In order for the animation to be as accurate as possible, I wanted to follow the movement demonstrated in the video clips as much as possible, even if I needed to use bits of many clips to get the one continuous action I wanted.
 
 
 
 
 
 
 
 

Walking animation:
From looking at the walking images, it was apparent that polar bears use their back legs to propel themselves along. As the front legs are not as strong as the back legs, they serve more to keep the bear balanced as it walks along. As the back legs are stockier with their strength, the front legs of the bear tend to stretch out much further than the back legs on each stride. To compensate for this larger stride, the back legs move a little faster than the front to keep the overall speed of the bear constant.

 The neck of the bear is quite long so that heavier back half of the bear is balanced. This length means that there is a slight bounce of the neck as the bear walks along. It was quite important to add this into my animation, as it was a vital clue as to the weight of the animal. 

The speed of the bear as is walks had to be taken into consideration, as I wanted to get across the weight of the animal in its movement. My first attempt at animating the walk resulted in a movement that was too slow, and made the bear look as if it was old and was having difficulty supporting itself. After speeding up the walk to a pace more in line with the evidence on the video, the bear movement became much more realistic.
The weight of the bear was also shown with the wobble at the neck and belly. As the bear walks along, so the vibrations of the impact of each of the paws as it hits the ground are felt along the underside of it. I animated this to a certain degree using the extra blocks I had made in the model for this purpose. Recreating the rippling nature of the wobble was not the kind of realism I was looking for in the animation, as I was looking for more of a representation of the movement. To this end, the two under-belly and under-neck blocks were simply animated swinging from side to side.

I had left the foot rolling part of the front legs as an animation decision when I built the model for the bear. So when I came to making the feet roll, I decided that the best way to represent the movement was simply to pull the foot block of my bear back and to the side, then up and round, in a much simplified version of the actual flip the feet actual make. Being as the actual movement is very quick, when I attempted to make the foot roll more realistically, it just looked as though the feet were snapping into the wrong positions and the blocks were spinning around their axes. Had the feet been made in more detail, this would not have been such an obvious problem, but being as this was how I had chosen to build my model, the simplified version of the action was much better for my simplified model.

Once I had made my polar bear walk along as part of my sequence, I needed it no bend to look at something on the ground. To achieve this I had to pull my model into quite a drastic position. This caused a lot if the blocks to cut through each other in ways that would not happen in real life. This was, however, the point of using parented blocks rather than deforming skin, as this meant that I did not need to change my animation to prevent these intersections. There were a few problems with the legs bending in the wrong directions when the model was bent in such a way, but nothing that couldn't be easily sorted with a little tweaking. So otherwise, if the blocks intersected drastically, I did not worry about it as long as it did not detract form the animation.

I had intended for the animation to end with the bear turning back up from its stoop and walking away, but once I had got the model to bend in such a way, it was difficult to get the model to revert back to walking. I could not get the back legs to move behind the bear as it stooped to look at the ground as the front and back legs would always intersect in a way that even a blocky model could not hide. The animal itself bends its body out of the way in a way that my model was not able to do. Once I could not move the legs around, getting the rest of the body to revert back to the walk would have looked unnatural, and would have involved trying to guess how the bear might stand back up without having any real material to back up. Being as I did not want any animation that was not based on anything I could study, I decided to have the bear look up at the end instead of walking on.

Hunting Animation:
For the first part of my hunting animation, I wanted the bear to look like it was being stealthy. In reality, the bear hunts out its prey under the ice by balancing itself as it moves. It does this in such a way that each paw is placed on the ice before any weight is put on it, so that any vibrations through the ice that might alert its victim to its presence are limited as much as possible. This means that the bear always has its opposing paws on the ice at the same time for balance while it carefully finds its footing with the other paws, all the while listening for movement under the ice.

As the listening operation is such a delicate one for the bear, I needed to slow its animation right down to give the impression of careful concentration. The wobble of the under side of the bear was also less important, as the slow movements keep the wobble down to a minimum. What was more important, however, was the neck movement. The bear cranes its neck around to try and catch the scent and/or sound of the seal, and so this would be the greatest movement the bear would be making with its body.

Once the bear had established that there was a seal under the ice, I needed my bear to sit back onto its hind legs before deciding where to finally pounce. To do this, the bear pulls all its legs underneath itself before raising its front legs. Unfortunately, the inverse kinematics on the front legs of my model meant that the action achieved was a little unnatural where the paws gave an impression of leading themselves rather than the lifting of the shoulders pulling them up. To combat this, I animated the paws swinging a little more than they did in the video clips to stop them looking too static.
I could not get my model into the same squatting position as the bear in the videos, as in reality the bears folds into itself as it sits back, and I could not get my model to do the same. I moved the bear into as close an approximation as the model would allow, however.

To animate the actual lunge of the bear through the ice, I looked at the different clips I had of real bears taken from different angles. What became apparent from looking at these, is that each jump is quite individual and only shares basic characteristics with other attempts. The front paws are not always held out at the same place below the body, the amount the body arches is not always the same, and sometimes the head is held up and sometimes it is arched down. I decided that I would incorporate elements of each leap as I felt appropriate. I decided to keep the paws quite central under the body as it fell, whilst keeping the head up. I wanted to give the impression of the bear stretching out as it hammers all its weight onto the ice, so I made the model stretch out as it jumped up ready to fall, even though this meant that the blocks of the body split apart momentarily as it rose. 

When the bear breaks through the ice in my animation, I simply let the bear disappear through the ground. I could have made a hole for the bear to fall through, but from the videos it was apparent that the ice is covered in such a thick layer of snow, the bear does just disappear from the shoulders up once it has dived. As the bear disappears further into the ice at the end of the animation, I simply moved the bear down through the ground. 
I wanted the final leg flick to look as though the bear was trying to get down to the seal under the ice. Being as the bears quite often kick their legs out as they slip further through the ice, there was quite a bit of video evidence, but being as the leg flicks are a reflex reaction and are therefore all different, I chose one image to base mine on. I decided just to have one leg kick up, as two would look too comical, and this was an image of the polar bear I was trying to get away from.

Conclusion

Home