Kiwi!

I stumbled upon this fantastic animation, which was produced by a student of computer arts. Here is a video link:

I find it both impressive and inspiring that this short was produced by an independent animator, studying for a masters degree. The movements of the Kiwi bird are very weighted, and I particular enjoyed the animation of the bird heaving on the rope attached to the tree, possibly demonstrating anticipation. Another creative feature as far as character design goes, is how Kiwi uses his feet to perform tasks. It adds a great comedy value to the short. When the bird hammers the nails into the tree, we even see the follow through principle as the body subtly wobbles with the motion of the foot. Another effective aspect of the animation is the story. We are made to laugh for the majority of the short, as we see this tiny bird trying to pull this colossal tree up the face of a cliff. The whole idea could even fit our module title 'Little and Large'. The story then takes an emotive turn however, as we realize that Kiwi is attempting to simulate flight, as he has no wings. Of course, this simulation does inevitably end in death, showing that the bird is essential committing suicide, to experiencing the dream of flight. This emotion is emphasized as Kiwi closes his eyes in preparation for his death. The character seems at peace, and we can see that.

Clearly, 'Kiwi!' is a fine animation, demonstrating many of the principles. What I took from the short however, is a look at how great narrative can be achieved through computer animation. Like our videos, 'Kiwi!' is only a short animation, yet it manages to on some level move its audience. It really demonstrates the power of animation.

'Little and Large' (Final video)

Here is a H264 render of my final animation. I am generally happy with the results, and I feel the colours and the lighting look very nice when rendered out. One aspect I am a bit disappointed with is the ray trace shadows. By forgetting to tweak the settings, some of the shadows are very sharp and blocky. Unfortunately, the rendering process took so long that adjusting the shadows would have cut my project worryingly close to the deadline. Another improvement I would have liked to make if time permitted, would be the easing with the bouncing ball animations. I could have really demonstrated an animation principle by easing out the ball to a halt. I believe I have still covered the easing principle however, generally with the considered curves in the graph editor. You can see the principle in simple animations such as the Lego man pulling down the switch that appears from off the screen.

Overall, I am happy with the animation. There are obviously still improvements that could be made in future projects, but as a first attempt, I believe the project has given me a nice introduction to Maya.

Three Point Lighting

I recently followed the three point lighting tutorial on moodle, using the key light, the fill light to bring out the shadows created by the key light, and the back light to create an edge around the subject making it pop:

Key:

Fill:

Edge:

I found the guide very useful. It covered everything from emitting speculars, to the relationship editor, which are all important to consider when rendering out high quality images in Maya. I plan to apply a lighting 'rig' in my 'Little and Large' Maya scenes, to hopefully bring out the subjects with in my animation. I plan on using directional lights for a more natural look, as my animation is set in an outdoor environment. Of course, this is something I will need to test first, as I am not fully sure how the objects within my scenes will react to the directional lights.

Revized Animations

I made a previous post about the final crit we had recently. In response to some of the comments I received, I have revized some of my Maya animations. Here is a short reel showing the improvements:

As you can see from the video, with the kick animation the ball now takes off at the moment of contact. I have also added more weight to the ball as it drops out of sight. I have also pulled out the tangents to add more weight to the ball as it bounces in to the shot on the second clip. Finally, I worked on improving the squash animation where the large beach ball crushes my Lego man. I simply inserted new key frames on the y axis in the graph editor to add extra bounces. The first attempt contain too many bounces, which made the ball loose its weight. In the end, I ended up using the single quick bounce that was suggested during the crit. The result is a ball that appears heavy, yet does not stop too suddenly at impact.

MegaMind: Behind The Scenes

Expanding on my previous post on the extreme deformation style of Dreamwork's 'Madagascar', I found this behind the scenes video showing the production processes behind the movie 'MegaMind', also by Dreamworks. The video gives great insight in to the entire production process behind a high quality animation. If we refer to the chronological order of the video, it is suggested that the voice actors come in before the animating begins. This makes sense, as I imagine it would be easier to key a CG model to dialogue, than for an actor to try and achieve sync with a computer animation. What I also found engaging, was to see some of the concept art behind the characters in the film. We see development after development, pages of faces and varying hairstyles. This gives an idea in to the amount of work that goes in to creating a character, before the modelling process even begins. It is also clear that different people are working in different areas of computer animation. We see one artist producing an epic fly-cam over a city landscape, where as one animator is working on the character's facial expressions. It is also interesting to see the facial animator manipulating MegaMind's eyebrows on what appears to be a graph editor, which I often use in Maya to try and achieve smooth animations. Seeing this, it becomes clear how much detail is required for such an large scale production, and also, that perhaps this requires more animators taking on different roles. We also see one team member who's role seems to be to review the rendered footage and make sure no mistakes are present. We see him point out that a small streetlamp is superimposed onto the camera view, ignoring the large object destroying everything in its path.

In response to the video, I believe it has given me a general idea of the type of working environment within a large animation studio. There seems to be many different roles, so that focus and effort can be distributed evenly, ensuring every aspect of the finished animation is achieved with time and care. 

Animatic

I have produced an animatic to help me plan out how each shot will flow, and the kinds of sounds that I'll be using:





I created the animatic by taking the frames from my storyboard, and placing them on a Final Cut Pro timeline. The animation will last around 37 seconds, which fits nicely within the time boundaries set. I tracked down royalty free sounds clips from the web, and went for a classic cartoon approach. I am currently in the process of rendering out my scenes, and will soon be ready to work with the image sequences. By producing an animatic before hand, I now have the advantage of not having to track down sounds while I work with the images.

Blur Studio

I recently came across this animation reel from 'Blur studio.' The company have worked on some major titles, including James Cameron's 'Avatar,' where they produced the space sequences. What I find most striking about the reel is the amazing render quality, which I believe is relevant to the process I am currently at with lighting and rendering out my scenes:





The clips in the video are generally cinematic, with very considered lighting and camera movements. The animations themselves are very dramatic and gripping. There seems to often be an emphasis on realism and detail. It really is hard to pin down one segment of reel to elaborate on, as the overall video is so powerful. Remaining on the subject of cinematic approaches however, the segments from what appears to 'Star Wars: The Force Unleashed' are rather impressive. at 2:38, we see a character pull back his hood. My first impressions of the scene is that the cloaked figure is dark and corrupt. The lighting implies this as the key light is rather strong, creating strong shadows below the subject's eyes. A fill light appears to have been used, but only very subtly. This leaves the character's face quiet heavily in shadow. To make the figure pop out from the shot, a soft blue edge light is generated from a light within the backdrop. Overall, the figure in the shot is left looking very mysterious and uninviting. In some of the proceeding shots from the same segment, we can also see what appears to be volumetric lighting. If you pause around 2:46, you can see beams of cool blue light pouring over the large creature in the shot. There seems to be a theme running throughout the short Star Wars segment. The strong use of cool blue lighting is perhaps used to generate a cold chilling feel, as I believe the game focuses on the dark side of the 'force'.


Overall, within this reel there are some stunning examples of mood lighting and fantastic renders. Some of the sequences appear hyper realistic and ultra moody, creating a very impressive collection of clips. I now feel inspired to experiment with lighting in Maya and render out my scenes.

Dreamworks Madagascar: Extreme Deformations

Casting my mind back to the time I watched the Dreamworks animation 'Madagascar', I remember seeing a segment within the special features where the animators spoke about new capabilities with the squash and stretch deformers. I believe I have already mentioned I enjoy working with diformers, so I wanted to find this video and be further inspired. Although I was not able to find the video I saw some time ago, I did come across the above embedded video in my search. The technical reel I found does not only give great examples of the use of diformers, but also shows other technical aspects like character rigging and 'systems' for the sand and waves.

elaborating on Dreamworks' great use of diformers, at 48 seconds to a minute, we see some very extreme examples. First of all, When Gloria (the Hippo character) kicks open the create she is encaged in, we see what appears to be a bend deformer and perhaps a little bit of stretch used on Alex the lion. This makes Alex mould around the fired lid in a very cartoon fashion. We then see some great examples of squash and stretch with the next two clips. With the chase sequence, where two of the characters run around the Giraffe (Melman), we can see stretching particularly as alex tries to spread his weight when executing a sharp turn. The next clip labeled 'squash and stretch' is my favourite example of extreme deformation. when slowed down, as Alex jumps into the frame he almost appears as if he is elasticated. Everything about the animation is extreme and over the top, particularly in the way Alex's hands are left behind by his body, and they pulled forwards with a massive stretching swoop.

When we look at Madagascar and consider the animation style, the exaggerated diformers really characterize the whole feature. For a beginner like myself key framing in Maya, the results can be very stiff in rigid. Dreamworks have really managed to achieve fluidity in their animations, which really inspires me to try and achieve more weighted and expressive results.

The Final Crit

Today we had our final crit, where we displayed our work so far to the group and received feedback. Here is a reel of the animations that I took to the crit:





The first animation is a test I created for my lego man kicking the beach ball. I am fairly happy with this animation, particularly with how the arms swing with the momentum of the kick. I tried to incorperate some squash and stretch, but the ball seems to have a slight delay after the Lego man's foot makes impact. I was going for a compression and then a release, but realistically, after being kicked the ball would imediately take off. I imagine this shouldn't be too hard to correct, I just need to make sure that the  curve for the direction of the ball is straight and does not ease in.


The next two animations are leading up to my final throw animation. Reflecting on the original throwing test, the ball seemed unweighted. This is what I have been working towards since, giving the ball a nice weighted feel. My first realization was that the point where the ball hits the ground and bounces should be pointed. the Y axis curve must meet at the ground point with no easing, and instantly change direction for the bounce. However, when adding the squash and stretch deformer later I realised a slight ease is needed for the compression, and then release of the bouncy beach ball. With the next animation where the ball bounces in to the shot, I was advised that to get the weighted feel, the tangents needed to be freed and pulled out to give hang time and then a heavy drop. This was a breakthrough with my ball bouncing animation. At this point after the initial tests, the clips are ordered chronologically based on the storyboard, meaning the final throw animation comes after the clip of my lego man picking up the beach ball. With the final throwing animation, I really tried to get this weighted feel by using weighted tangents then pulling them out. This gives a cartoony exaggerated motion which I plan on going back and applying to the ball bouncing in to the shot.


Another other key animation that I have produced is the pull chain shot. the chain drops into the shot and eases to a halt. I then used a stretch deformer to manipulate my Lego man's head so that he is looking up and investigating the mysterious source of the handle. The Lego man then simply reaches up and gives the handle a gentle pull.


There is also the 'squash', where the handle being pulled results in a huge ball crushing the Lego man. Although the large object still shares the physical appearance of the soft bouncy beach ball, for comedic value I wanted the large ball to act almost like a big sphere of concrete. The ball stops dead at impact, leaving only the Lego man's feet showing. The huge ball then does a gentle roll revealing my character laying flat on his back. There is a subtle ease in and ease out with the rolling of the ball after impact. It was suggested however, that it might look more convincing to include a very subtle bounce at impact. I plan to try this and compare with this original test.


I feel confident that the animation process is very close to completion. However, I must manage my time well as the lighting and rendering process could take some time. I then need to locate sounds and place together the sequence on a timeline. I felt that the crit was helpful, as it allowed me to compare my progress with the rest of the group, and see where I am doing okay and at the same where I need to catch up. A final thing that was mention was that perhaps a displacement map could be more effective for my Lego floor. I did a quick test with this but it was unsuccessful. the texture map (bump) did not translate anything like the original UV, but hopefully I will find the time to experiment more with the settings of the displacement map. (Also note that as I have corrected my maya timeline playback speed, the later animations have more realistic timings to save time when rendering).

'Little and Large': Throw Animation

This particular animation took me a while to get right. Perhaps it should not have taken as long as it did, but I am pleased with the results nevertheless:

I began the animation by having my Lego man drop his arms and throw the beach ball to the ground. I figured this would look nicer than trying to have my character throw the ball in the air, his arms being so tight to his body. Probably the hardest part was getting the right hang time at the top of each bounce as I keyed  each frame by hand. Looking back perhaps I could have utilized the graph editor for the balls motion, drawing the path inside the graph editor window instead of hand keying it along the timeline. I will try this when I animate the regular sized ball rolling in to the second shot. Perhaps the most tedius part of the animation was keying in the rotation of the ball. As I did not do this as I went along, I had to go back, rotate the ball, write down it's current coordinates , apply these coordinates to the next frame, rotate the ball some more, write down the new coordinates for the next frame and so on. The squash deformer was quiet nice to use. I had the ball squash each time it hit the ground, stretch out as it bound in the air and then rotate on release. I also had the arms swing after throwing the ball, keeping the follow through animation principle in mind.

'Little and Large': Beginning the Animation Process

Fitting with the title 'Little and Large', here is my storyboard:

In my animation, I plan for my character to be met by a small beach ball. The Lego man will then toss this out of the frame. After this, a small pull chain will drop and my character will curiously pull it. This will result in a giant beach ball falling on top of him. The 'Little and Large' concept comes with the slapstick nature of the small and then hugely oversized beach ball.

I initially had the idea of using a beach ball in my animation, as I enjoy working with the squash and stretch diformer. This is also an opportunity to demonstrate squash and stretch as an animation fundamental. The 'ease in ease out' fundamental will also be covered as the ball slows to a stop in the second shot. I believe I will also be able to show anticipation as my lego man lifts the smaller ball and prepares to throw it. However, after presenting the story board, the point was raised that it may be hard to achieve the motion of my character bending down, picking up the first beach ball and then throwing it. It was suggested that perhaps using a kicking motion could still be successful, yet easier to animate. In response to this I plan on producing two tests, one that keeps the original idea of throwing the ball, one using the kick.

After gaining feedback on my storyboard, it was time to begin the testing process. The first step was to produce the surface, which later will become part of the environment. I felt I needed this surface to keep everything grounded and on the same level. I sometimes find that using the grid on Maya can be less effective than a solid plane. I had the idea of producing a lego surface with the cylindrical connecters spread evenly over the plane. I thought however that this might be tedious to achieve the small cylinders, so instead I used UV and bump mapping. Here is the result:

The next stage was to produce the beach ball. Again I used UV mapping and a subtle bump map:

I managed to animate my Lego man bending down, taking hold and then lifting the beach ball. To achieve this, for the bending down motion I grouped all of the upper body down to the hips and crotch joint of my Lego man. I then moved the pivot point holding 'D' on the keyboard to the centre of the crotch joint. This enabled me to get a turning motion where realistically a Lego Man would bend. For the movement of the arms, I grouped the left and right arm together. Again, I then moved the pivot point to the centre of the shoulders. This meant the arms kept in position in relation to the upper body. Keeping the arms in position I then animated my Lego man to stand up. After this I simply moved the beach ball in to position, keeping it in time with the hands as best I could. Then with the upper body set back to the 0 position, I animated a final lifting rotation of the arms. Again I tried to mimic this movement with the ball. Here is a Playblast of the finished result:

It seems however that my maya timeline plays back too quickly. This is why the animation here is very slow and drawn out when Playblasted and also rendered. I could either re-animate this part or I could speed it up in post production. Nevertheless, hopefully this shows an effective picking up motion. I now need to animate my Lego man throwing the beach ball. Once this is complete I can create a kick animation and compare the two.

'Little and Large': Modelling my Lego man

For our brief within this module, we have been given the task of producing a 30-45 second 3D animation that demonstrates some of the fundamentals of animation. I began the process my searching for reference images online. Although I had my own Lego man on hand to look at closer in more detail, to get the basic scale I figured I would be able to find some decent scale references online. This is what I found:

With my image planes set, I now began modelling. My methods heavily involved dropping cubes and manipulating the vertexes. To achieve some of the more complex shapes such as the arms and hands, I worked in segments, extruding the faces and continuing to move the vertexes in place. Hopefully this image should demonstrate what I mean:

To achieve the shape of the legs, I had to get the smooth circular thighs. You can see this in the side on reference image. Here, I placed a cylinder and rotated it onto its side. I then grabbed the relevant vertexes and extended them down, creating the legs. I added the feet on last as separate polygon shapes. This image should highlight the process:

To create the head I also used extruded faces, this time with a cylinder shape. I carefully went around the vertexes and tidied them up to create a smoothed out look. I believe I could have done this much easier and faster by instead of manipulating the individual vertexes on the extruded face, using the scale tool keeping the vertexes synchronised. Here is the head:

It was now time to add colour to my character. I wanted to go for simplistic classic lego man look, as I didn't wan't to overcrowd the model with too much detail. In the Hypershade menu, I created bold 'Blinn' materials for the legs and upper body. I chose the primary colours red and blue to maintain this idea of simplicity, and used the Blinn option to give a shiny plastic look. The hands were given the classic yellow skin colour we see in almost all Lego men, and the head was created using UV mapping. I opted for a simple smiling face, to keep a very purist approach. Here is my Lego man in full colour:

I am quiet satisfied with the results of this character model, although I do believe some of my methods were perhaps amateurish. Ideally, I would have liked the legs to be one solid shape, but the way in which I achieved the circular thighs meant I could not extrude out the feet. Also, the arms were created from extruded cubes, and I was not able to achieve the circular shape where the hands slot in on a Lego man. Instead the arms look quite square near the wrists. I do believe however, that the modelling process of this character was a great learning curve with Maya. I look forward to developing my methods and hopefully overcoming some of the limitations that compromised my character model.

CG in the Original Star Wars Trilogy.

In re-released version of 'Star Wars IV: A New Hope,' There is a scene where Crime Lord 'Jabba the Hutt' threatens to put a bounty on Han Solo's head. The scene includes a CG Jabba interacting with Harrison Ford, however the scene was original shot with the intention of adding in a stop motion Jabba the Hutt. Because of costs and time, the scene was not included in the original release, but the improvement in CG technology, was re-introduced in later re-releases. I managed to find this compiled video taken form various sources, showing the original scene with Irish actor Declan Mulholland filling in for Jabba:

From this original scene filmed with live actors, the challenge at this point would have been placing in the CG Jabba the Hutt. There are points where Han Solo is standing in front of Jabba in relation to the camera. Today I imagine a green screen technique would be used for such shots. Perhaps with the ability to manipulate a 3D model and play it alongside the original video content, a live actor to fill in for Jabba would not even be needed. Obviously the process of planning and then producing a stop motion animation to be placed next to a live actor would need to be much more intricately planned, the original actor being used as a marker for each movement of the alien crime lord. Take a look at the finished sequence:





The visual effects supervisor Joseph Letteri said that the aim was for the revisited scene was for it to look as if Jabba and Han Solo were interacting with each other on set. Apparently the team had trouble getting the CG Jabba to resemble the original puppet used to portray him in 'Return of the Jedi.' After the 1997 release, they even polished the CG Jabba later in 2004. However, the computer generated results still do not fully resemble the puppet.


One thing I found interesting was a problem that arose and how the team tackled it. at around 50 seconds, Han solo walks behind Jabba. This was not considered in the original take, and the later concept used for the puppet meant that the alien's tail would be in the way. To solve this, they animated the CG Jabba to let out a yelp signifying that Han Solo had stepped on his tail. They even managed to elevate Han Solo, emphasizing further that Jabba's tail was over his path. I believe this was a very clever way of making the live actor and the CG model appear as if they were both sharing the same physical space.


Despite the CG model not reflecting the original puppet used to portray Jabba in Episode V, I find it very interesting how a 3D animator can create the illusion that a live actor is truly interacting with a computer generated model. I have already visited this concept with my last post on Jurassic park, and the placement of CG in the original shots. Although puppets and animations look more realistic in my opinion, I am beginning to see that working with CG can unlock many more capabilities.

Money For Nothing

Remaining on the subject of CG landmarks (Star Wars IV), I thought I would mention the first use of CG in a music video. I am a fan of the band Dire Straits myself, so I was surprised to discover that their 1985 music video for 'Money For Nothing' is very significant in the history of computer generated imagery. 





It is noticeable that the 3D technology here is very limited. The CG characters in the video consist of very few polygons, making them appear very square and un-detailed. Also their movements are very slow and rigid, and lack the expressive squash and stretch element that many 3D animators use today to create fluidity in the movement of objects. Not to take anything away from this video, at this time in the 80s human characters conveying emotion where only beginning to appear in features such as 'Adventures of Andre and Wally B' and 'Tony De Peltrie.' Although these characters are much more rigid than many in the features being produced by studios like Pixar and Dreamworks, the introduction of CG in to a music video must have been ground breaking at the time as music appeals to such a global market. The possibility of creating computer generated characters would have been seen by so many, and I do not doubt that a higher level of interest towards CG was generated.

Early CG in Star Wars IV

As I have recently been re-watching the original Star Wars trilogy, I thought I would write a post about the first use of three dimensional CG wireframe model rendering in film. It is fascinating to think that a 1970s computer had the capacity to create 3D models. We seen in the video that the creator of the sequence, Larry Cuba, was still working with dials to control and navigate around his work. It is humorous to note the time between each click when he uses his graphics tablet. Clearly the equipment he is using is not very responsive in comparison to the computers around today. Everything about this process is tedious and time consuming. He explains that he 'wrote a program' to combine the six trench modules to create the whole structure. He also describes the stop motion-esque method of recording the trench run simulation. It took 2000 exposures to show the approach toward the Death Star and through the trench. Just the thought of this drawn out method causes me to cringe. In programs today like Maya along with much faster computers, 3D models and animations can be created with ease. A playblast to preview an animation takes only seconds to be converted into a quicktime movie. Seeing methods such as this really helps one appreciate the evolution of 3D animation. The software and machines around today are highly advanced in comparison to the tools used by Larry Cuba, and it is hard to comprehend how much the tools for a 3D animator will improve in the future.

Jurassic Park Before and After

Following up on my post about Spielberg's choice to switch from Go motion to CG in Jurassic Park, I found another interesting video. Although I praised the advanced stop motion technology which Spielberg originally planned to use, I do believe the CG effects in Jurassic Park still exceed many of the films out today. This video shows the process of placing in computer models and green screen actors in to some of the shots in the movie:





What is perhaps the most fascinating and also relevant part of this process, is the attention to detail when placing the CG creatures in to a live well lit environment. We can see in the 'First Sight' segment, the body of water was 'painted' in separately and then 'prepped' to take the reflections of the dinosaurs basking in the small lake. I would also be keen to find out what a 'heat map is' as apparently one is used for this particular shot. The 'T-Rex Vs Car' scene also gives us a look at the rigging inside of the T-Rex model, as well as the 3D car model, which I always assumed to be real. Even the actors are absent from the initial scene and apparently were placed in later, which shows the extend of polished modern techniques where everything is slotted together in post production. Building further on responding to real lighting conditions, we also see various 'shadow elements being placed into the scenes. For the 'In the Kitchen' segment we really get to see CG models responding to live actors in the scene. My favorite shot is at 3:50, where the boy is hiding from a raptor lurking just around the corner. The movement of the raptors head as it swings over the top of the kitchen surface creates a huge amount of tension, and leaves you at the edge of your seat.


Overall, this video shows how 3D modelling and animation comes in to film, and can be used in unison with live actors and lighting conditions. It shows how you can create realistic shadow elements and reflections to really place a CG model in to a real world.

Deformers, Hypershade and UV Mapping

Building on on our understanding of modelling and animating in Maya, we were introduced to a few new tools. Firstly, we were given a brief introduction on using nonlinear deformers. by manipulating these deformers and setting key frames on the timeline, there are many smooth animations that can be achieved. With this knowledge, I was eager to try out a squash and stretch animation using a simple sphere shape. I would firstly like to slightly digress and show an animation of a solid heavy sphere being dropped. I created this animation by applying key frames on the timeline, and levelling out the top of each ark in the graph editor to add a very subtle hang before each downward transition. notice how each bounce becomes smaller until the ball is stationary:



Here is a screen shot of the graph editor showing the curves, each time growing slightly shorter and not as high as the last:

Now compare the hard ball to this exaggerated squash and stretch ball I created using a squash nonlinear deformer:

I like the over emphasised cartoon-like bounce using the squash deformer. It is perhaps something I will incorporate in my character animation in the very near future.

Building further on our understanding of Maya, we also looked at the Hypershade tool. I added a football texture to the ball and a grass textured plane for the ball to bounce on. I also experimented with lighting briefly, which is not visible in the following play blast:



Looking deeper in to the Hypershade tool, we also tried some UV mapping. I found this stone paving image from a google search and used it to create the texture for a cube. here is the jpeg net:

To add realism, a greyscale version of the above texture was added to the bump mapping layer to add the illusion of depth and texture:

I really was impressed with how the quick and easy application of a bumping map image made the stone block look so realistic and textured. To emphasise the realism even more, I added some moody lighting to the image. I took inspiration from the orange to blue colour bridge in art colour theory. This meant that the colours worked together and complimented each other. also the warm orange really jumps out from the cool blue, added exaggerated depth. here is a screen shot:

I found the UV mapping process very fascinating and feel I could really use it to create realism when modelling my character.

First introduction to Maya

We were recently introduced to 'Maya' for our 3D modelling and animation module. Our first task was to follow a step-by-step tutorial and create a small truck. Although the three dimensional model was in essence very simple, I at first found this new software very hard to work with and tedious. There were certain commands such as the grid snapping and moving the pivot position which I struggled to understand early in the process. Also I often got confused with the left right and middle mouse click functions. Although this task took me a while to complete and was frustrating at first, I believe it really helped me understand some of the fundamental modelling skills in maya such as working with faces and vertexes, and also working in multiple views. I found it important to utilise the different camera views when attempting to line up objects accurately, and the perspective view was useful when assessing the model as a whole. Here is my original model:

After completing the first model, we were encouraged to add to the truck as we pleased. As I had already spent a lot of time on the tutorial and at this point I was still finding Maya quiet confusing, I opted for a simple approach. I simple added a small trailer onto the truck, making it appear more like a van. Here is the result:

With our truck model, we moved on to the process of animating. We began simple, translating our object into different positions and setting key frames on the timeline. I found this fairly straight forward and attempted to produce a sliding truck by translating and adding a slight rotation to the object to show the back end of the object swinging in to the turn. Here is the result:

As well as this, we produced a simple swinging pendulum. We used a similar process of adding in the key frames. This time however we used the graph editor to add a sense of a hang at the top of each directional swing. This made the animation look much more smooth and realistic. Again at this point this point I was finding the process fairly steady and nice to follow. Here is the first swinging pendulum:





This next exercise was where the process became difficult and confusing. We were next given the task (still following a tutorial) to animate a second swinging pendulum with joints. I will begin by showing a screen shot of the finished graph editor:

Because the second pendulum had joints, the outside fragment had to finish it's swing last, at the point where the fragment closer to the pivot had to be already beginning a downward swing in the other direction. to keep this animation smooth and flowing we had to extend the curve by selecting the infinity option. I found it very difficult to emulate the curve in the tutorial and still now believe I would struggle to create this smooth swinging motion. I plan to experiment more thoroughly with the graph editor and improve my understanding of it when we begin animating our character. For now though, here is the second pendulum:

Finally, we followed instructions on how to create a CV curve, giving our objects a motion path. As I produced this at home and didn't have my truck model on my external hard drive, I created a quick and simply vehicle model to travel along the track. This techniques was fairly simple to follow and I was very interested by the quickness of the process. I believe this technique could be very handy for smooth flowing objects such as vehicles or objects that hover, but perhaps would not work quiet as well with a walking animation as there is a constant smooth gliding tempo which could look un-realistic. Here is the finished animation:





So far I still believe there are fundamental elements I find difficult to understand, the graph editor being one. I do feel however I am beginning to grasp basic modelling fundamentals, which could obviously still be expanded on. I have though, acquired a student version of Maya on my personal computer and plan on experimenting with modelling and animation to build on my understanding of the fundamental tools within the software. 

The Thing: Models vs CG

I have already mentioned the new 'The Thing' prequel in a personal and professional practice post. Having seen the film, I would like to present a comparison between the visual effects in the first movie and the 2011 prequel for my 3D modelling and animation module.



Rob Bottin was behind the creature designs in John Carpenter's version of 'The Thing'. In my eyes, what this scene really shows the viewer is how the thing's ultimate goal is survival. Ever single cell inside the creature is independent and will try to escape when threatened. This is shown when the detached head of the thing sprouts spider-like legs and slowly crawls away. The whole animation has a cringe worthy quality, particularly as the skin on the neck tears leaving the creatures head on the ground. As real materials are used, the tearing of the flesh and liquid substance oozing from the creature is much more believable than what we might find in a CG sequence. As every movement of the creature is created by a mechanical motion, there is a very jerky uneasy look generated. This only adds to the realism in my opinion as I find sometimes CG can be overly smooth, as mentioned in a prevous post regarding Go motion in Jurassic Park. The use of physical props makes the contrast between live actors and the creatures itself much smaller, making the whole viewing experience much more absorbing and authentic. I find the head using it's tongue to pull itself across the ground particularly realistic. The head animatronic has a great jaw action, which is so constant that I often find it hard to convince myself I am witnessing a prop. Let us know take a look at one of the memorable CG animations form the 2011 prequel:



This particular animation is much smoother, gracefully gliding from side to side as it stalks its prey. There is a very nice stretching motion of the exposed muscle tissue between the two fused faces. There is also much more freedom in the eyes, making the creature appear much more menacing as it scans its surroundings. Although I have expressed my love for the visual effects in John Carpenter's 'The Thing,' I must admit I think this is a fantastic 3D animation. The level of detail really is quiet impressive. From the visible gums to the individual strands of hair, the model is highly realistic which cannot always be said about some of the animations in even the highest budget films. Even the folds in the skin and the movements of the bulging neck muscles are quiet baffling.

For me though, without sounding like too much of a purist, the visual effects in the original movie that inspired the prequel are irreplaceable. CG in film (in my opinion) always runs the risk of clashing with the live actors and surroundings in the frame. I do think that computer generated animations have their place though. In games, 3D animation is growing ever more exciting with the introduction of 3D scanning and motion capture. I believe this level of hyper realism could really enhance the whole immersive gaming experience.

Stop motion tested for 'Jurassic Park.'

Before deciding to switch to computer generated animation techniques, Spielberg original planned to use state-of-the-art stop motion animation for the movie 'Jurassic Park.' Often, when we thing of the traditional clay or model animation, we associate the styles with quite jerky un-realistic illusions:



Here is a fantastic stop motion animation produced by Ray Harryhausen for the 1963 film 'Jason And The Argonauts.' In my opinion, the way in which the skeleton animations interact with the live actors is an animation marvel even today. Back in the early 60s, this sequence, I imagine, would have been an immense achievement in movie special effects. Still though, some of the movements aren't fluid and the interactions with the live actors are still limited some what. If you jump to 4 minutes and 4 seconds in the above sequence, you can tell that the actor is being very reserved and thoughtful with his sword movements. This is likely because he is aware that a painstaking process of molding the skeleton figures frame by frame in to the sequence is yet to come, making his actions more reserved and less complex. For these reasons, it comes as no surprise that Spielberg was seeking something a little bit more cutting edge for his 1993 high budget feature. Bellow is a stop motion test for the Tyrannosaurus:




I believe that this animation is very impressive. It achieves a kind of realism you don't always get with shiny overly smooth computer images. The skin appears to stretch as the figure walks, which is a very life-like quality. Even the slight jerks in the movement do not take away from the realism, as we imagine an animals movement not to be perfectly smooth and flawless. Phil Tippet was responsible for this walk cycle, using his 'go motion' technique. This technique aims to create a realistic motion blur by instead of taking frame by frame stills, an animation sequence consists of small split second movements to eliminate the jerk we see in traditional stop motion.

So why did Spielberg switch to CG even after seeing this great looking walk cycle? Perhaps it was because the whole stop motion process is very painstaking and time consuming. Also, when working with computer images I imagine it would be much easier to go back and manipulate the animations and make changes.

It is a great shame that this 'go motion' process was not used in Spielberg's masterpiece. I think that the walk cycle looks highly effective and could easily stand up against, if not exceed the computer generated animations used in the final movie.