And finally, its all over. Im just collating all my files together ready to hand in. Can’t believe how fast the time shot by. Also can’t believe its almost time to be a 3rd year. Crazy!!
Anyway, here is the final showreel of my 2nd year work. Unlike the longer coursereel, this does not have the entirety of my work on it. This simply has the best of what I’ve done during this major project. Plus it has music. Woooo.
Finally managed to locate the cause of the issues with the spine when the dragon was rotated to face in a different direction. The base bone in the spine was flipping by 90 degrees whilst the others remained fine. It was due to creating the spine as an IK spline set up and wiring the rotation handles to controllers. Although this set up works fine in an upright spine, it appears to cause the handles to rotate strangely when the spine is horizontal. I have deleted the IK and re-linked the bones together and wired the controllers back to the correct things. Hooray for a working spine.
I also realised that my centre of gravity controllers were not set up correctly. As I had placed them above the dragon, I was causing the dragon to rotate around the pivot of the controllers and not the pivot of the dragons hips. Some extra point helpers and a few extra position and orientation constraints have fixed this. Also hooray!
To try and give the animators an easier time, as well as improve the general look of the animated dragon, I have added a flex modifier to the dragon’s ears. This was done by first adding an edit poly layer in which I soft selected the tips of each ear. The flex modifier was added above this and then only affects the sections I had selected in the edit poly.
The skinning is finally almost finished, and I am now working on getting the deforms as smooth and natural as possible. One of my biggest problems has been the deformation of the back leg. When the knee twists away from the body, it causes the mesh behind the leg to sink into the body and form a large indent. There are several possible solutions to the problem. The best one would be to create an extra bone linked with the leg that the mesh is skinned to that can be animated to force the mesh back out when the leg twists. However I am already behind on this project and although it would be the neatest and most effective solution, I simply do not have the time available to experiment and get it working properly. As such, I have instead created some simple morphers that can be switched on and off to fix the dent when the leg is brought forwards/twists.
Originally I had intended to only need to spend a day or two updating the original dragon mesh from the advanced tech project. However I hit a huge snag when I first started editting involving 3ds Max corrupting any file in which I tried to edit the dragon mesh. Its taken me about a week to track down the problem which seems to have stemmed from using a symmetry modifier, and collapsing it down during advanced tech once it was time to texture and rig. When I then tried to edit again at the beginning of the major project and add a new symmetry modifier Max just couldn’t cope and crashed, corrupting the file in the process.
This has unfortunately already pushed me behind schedule and leaves me scrabbling to play catch up. None the less, here is the finished dragon mesh, now with wings and a tidier back (which it did’t have during advanced tech).
So it would appear a pair of 3rd year VFX students had their own project idea that involved compositing creature animation into live footage. They actually were aiming for a bit of a “His Dark Materials/Golden Compass” type theme but this still fits well with my original plan to try and composite some of my animation into the streets of Cardiff. However it appears they don’t have quite the numbers of animators/artists (so far at least) required to hit the scale of film they had been planning – which is a shame because it would have been awesome if this had turned out to be the big project of the year. Ho hum. Anyway, it looks like it may just be one 2nd year artists and myself doing the CG side of the work which rather limits the number of creatures they can have. In fact, Paul and I have agreed that we can only manage two creatures. While Paul handles the modelling and texturing, I will be rigging/skinning both models and then creating a walk and run cycle for both as well as a single non-cyclic animation (ie standing up from sitting or something similar). So, that covers 6 of my 12 weeks on the major project.
For the other 6 weeks I will be working alongside Elaine and a few others to help produce a short animation we have currently nicknamed “Project: Dragon”. It involves a rather silly chase between a hapless knight and an overly playful dragon. I’ve set myself the rather terrifying project of both modelling and rigging the dragon (eeep!). However, we have an artist who loves to draw who will be doing all the horrible design work of the characters. As such, he will hand me completed turn arounds so I dont have to do any of that nasty stuff involving pencil and paper. Or at least… not in designing the look of the dragon. I still need to work out the rig structure. The modelling of the dragon (and first attempt at rigging) will actually occur during our final advanced tech project. Three weeks to model, texture and rig (a week for each). I will then probably pass the model to Ruth who may tidy up any poor edge loops to aid with skinning the final rig. Jess will also completely re-do my textures. Once that is all complete, I will build a completely new rig, get the model all neatly skinned and set to work creating clever controllers so our animators can have a great (and easy) time of animating him. So that takes up another 3 weeks of major project time. The final 3 weeks will be devoted to doing some animation with the knight in the project – a run cycle, possibly a walk or jog and then some sort of non-cyclic animation.
Scarily, we are almost three weeks through our pre-production, and currently I feel as though all I really have is research. Will have to slave away at it all after Christmas.
So, the last two weeks involved learning about more complicated lighting within 3ds Max and how to composite CG objects within a real environment/photo. First there was a short afternoon class exercise to learn about the standard lighting in 3ds Max and how to replicate common phenomena by hand (ie place lots and lots of lights for reflected light etc) rather than with clever programming. We were given a room set up with 3 spheres, one of which we had to texture as metal, one as glass and one that was self illuminating. We also had one red wall and one blue wall and a light in the ceiling.
Next, we had to take our hand model from the previous advanced techniques project and create a “studio lighting” set up that would show our model as well as possible. This set up is extremely useful for any future projects as it provides us with a pre-made lighting set up for any renders to best show off our work.
Finally, we got to the main section of the project. This involved taking a couple of photos in very different lighting conditions and then placing a CG object within the scene and trying to make it look as realistic as possible. It was theoretically a very interesting project and the theory learnt will be incredibly useful. None the less, I can definitely say that I do not find lighting even remotely interesting or enjoyable. In fact, its spectacularly dull.
I think the desk scene with the pen was far less successful than the dice. I still haven’t really worked out why that was. I just struggled to get the lighting to work. Ho hum.
This project was based around the idea of the 11 second club. For anyone reading this who doesn’t know what that is, its awesome to look at. A nice monthly competition to animate to an 11 second voice/sound clip.
Anyway, this project was about four weeks long, one of which was pre-production and one of which was reading week. It was a hard slog, seriously hard work but a good introduction to animating and the stages that go in to creating a polished final piece.
We broke the animation down in to a series of stages to get done one at a time. Firstly we blocked out the main poses, then added some in between poses. Over the first week we just built up the in between poses in a stepped animation format. Once we had got out blocked animation fairly cemented we were able to move onto the smooth animation and start tidying up. Once things were tidy we could finally set to work on the facial animation and the lip sync. The eyes/eyebrows I found tricky but was pleased with the final result. However the lip sync, although pretty much in time to the clip, the actual lip movements were very unconvincing in places. Sadly, I just ran out of time to be able to keep playing with it.
So, I said weeks and weeks ago that I would upload a slightly slower version of the wasp turnaround. Here it is… at long last!
Soooo… the trouble with a busy, manic course is that there is no time around the work to actually keep this blog up to date. Looks like I may have to try and set myself a time once a week to sit down and make blog posts. However, since I have found an oppurtunity today, I shall take full advantage of it.
The rest of the first two advanced tech course works involved modelling a realistic hand in one week and then rigging, skinning and texturing it the next week. To practice skinning techniques we also had to rig and skin our bug in another 24 hour session. Needless to say, it was a pretty intense couple of weeks. Still, the modelling side of things turned out pretty well. I struggled at first while trying to get my head around the idea of “edge loops”. This basically means keeping the flow of the edges around polys moving nicely around the model and trying to ensure there are as few as possible, so that they actually just loop around and connect in various places.
Skinning and rigging the wasp turned out to be a real struggle due to a few mistakes made during my modelling process. As I had modelled the legs quite bent, it was impossible to make them straighten nicely with the rig. Also, thanks to a lack of edges around the leg joints the bends were very untidy.
Having completed the wasp, we moved on to rigging and skinning the hand. This went very well and I had it mostly completed on Friday before stumbling across an error. One of my tutors helped me to find a work around for the problem and a way to reload all my work on to the rig after the fix. This worked great on the Friday. However, I came in on Monday expecting only a small amount of skinning left to do only to find that the fix I had been shown on Friday had now deleted all my work and I had to start again from scratch. Unsurprisingly I was absolutely gutted and this seriously affected the quality of my skinning. There were several errors I failed to notice in my rush to get everything complete and I watch my deformvideo now with frustration that the deforms are so much poorer than my first attempt.
We also then had to texture the hands. Rather than using photoshop, we used procedural texturing, using various noise algorithms to create the random colour changes etc that you see on skin.
Finally, once the hand was rigged we also needed to pick a final pose and animate the hand moving into this pose.