), they were modelled to just get a semi basic shape, and not be flexible at all. So I started following a few tips from the old source that is the internet, and seeing if there are any basic guidelines to make a much nicer looking figure, with minimal polygon count. This is VR after all - it’s good to keep things simple.
I’ve played my fair share of VR games, so I kind of know what to expect in terms of quality. You’re not going to get new Final Fantasy cinematic styles or anything to that detail, but it won’t be as blocky as Final Fantasy 7 either.
I had decided to start with a more realistic human proportion, so I decided to start looking for real life photos. Meanwhile the body and even hands were quite straight forward, the head/face is really quite heavily intensive characteristic of the human - essentially giving it’s identity. Not only that, but the consideration of future proofing this in case it needed to speak (rather than just have a photoshopped mouth) was also something I had to think about.
I had found a few links that helped me get my modelling up to scratch again. Check them out below:
Here’s also a good reference guide to a low poly head topology.
The head was the longest part to do, as it was about sculpting the fine lines that would eventually make the muscle features more prominent and less flat. The trick was to not get too carried away and add unnecessary polygons to smooth it and add to the poly count. Overall - the head alone was about 640 polygons that I managed to make it towards, so I think we shall see what that will come out to in the final render. Stay tuned!