Final Evaluation

Overall, I’m very happy with the way this project has unfolded. I feel that it had a good scope from the outset which I’ve met in its entirety, that it covered a broad range of skills and programmes and that it allowed me to learn a lot about the character art pipeline I chose to use in a relatively short amount of time. Of my initial goals set out in my pitch document, the only one that I feel I could have met to a better standard is that of the overall aesthetic direction, which I think could have been a little more refined. However, this is something that I would put down to time constraints and to the fact that this project was largely about skills acquisition for me personally, and so at some points I was worrying more about the technical aspects of the process than the aesthetic ones – my main goal was to produce something functioning and game-ready, which I feel I can say I’ve achieved. That said, I’m still very happy with the overall look of the two models, and I think I can now produce a much higher standard of work as a result of creating them.

The difference in quality, particularly in the area of topology for animation, is fairly significant between the two, and I’m confident that if I had produced a third model, it too would have ended up being a large step up and produced faster as a result of what I’ve learned during this project. I feel that I’ve taken a big step forward in being prepared for industry work and also in being able to pick up new skills as and when I need them, something which will serve me well moving forward. Part of me would like to have tackled something slightly more high-poly and complex, however I think that keeping my scope small and simple was ultimately the best decision, and allowed me time to make mistakes and learn from them without having to cut corners. This project has soundly taught me the value of working from the basics upwards, as I now have a foundation which I’m confident moving on to more complex projects with, and in conjunction with my BA3a project I now feel that I could tackle a higher level of realism and quality without feeling out of my depth or making fundamental errors.

In summary, I am very pleased with the outcome of this project, and while I might have liked it to be a little more polished in places, I value it immensely for the amount I’ve learned from it and for the skills I’ve gained.

Week Summary - BA3b Week 13 (Submission Week)

Low-Poly Model

I had to very carefully plan my time this week to make sure everything was going to be completed, after the setbacks I had last week. Thankfully, I found the rigging and animating process much faster than last time, having completed it once already. As before, I used a mix of tutorials depending on which sections were relevant to my model as well as asking tutor advice where appropriate, and with some minor issues (documented below) due to accidentally double skinning the model as well as some errors I took one and a half days to rig the model and another one to animate. 

I used the remaining half of the Wednesday to put together my showreel, which I created using Marmoset Toolbag and Adobe Premier. Below is an example reel of both my models in situ, followed by the final showreel that I will be submitting for the exhibition. 

Wax Model

This week we added the latissimus dorsi, pectoral muscles, and deltoids. I also made some changes to the serratus muscles to bring them in line with where they should be and cover up the last of the exposed ribs.

Unfortunately we didn't have time to finish the model in the time available, however extra sessions have been organised for after the submission date in order to complete it. I really want to have the model as complete as possible as I think it's going to be an incredibly valuable reference to have in the future. The actual process of making it has been extremely interesting and very informative, too, and I've learned a huge amount over the last several months.

Week Summary - BA3b Week 12

Low-Poly Model

I've started on the rigging now, I'm still running a little behind but again am confident that I have enough time to complete the project as intended.

I used locators to place the joints for the the character's arms as they're curved and so would have non-zero joint rotations if I tried to rig them the same way as I did on the previous character, so this helps to avoid awkward movement later on. The character's legs, I rigged by separating a duplicate from the rest of the model and placing the joints on a straightened version of the geometry before moving the bones in line with the original. They will swing outwards a little when moved, but as I'm unlikely to have time for a walk cycle and will probably be creating an idle to complement the animation for my first character this isn't likely to be an issue. The screenshot below uses a reverse footlock, however I removed this from the final rig because again, it's unlikely I'll be moving the character's feet a huge amount. 

At this point I started experiencing some issues with Maya, wherein it would experience fatal errors and crash with little to no warning. Nothing that I or my tutors did seemed able to fix the issue, nor could we figure out the cause - it seemed to be something to do with the model geometry, but the exact issue with it was unclear. Thankfully I was only making small changes to the actual model geometry, so it was fairly easy to work around, but it did set me back a couple of days while we tried to figure out the cause.

I used the time to clean up my texture seams, using Substance Painter moreso than Photoshop for its ability to paint nicely over seam lines. I could have used Substance for more texturing than just this part, however I'm more familiar with Photoshop and its layer structures and so it was much faster for me to create the bulk of the texture in there before adding the finishing touches in Substance. 

I also imported the model into UE4 just to get a feel for how it would look in-engine, although at this point I've decided to create my final submission turntables using Marmoset Toolbag. After some brief issues with the transparency map (below) I did manage to get it working as intended, and having it in-engine alerted me to the fact that my texture was overall rather dark, so I took the opportunity to brighten it up a little and give it a better read. 

In addition, I added some variation to the size of the fur planes to make things a little more organic-feeling. It's not an enormous change, but I think it makes a difference to the overall feeling of the model.

At this point, I was really getting into the finer details of the texture and really just addin more polish. Up until this point, I couldn't find a way of painting the fur on the backs of the hands that I was completely happy with, and I had the idea to use the fur planes instead. I think it looks much better and provides another nice little break in the silhouette in a similar way to the planes on the shoulders.

Given the lost time, I'll have to use next week to finish off the last of my rigging and animate the model, but I'm confident I can get this done. In the meantime, I've been updating my portfolio and working on my submission pack, so I'm still well on track.

Wax model

This week we added the neck muscles, abdominal muscles and rhomboids.

I found it really interesting how the obliques create the distinctive six-pack shape with the aponeurosis, I didn't realise that this was the cause and it's a strange fact to consider. 

I was also introduced to the website zygotebody, which is an online anatomical model and already an incredibly helpful resource. It's very easy to use and has all the muscles labelled on the body, as well as having the skeleton and the ability to peel away muscles layer by layer to see how everything fits together. I've already used it a few times for reference so far and I can see it being something I'll frequently be returning to in the future. 

Week Summary - BA3b Week 11

Low-Poly Model

I worked on actually painting my texture this week, after using last week to lay out. I put a lot of thought into how to approach translating the fur from my design into something appealing on a collection of flat planes, while still keeping the style in line with that of the previous character. I didn't want to hand-paint the fur completely, because that level of detail would be incongruous with the other character, but at the same time it needed some texture to it.

In the end, I settled on using the lasso tool to create block shapes which I could fill in with a soft gradient. A quick test revealed this method to have a really pleasing effect on the actual model, and it was then a matter of applying the same technique across the model, being mindful of texture seams and places where the image might end up stretched.

At this point, I asked for a few opinions on the design, and decided that the suggestion to lower my contrast was one that would be very helpful for me to follow, and I think it makes the design much easier to look at as well as focusing more of the viewer's attention on the face by keeping the highest contrasts in that area.

I did try to get a preview for the transparency map in Maya so I could get a better idea of how the final product would look, however I couldn't get it to work, and decided that it would be much quicker to put the model into Marmoset Toolbag or substance painter for my preview than continue to use Maya.

I would consider the textures at this point to be almost completely finished - I'm a little behind in not having my rig done but not to a point where I'm worried about not completing the project; I have the next week to rig and animate. I've had a few issues with Maya crashing that I'm hoping to get resolved, hopefully this won't pose too much of a problem.

Wax Model

This week started with adding more of the upper arm muscles and the erector spinae group. The shape here is really starting to fill out and look like a completed model.

In addition, we also added the abdominal muscles, including the external obliques. The internal layer of muscle wasn't added for the sake of simplicity.

I realised while doing this that my hips were very much the wrong shape, and needed flattened off and adjusted.This was mainly something that could be done just by heating and carving the wax, so it was fairly easily remedied. I've also carved out some sections that need fixing for next week, including areas where the glutes could do with being filled out a little more.