Learning Report

Assessment Criteria

1. Generating & Selecting Ideas: Character research and choice of final designs

2. Investigation, enquiry & visualisation: Character sketches, model sheets, reference drawings

3. Process &/or concept development: Industry process and ability to learn software

4. Research & Analysis of information, contextualization & critical thinking: Reflective blog & learning report

Planning / Designing

Whilst planning, or researching for the design of my character, I got really hung up on someone saying to take inspiration from real life people. So I did a lot of searching up actors that I really liked or models that were pretty and stylish. I dare say I can be very creative but when it is in a situation like this, I feel I can easily get the wrong idea or at the very least get hung up on something unnecessary. So instead of thinking of making a background and "building up" a character from my own imagination - I just got glued to the idea of making a "Good looking woman" for my sculpture/3D model. It took a couple of in-person sessions of feedback before it finally clicked for me that I needed to get away from this idea, as it was getting me nowhere by just thinking of making a "pretty" character.

So then I basically started thinking of things that I really liked, but not something very obvious like games, or something similar. The idea of taking inspiration from foods that I enjoyed came to mind, and creating sketches involving things like buns, weird hair styles that resembled soft serve ice cream, clothing that took inspiration from candy and things like that. It very quickly came to me that the "person" I was creating was very much energetic, happy, go-lucky and positive character and the clothing/look of the character I felt resembled that as well. This also made me realise that this is something I very much like in characters from other series/films/games as well.

For generating the ideas and doing the research I could have expanded a bit more than what I did, exploring other types of sweets or foods or even exploring a different theme overall. I was very quick and decisive when it came to choosing the kind of inspiration that I wanted to reflect in my character (hair, clothing) so I only made a couple of sketches and had already decided by then what I wanted to go with. I did explore some different colour variations but I quickly ended up selecting what I felt looked more pleasing to me and what I was happy with. The methods I used were also not as expansive as it could have been. Focusing a lot on body shape I could have taken this into silhouette drawings where I explored different poses and even more shapes, as well as after having done the clothing design - made sure the silhouette was easy to make out what the character looked like and made it very clean. Something to focus better on in future character designs.

What did I learn:

Learned a completely new program: ZBrush.

After learning most of the basics of the program, I realise what it can accomplish. And compared to other programs like Maya, it personally felt a lot easier to use. In Maya are starting with different shaped polygons and use mouse clicks and drags, and other actions like "Bevel", to accomplish what you want in your model and/or object that you are creating. In ZBrush however, you can either choose to start from an already made skeleton/model that has the basic shapes, or you can start from a cube/other shaped object and draw on top of that shape to progress your creation. There are also tools like the "Move Polygons" tool where you can drag and shape the object very roughly to get the base shape you need to then build on top of that.

To me these programs feels so completely different to each other as I, myself, feel like I have more control of what is going on during the modeling progress in ZBrush. If I say, in Maya, I started with a cube, beveled it or smooth it out I depend on the program itself to show me what it would look like in the end. I am sure with more Maya experience you most likely will end up knowing what the different outcomes with certain actions at what time will look like. But ZBrush and painting makes me immediately see the progress as I keep working and gives me a better oversight. This type of workflow I can definitely already tell fits me the best, and for sure I will mess around with this program in the future on my very own projects.

When it came down to model the hair and folds for clothing in the model, I personally looked up tutorials on YouTube as well as courses on LinkedIn Learning, to educate myself on different ways to do so. As was said over and over again on both websites: there is no right or wrong way to do things, as long as you get the desired result in the end. There are however, easier and harder workarounds when doing different methods.

Along the way of sculpting I had some issues when doing the skirt. I mention it in my blog but when using the Move Polygon Tool to create the "puffyness" of the skirt, it wasn't quite as neat as I would've liked it to be. It took a couple tries to get it correct and in the end when I got a shape that I was happy with, I cut off the bottom/ends part of the skirt so it was a clean and even cut. I duplicated the top skirt to use it to shape my underskirt as I was already happy with the shape it was in, just needed some adjustment and scaling down to be smaller.

When I had imported the sculpted mesh into Maya, I quickly realised that my file size was way out of control and the high poly was, uhm, very high poly. I think from memory I was up in millions and it took me some time to Re-mesh and fiddle around with the sculpt to get it to an acceptable size and for it to still look good and detailed. Even after downsizing and re-meshing some of the mesh was still very dense and some even happened to go into Maya reversed so I had to go and "Reverse" the mesh for it to not appear black on the screen.

Using Photoshop (in my case Clip Studio Paint) to add textures

One way to texture models made in Maya is to create UV Image Snapshots, which basically gives you an image of your UVs, with the lines in a PNG (or whichever file you choose) that you can edit however you want. So you can paint colour and add things like fabric texture, grunge and shading to clothing and other things to make your model look nice. I originally vaguely learned how to do it with Substance Painter, so learning another way to do it was really neat. The only thing I found off with using UV Snapshots is how difficult it could be when shading and adding details to the texture. It really depends on your UVs being really nice and clean to line up the textures properly, so that you don't get a weird seam or unevenness when importing it back into Maya and onto the model via Hypershade.

Auto-Rigging with a Maya Script

To rig a model in Maya you can either do the entire process yourself, by making/drawing joints and joining different stuff together to make the skeleton. But for this module we used a script for Maya called "AdvancedSkelton5" and followed a tutorial on YouTube and OneDrive for instructions on what to do and how. In general this seemed really simple and quick to do, but to get the right placements there was a few trial and errors. When exporting/importing the rigged file into a new Maya file where the light scene was already set up - I was faced with issues trying to scale my model to the right size as it would blow up quite badly. This made me incredibly frustrated as I had followed both tutorials to a T (or so I thought) and it prompted me to go back a couple files (thank god for multiple saves) to see what I had missed or done wrong.

My assumption is that I had maybe messed up the grouping of the different things in the Outliner as it was the only thing different between the two files.

Basically what I did is open up one of my back-up files - check that everything was working as intended and edited the rig a little bit. I had to edit the rig as it was one of my earlier tries and wasn't to my satisfaction when it came to how different parts of the body was bending. The knee joints were too high and the feet ones weren't proper so the shoe heel would completely bend when it was moved around. The shoulder wasn't perfect but okay for what it needs to do and for what we need it to do with the posing, but since I was editing the rig anyway I moved it along with the elbow to create a smoother, and better bend. Will have to keep an extra eye out for my Outliner changing in the future.

ZBrush

When it came to ZBrushing it isn't much what I would do different, but what I would look out for in future projects - I would keep in mind to sculpt details and rough outlines in the low poly state before going into sculpting the fine details and adding extra things. After not paying too much attention to doing low or high poly I ended up with a massive file (in the millions of high poly) and had to decrease the quality and re-mesh most of my sculpture in Zbrush. The model was also pretty tiny compared to what I saw in the tutorials videos - and when I took it into Maya in the later stages it had to be scaled up a lot.

Re-topology

For the re-topology part I struggled with keeping things even, as well as thinking ahead of myself when having to make it look decent and for it to make "sense" in a way. The topology needs to flow nicely so that when we UV the model later, it won't create weird stretching and/or just looking weird in certain places with having stars where we could avoid it or having very dense areas.

That said as I was doing the re-topology and realising that I needed time to texture and rig the character - I had to cut some parts of the model out / off. I decided that small elements to the character - like the ribbons on her leggings, and the candy attached to her apron could be skipped as I personally felt like they weren't essential to the character, and it wouldn't "ruin" it if they weren't there. So for the future I know have an idea of how long re-topology can take, and having a lot of intricate and small additions to a character might need to be thought over better for future projects. And to mention, the hair was really difficult to fix the topology of, as it was "swirly" and very unique. I felt very frustrated with myself when I realised I had no idea how to do the hair justice with the re-topology and honestly just did my best in according to what I thought would look decent in the end.

UV-ing

When I got to the to the stage of UV-ing my model, I realised that if I had kept in mind that this stage was something I had to go through after the ZBrushing - I would be creating a simpler model with some characteristic. And by that I mostly mean the hair, as mentioned before. Since I made it kind of a swirly soft serve ice cream kind of hair, I took notice of it being a quite difficult thing to retopologise and even UV in the end. With that said I did the best I could with what I had done with the model/hair and I tried to make the topology make sense for the kind of hair style that I gave her, and tried my best to not have any weird tearing or seaming pop up. In my opinion I did a good job, but I think I could always strive to be better even though as an artist nothing will ever be "perfect" and I suppose as long as it doesn't look completely off - it's passable and just something to keep in mind for future projects. This is my first time doing this kind of work after all, and it can only get better from here as well as I can always learn from this.

Time and planning

Need to work on how to plan and estimate time doing different tasks a lot better. At least I got a decent idea of how long certain things take during this module - for me the planning took long just because I got hung up on something unnecessary, so ideally if I had started Zbrushing quicker I think not only would I have got it done before Christmas, I would have started the re-topology before Christmas as well so I not only had more time for textures, but more time over all to either look over for errors and perfect certain things.

Speaking of time management - I feel like had I done it better I could've had more time to explore the UV Snapshots and making textures that way. As well as adding shading to my textures to create more depth and made the model pop out a lot more for the final render images. Although I am happy with how the Substance Painter textures turned out, I could have tried more to make some textures of my own to add onto these.

Overall I am very pleased with the result of my project as it was my first time using ZBrush, Quad Tool in Maya, Texturing a model and even using an Auto Rigger to animate it. I think under different circumstances I could've been able to do a better job, and as already mentioned before with a better plan and timing I could've gotten more out of the different stages than I did. But in the end I did manage to design and sculpt a model on - what I think, is a really good first try in ZBrush, as well as adding some basic textures and learning how the auto rigger "AdvancedSkeleton5" script in Maya works. I was really excited to see the model come to life with the animation testing, even if some parts were a bit wonky, which is to be expected with an auto rigger. If I spent the time to rig the entire model myself, which I imagine would take weeks, it would definitely look and work better than the current set up does.

I was for sure a bit scared and nervous when I looked at the workflow and what needs to be done after sculpting the model for a while, but I after actually having gone through the process I am more confident in my abilities. If I had another go at the entire process I have no doubt that I could do a better job, and using what I experienced and learned from this to do things differently next time.