Friday, 20 April 2012

The Blunderbus Returns

My redesign of the Blunderbus has been finished, and I have continued to learn the importance of lighting and texturing correctly. I could explain HDRI ranges and all that crap, but I don't really understand it myself! Instead I'll let the two images speak for themselves.

A regular old Mental ray render to start with, showing the new masucline design of the Blunderbus and his reflective and emmisive textures. Rather mediocre image at best.

HDRI rendering says hi.
And the posed, HDRI final gathered image rendered at production quality from Maya, a completely different image and mood and a much more successful image altogether. Notice the changing colours all over as I slowly re-rendered after editing parts of the image. Then I discovered IPR rendering, cutting the wait time for renders down even further!

Not Work But Work 2: The Skivining

Continuing to make background models for my other courses to use in Vue Xstream has helped me understand Maya's materiala and texturing to a much higher degree, for instance this Lancaster Bomber made for my time based narrative course.

It's very pretty, and even taught me something!
Using Maya to create simple backgrounds for Vue has made me realise just how important a workflow is to creating and managing assets on larger projects like this, as well as when materials>uvmaps.

Tuesday, 3 April 2012


A recent inspiration in texturing for me has arisen from Capcom's work on the Street Fighter series' recent incarnations, most notably Street Fighter 4 and Street Fighter X Tekken. While I enjoyed Street Fighter games as a lad I have pretty much ignored them for the last ten years or so (perfection was reached on Street Fighter 3: Third Strike in my humble opinion)

Recently a friend invited me to play some of the latest releases, and while they did not exactly blow my mind gameplay-wise I was amazed at the graphical style that Capcom had achieved.

Big for more detail
Capcom are a company that has always had my respect for understanding how classical techniques can inspire and create. The older (2d) street Fighter games looked much better than any of their rivals, because Capcom took the long (and a bit mental) path of hand animating all characters before digitally transposing them into the game. For their later games they have perfected the hand-painted 3d model style. The pictured game in particular is a fantastic example, when characters stop moving, you could swear they were simply painted in their position rather than being a 3 dimensional model (maybe not in the example pictured)

I will look into this form of texturing, as I feel it brings an organic beauty to 3d models that most  models lack - my own models of Lucy I felt were too clean, and her style didn't allow much time for detailed work. Hand painting her textures using Sai Paint Tool or Corel Painter could give her the detail that the model desperately lacks and hopefully make her appear a little bit more alive.

I will not be using Mudbox or Photoshop's 3d paint tools however, as they are absolutely awful and a real let down considering how amazing both of the programs are at other aspects.

Blunderbus Designs

While working I have started to think about the designs of my characters, specifically who they are designed for. The original project had them in the starring roles of a family game, designed to be played by families with the older members of the family controlling the Blunderbus.

This brought the idea of who the Blunderbus had to appeal to. The girl character of Lucy is obviously designed to appeal to the younger gamer, but I felt that the Blunderbus failed at attracting older gamers. My own perspective would be playing the game with my nephews or niece, and I don't think that the robotic character would be that appealing to me or other gamers in my position. Considering the only reason the character exists is as a glorified babysitter for the younger player, they would have to be appealing for the older player to actually play as. Obviously I cannot rely on game mechanics to convey this - as there aren't any - so I have decided to look at the character and redesign him to fit with the (probably) male players that will control him.

This is a redesign of the factory background. Modelling the industrial robot arms made me realise how little the Blunderbus looked like any kind of industrial robot. The specific question was raised, how can he actually fix anything?
Working my way from the top to the bottom of the character I have started reworking his design, keeping recognisable shapes while adding more masculine forms and tools to allow him (Notice I always refer to the Blunderbus as a he? That was my main reason for deciding to make him more masculine) to actually work as a repair robot/janitor. Notice the robotic welding arm and the vacuum cleaner added to his midriff.

I am only working on the left side of the model, planning to mirror parts for the right side. The lights on the shoulder are emissive (light casting) textures rather than actual scene lights to keep the render time down and the lighting controllable. Although the 'codpiece' has been reworked, I am yet to start on his hips

Friday, 16 March 2012

Normal mapping

After an incredible amout of work, I've been able to get my model to convert to a subdivision from a polygon, allowing me to add extra detail in Mudbox. I prefer working in Mudbox to Maya, as with my Wacom monitor/tablet it creates a very natural feeling interface to model with.

The normal mapping process will affect the UV mapping, so the maps will have to be redrawn- it was already my objective to upgrade and reskin the model but now its a necessity. The useful part of this process is that I can take this very high poly model as a normal map within Maya, allowing me to keep the poly count down on my already loooong renders.

I have also been thinkig about the composition of the pictures I wanted to create, and will be working to improve the backgrounds in texturing mostly. Now that I have mental ray rendering working again I can start to use some of the advanced materials within there to make the environments much more interesting.

Thursday, 1 March 2012

Not work but work

For another course, where I have been tasked with creating an animation. For practice and for a much smoother output I have decided to use maya to create backgrounds for this hand drawn animation. Even if I simply use these renders to paint over they will be invaluable in keeping continuity between scenes.

The above shot was created quickly using a simple schematic of a Douglas DC-3 found through google images, with reflective textures added. Although the shot is missing several important features, like windows or landing gear, it lets me see the correct placement of shadows and reflections in the aeroplane's surface.

This may not seem entirely useful for my current project of Lucy and The Blunderbus, but it shows how quickly a model can be created and how much of it can rely on the correct texturing of said model. Despite the 2 million odd blades of grass, the reflective textures and the 2 hour (I'm not lying) render the picture is absolutely wrong looking. It's a good thing I'm planning to paint over it

The Blunderbus

The Blunderbus had a character design far more complex than what I ended up with, with many areas being dropped for simplicity. One aspect, for instance, were his fingers - which were supposed to fold out of his arms to create his hands. This was entirely dropped for time but now I can focus on actually getting them modelled correctly.

His elbows and knees both suffered as well. In the design phase I created joints for his arms to move on but again had to resort to simple spheres to fill in.

A very early design that featured a few elements dropped for time. THEY WILL RETURN
It's been a while since an update so I will do my best to get everyone up to date :)

My models came out reasonably successfully for the first semester, but they do need work for semester 2, so first I will outline my objectives here before talking about progress.


I like the Lucy model I created but I am not entirely happy with the finished result. My main weakness when starting this project was the modelling side of the characters, as I had never managed to create an organic model before. However, by focusing on creating an acceptable 3d model I have ignored one of my skills at texturing.

Although I am happy to have created an organic humanoid model the texturing on the hair and clothes is very basic

Lucy will be getting a makeover to reduce topological errors, such as triangles and polygons that share vertexes, but the modelling side of this character is pretty much complete. Lucy's final model featured simple specular and bump maps - but these are what I am targeting for improvement in this semester. These include the file problems that caused broken connections most of the time when importing or opening the file on other computers than my own. One aspect of correction with topology will be for normal maps to increase the detail on her tracksuit and facial features.

The Blunderbus

Because the main objective of the project was to create an organic character the protector character of The Blunderbus (Don't know if I've explained on this blog, but he's named after a camper van my dad used to torture the family with - the point being that it was an ex-ambulance that was not built for the job but kept us safe anyway) was largely ignored. The model was very simple at the best of times and because of the focus on modelling Lucy he was left with many placeholder elements - the feet, for example, are simple pyramids that I meant to upgrade but ran out of time for by hand in. These are one of the  many modelling based examples that the blunderbus needs.

I have to admit this model came out far worse than I hoped.
So I feel that by focusing on improving weaknesses I have actually failed myself by not allowing myself to use my strengths to improve characters. I have always been good at making mechanical models, but did not allow myself to model the Blunderbus successfully because I wanted to improve skills in other areas. I have been texturing for ages but did not allow myself to use my full skills because I was too busy concentrating on skills that I had not learned yet.

With another semester to improve my models I am hoping to improve where I failed by allowing myself to use my skills.

Friday, 27 January 2012

New Semester

New semester and time to continue the development from semester 1.

My main objective is to increase the detail on the characters I have created, and to make them ready for rigging (I will probably rig them as well)

By the end of the last semester I feel that my  characters strayed from their original designs so I will look at returning them to the forms that inspired them - right now that means looking at the dimensions of Lucy and making her more realistic - I feel tshe's too tall and skinny for a little girl. The Blunderbus also needs development to make connections between the geometric shapes of his joints and limbs