Whillock: May 2010

I decided to have a go at building something from scratch in Maya, just as a bit of practice modelling things. I chose to do a robot as it should mostly be geometric form and few organic curves. Anyway, here's my attempt so far:

This project was a complete dive in at the deep end for me, having only touched on 3D modeling in simple game/level editors such as Unreal Ed, Hammer, and a couple of others. Realising how different and much more in-depth Maya was, was daunting, but also inspiring to see that the possibility of creative freedom was virtually limitless. From poylgon and NURBS construction and modeling, to texturing and using lamberts/phongs/blinns/other materials, and then lighting a scene, working the camera(s) and lighting, you are in control of everything, and it enables you to become the sole producer of the piece.

My final animation achieved exactly what I wanted too, although wasn't perfect. This to me was a good opportunity to try a few different aspects of Maya and just get the general workings of it into my head. I wanted to keep it relatively simple throughout, and after learning the lesson that everything you make might not actually be in shot, decided to tone back some of the detail for things just out of shot, or only on camera briefly. A more comprehensive storyboard may have helped solve this from the beginning.

Learning to use certain properties of the types of modeling to your advantage was one of the main learning points in this module. Choosing between NURBS and polygons for different objects was sometimes obvious and sometimes not so much, and the same goes for which shaders to use to get the right sort of material you are looking for.

There is so much to learn in Maya, I think this kind of approach of producing small projects with challenges is the best way to go about getting it all to sink in. If you tried to tackle it all at once it'd just be too much. I am going to keep plugging away at Maya as it's interesting and, as I said before, full of creative potential once you have the basic skills under your belt.

Changes I would make:

introduce some dynamics in the physics of the objects falling, to make it look more natural. i.e the fall and bounce of the glass jar onto the plank.
add more detail onto objects such as the books/tomato
remove some of the clipping issues such as the needle and tomato point.
learn bump mapping and use it correctly.
storyboard more thoroughly next time.
render time planning will be easier next time now I have done it once and can see averages. It was hard going into it with no idea.

I was particularly happy with:

certain objects I created, such as the fan.
certain aspects of the scene lighting to change atmospheres midway through the animation.
keeping a set theme and feel to the piece.

What never made the Cut:
Here are a few of the items that got cut from the animation, usually for rendering time constraints. So I thought I'd share them with you here:

The final rendered version of my animation for my "Introduction to Digital" module, It was produced as an exercise into modeling, texturing, lighting, and animating in Autodesk Maya 2010.

Big thanks for help from Toby Rutter and tutor Kelvin on the bits I got stuck on!

Let me know what you think of it!

Evaluation and process coming in the next few posts.

With great timing, the folks at Wired Magazine have this article about how Toy Story 3 is created:

Here is the Article

From what I took from that, not only how long it takes for a modern feature like Toy Story to get from first ideas to final production, is the whole creative mindset each individual who works on it must possess.

It's not just a 'job', it's always living as a creative, letting your mind be open to new ideas from the people and the world around you, and finding joy in a million different things, showing narrative where others would never expect to find it.

This also reminded me of this video I discovered through Roxy Gomes which is all about just how much hard work plays in getting far in the creative industry. It's all about the mindset and determination, as well as having the creative drive and imagination:

Just a few renders from the almost-completed lab set-up, now with a lot more detail, and don't worry about the sparse areas and shelves etc as they're not in shot in the final film. The lighting is pretty much done now, and I've got the feel I was looking for - dim lab mostly lit by spotlights/lamps, and a dark side lit by the eerie glow of machines/experiments.

Looking at a few references of laboratories and mad scientists (very much the movie-versions)

And so I've started to add details to make it a bit more 'mad-sciencey', like making the liquids glow a bit, warning signs etc.

Quick playblast with further animation

Looking at render time differences when adding complex elements

Rendering the electricity

The scene so far

For the Maya project we have been given four weeks to produce a short animation based on a Rube Goldberg (or Heath Robinson) style contraption. I have loved these style of machines ever since I was young and used to build them around the house. Here are a few of the best examples of them I have found around on the web:

I haven't said much in my blog posts of late as I've been knuckling down with the Maya work, but this seems like an opportune moment:

Firstly texturing has been gotten the hang of slowly, great fun producing the blackboard:

Playblasting has really been helping me get a feel for the timing and how the animation will look and I now feel most of the camera work is done except for minor tweaking. The same goes for the main parts of animation - it's all moving now, albeit not perfectly but its getting there:

So pretty pleased with the progress I've made so far, with some of the detailing and bits and bobs left to do. Any feedback welcome!

So far, my rendering times haven't been too bad, usually 50 seconds or under for a frame, but with the addition of this lamp, they have started to creep up, to a maximum so far of 1:37.

I have to be careful as I don't want this to take days and days to render, but hopefully if I can keep each frame around this level, it shouldn't take more than a day or so:

1m 38s = 100s (approx) per frame.

Therefore: 25fps x 100s = (2500s) 41.6minutes rendering per second of footage .

So a 30s animation = (1250m) 20.8hours to render.

Fingers crossed!