Evil Baby Character

Ok so for the past few days I've been conceptualising my evil baby character and trying to model him. I've looked at many images but the main one which really stood out and inspired me was one I found from deviant art:

After finding a nice orthographic picture of a baby on-line I decided to use it as a base for the character giving it an over exaggerated look to the character with a huge head and small body. I did this because i did not want to fall into the same trap as I did before in making realistic models. It only made life hard especially when using low poly geometry modelling. I figured that since the baby wont be focused on as much as the main character it wouldn't be necessary to have as much detail in the model. Learning from my prior mistakes I realised that I should base most of the detail of the UV maps instead of the model itself. There will be large amounts of babies on the screen at the same time when the game in complete so this also emphasised the point that it did not have to be very very detailed. As long as there was enough geometry in the joints to move and animate I was happy.

As for the texture I chose a nice flaky effect with gave the baby a zombie like appearance mixing dark colours and blood together. I contrasted this with the colours of the nappies to give it a strong aura about the character. It looks scary but at the same time cute. Ultimately I am extremely happy with the final outcome. Instead of painting I primarily used pre-made textures that I overlayed on top of each other to give the desired effect.

                                                                       

Walk Cycles

I have also been putting a lot of my time and effort into walk cycles of the main character. It meant that I had to set up a mirror in my room so that I could watch myself walk across the room. I examined human movements and used this information to make a walk cycle. The first walk cycle went well, however I realised after making the rig that I had not froze the transformations of the controllers which meant it was very hard to make it symmetrical  Not only that but I did not attach a global control to the rig which meant that the model walked across the canvas instead of in one place. This meant that when the model was imported into unity it actually walked away from the camera when you pressed 'W'. 

Because of this I had to create a whole new rig and start the animation from scratch which allowed me to correct my mistakes. In the end this was the best thing for the walk cycle because the final result was far superior than the first. Practice makes perfect. Not only that but from watching the character in the game I realised that in order to go up stairs the character had to jump so I thought it would be wise to add a jump cycle as well as a stationary pose. Here are the final results: 

Overall I am quite happy with it at the moment, the feet look quite static but that's because I used a reverse foot lock technique in the feet, evidently it works great on high poly characters but not so well with low poly. I will keep this in mind for next time. 

Environment Modeling

While working from home I also began working on the crypt model within the game. This modelling is A LOT easier than character modelling as you can block out major shapes with simple objects like walls/floors etc. I'm developing a workflow for buildings where I can reuse objects to save myself a lot of time. This was the first attempt at modelling a building and it turned out better than expected. There are 2 tombs, with stairs in both leading down to a secret room where the player will encounter the 'music box.' This is just one of five scenes in the game that players will encounter. All I need to do is UV map the building and its finished. I have however already given Tim the base mesh for him to play around with and position within the game.

I modelled a few different versions to show the team and find out which one they liked best. Here is the final outcome:

UV Mapping

I was not happy with the colouring of the last character. This was because I was using straight colour shaders from Maya which made the character look very plain and bland (what I like to call Iceblock material). This really made the character feel tacky. I wanted the audience to 'believe' the character in the game world and so I decided to UV map the entire character instead. I also did this so that the group could have some input as to what they wanted the character to look like. With my workload already taking up all my time I figured it might give other team members a part to play in the character. However in the end I decided to do it myself as it was too hard to explain how to paint the UV map and know where to place certain shading to emphasize shadows on the model. To help me regulate where exactly I needed to paint I used the 3d paint tool in Maya to highlight areas where I wanted to add more detail.

From there I imported the files into Photoshop where I used an oil painting technique to layer on the base colour of the character. From there I added details and shadows to the character. The last step was to use the Photoshop function 'Overlay' to combine real life textures of the top of the clothing, skin, and hair to give the character a more realistic look. 

NEW CHARACTER

Ok so I made the decision to make a whole new character because I was not happy with the original character. Not only was it very high poly but the mesh did not distort well when a skeleton was applied to it. I decided to simplify the character down to be sure that the poly count was under 4k. This would ensure the game would not slow down if it was to be run on a laptop.

I found that with designing low polygon characters it was a lot more challenging because you have to use every polygon efficiently so that the character will distort with the skeleton without artefacts appearing on the mesh. This was harder than I thought it would be. I learned what I needed from the first character so not all was lost from that time spent. I also designed the new character keeping in mind the earlier problems. For example, I made the limbs slimmer which would deform better and studied human anatomy to work out which direction muscles flow in the body to generate more life like rotations and bends.

Communication is Key.

As the title states, I am finding that as we draw closing to the due date me and Tim are having to communicate a lot more which is changing the dynamics of our workflow. Because we are both learning the programs we have no idea what will and wont work. After creating a whole new model and walk cycle animation I gave the FBX to Tim which produced some very interesting results. Results that transformed my character model into something that resembled a retarded Tweedledum.

I was trying to solve the problem for days, remaking the rig, altering scale ratios of the rig controllers, changing units. But nothing seemed to work. Surprisingly enough after a bit of playing around I found that the problem could be solved by changing the settings in the actual FBX file from inches to cm. This fixed the problem because unity works in metric units and was being confused by the FBX format. We sighed with relief because it meant that Tim could play around with the animations.

That is another issue with our workflow, it seems that for a specific task to be done it means that someone else has to finish their task. Its like a hierarchy. I cant model till I know what exactly is in the game and Tim cant progress in unity until I finish specific models. If someone doesn't work it means that all of us suffer.

Insane Workload.

I think we should definitely have thought more about what roles were assigned to people. Im unsure of what everyone else is doing. The only people in the group with a solid role is myself and Tim. It seems every time I come into uni people are just doing their own thing instead of focusing on the project and hacking away at things which really need to be done. I keep coming into uni and seeing conceptual drawings being done by other members in the group but it is far too late to keep producing them. I think the group needs to focus on the final product and stop drowning in concepts. I keep saying that we need to actually start producing items which are going to be USED in the final presentation. Such as the animation cut scene at the start of the game. I'm becoming increasingly frustrated when team members play games during studio time where I'm working my ass off all day everyday till 2am trying to learn Maya. This seems to be an everyday thing and I cant stand it any more especially since we have so much to do in so little time. I know that with only 2 people on the technical side of the game we are going to bare the brunt of the majority of the workload. With that said I am getting a lot of work done and I am now fairly fluent with Maya.

This also bring up an important issue that I am having. Because I am working so hard to learn this program it does not really allow me get too involved in the conceptual side of the project which worries me. We agreed at the start that we would have people dedicated to sorting out the story but it has just taken far too long to get a clear idea as to what exactly I need to model and animate because things keep changing. I wish I could partake more in the conceptual side of things but with 5 people in the group things are just getting way to complicated. In a way I am glad that I have a concrete role in the group because I know exactly what I need to do for the day when I wake up in the mornings. There's no "wondering where to go from here" thoughts which is absolutely great.

I  think from now on I will just work from home as I just get far too distracted at BCT, its not a good environment for me to work in especially modelling on a small laptop. I haven't been to uni in a week and I have accomplished more than I have in the entire semester.