Final Film

I hope this blog has shown the work process of the film as a decisive process, that no elements where left to chance.

Our gant chart can be found here:

Whilst our gantt chart indicates that we have been blogging all year, it is in this past week which we have been back and tidied and written in a way to show a retrospective and reflective journey. We thought this would aid the flow and coherence of our reasoning and plans and results of these in regard to the project as a whole.

Attached are two business cards designed by Catharine Downes as hours for the Group Project. Cathrine did the wording and then we inverted the colours and added a picture. 

CVs can be found on our individual blogs:

lpriestanimation.blogspot.com

rbl3animation-groupwork.blogspot.com

Here is a poster we made for our film: 

May

The end was near

What was left to do now was one of the hardest scenes - the powerup gap scene. This is the bit near the end where the neurons are being lifted up by the white sparks and as we see them rise the whole environment lights up.

In the animatic, simply boosting up the contrast was simple enough to show a change but in the actual scene there were many intracacies (far more than was proportional to its length, which was quite annoying!). There were many elements and it was of course the scene with the hardest composite.

To start with the actual gap scene was properly created. There were three shots where we saw the WS of the gap and the characters on either side. 

To begin building this, we used combinations of the concepts of the gap we already had. As the environment, up until this point, had been quite dark, when the light up occured, we wanted to show the depth and scale of this environment. Therefore, a lot of the scenery in the environment was to be painted in photoshop. The way we did this was to generate a random alpha using a dirt map in maya, UV mapped the texture and exported this into photoshop where we painted on top of to get the right shape. As long as the alpha remained intact, it didnt matter that the colour parts didnt match the exact shape of the geometry in maya:

To get the environment to light up we would use a dark environment and light environment and reveal the light environment by means of some kind of masking. To save time from us masking each layer of the scene, we thought we would use the random motion of sprytes to generate an alpha.

Here is an example of how we were generating this alpha:

This worked well as can be seen in the final film.

As we were compiling and editing all our scenes together Sean commented that it would be nice in some scenes if we were to have some kind of cellular bodies floating around, to enhance the idea that we were constantly on the cellular level.

Whilst not having the time to put them in every scene, we did think it would be really good if we did try to concept them and at least put them in a few as we thought it was a good idea.  

We used the cell shapes from the cells which make up the environment and used them as the participating media. Here are a couple of tests which show the effect of this cellular effect which were going for:

Still

Moving

Here is a more comprehensive break down of how the system works for the cellular bodies:

This was the final hurdle in finishing our project. It was just about editing now and bringing it all together. 

Jack gave us an array of soundeffects. We also had recorded some of our own sound effects and we therefore mixed and matched to create some new sounds and layered them up until we got the weird, quirky sounds to match the piece. We were pleased with the result of this in the final film, especially the gloopy noise which makes it sound as if everything is taking place in some viscous liquid.

Unfortunately, we had to cut a small section at the beginning, not because it was unfinished but simply that it didnt seem to fit. Previously we intended to move through this environment and then zoom onto the character. Yet when the panning of the environment was created and inserted at the beginning, the intermediate zoom did not fit at all: it just seemed to be far to short and far too fast. We played around with different combinations but in the end we resolved to get rid of the beginning shots completely and open on the blue character opening his eyes. 

This seemed to be too quick an introduction but we had no time left to make improvements. We hoped the pan we put in to show the characters relationship in location terms would be enough to get across the environment and its mood.

Overall, we were pleased with the piece though many changes and improvements will be made for our degree show.

As April turned, whilst still animating, we were still reviewing the animatic and what we could do to improve it or make it clearer.

We thought that the gap bit was the most poignant part of our film so wanted to emphasise it. Instead of a simple cut to reveal that the characters were on either side of the gap, we shrouded the gap by placing "foliage" type cells in the forground and nesting the camera in it so we could only see jay and kay on either side of the screen, in the distance. When they started to move towards each other the camera moved in, through the foliage and finally, when it emersed from it, the gap was revealed. This had to be quickly boarded and made so we could see what it looked like.

As some of Bronny's group work hours, she animated both the characters in this shot, moving towards each other in the distance (Brons blog - adventures-of-bron.blogspot.com)

Here is a playblast of the new scene:

 

At this point, we were still aware that the inital location of the neurons at the start may not have been as clear as it could have been . What we were thinking of putting in was a pan through the environment near the beginnning which would clearly show two small points of light at either end. A background for such a scene was made and then created in maya

Here is the backround that we used:

Here is a movie of one of the tests we did to see how this was looking (Bron also animated these characters):

This shot didn't quite gel well when we inserted it into the animatic. (It was inserted after the top down shots at the beginning.) We would cut to this scene, panning across the environment and then into the pink character. However, an intermediate shot was required as the cut was a little too severe. We thought it would have to be some kind of wide shot just not as extreme as pan shot. The following playblast shows what we came up with - we were pleased with the result:

Things were going well. A lot to animate but we were confident it would be done. There was always more to problem solve though.

In our animatic, when the blue and pink sparks from the neurons came close to one of the characters, we wanted a nice glow to appear on the characters, just as in the animatic. However, this was not as simple as it sounds to do because the shader used on the characters did not respond to light in a normal way. Heres a playblast of one such scene where this was needed to work:

It took a while to work out but a solution was found. 

What had to be done was set up two additional render layers; one where the character rendered totally in the highlight colour (e.g, if it was the blue spark, then the character would be blue) and one which textured the character with a lambert shader. This was affected by the sparks light in the expected way, creating shadow and highlight. These values were then used to drive a key mix in compositing where the value of the lambert shader corresponded to the coloour which was shown in the mix.  

Below is a clip showing the stages in how we got this effect. Additionally, it also shows the stages gone through in making this particular scene, involving adding backgrounds, and the spark rushes. This was a small scene yet quite a few processes were needed in the compositing stage:

Into march and we had our characters, our assets to make the scenes, it was not just a matter of creating the scenes and animating the characters.

Firstly, before creating the scenes, we thought it would be useful if we pinned down the compositions for some of the pivotal shots. 

We were aware of the flow of the eye within pieces and actively tried compose the scenes in this vein. 

Below are the various scene compositions:

Heres an example of a composed scene can be seen below. It was composed using the above composition.. At this point in the story, blue is seen searching around for pink. The animation using the rig has been blocked in. The compositions were assigned to a polygon plane in the scene and the geometry referenced in and composed this way:

It was in kay hill scene that we chose to do one last test with all the assets just so that we could be certain that after had animated the character in the scenes, we could actually achieve the look that we wanted through compositing. 

Below is a  number of test composites with the two characters in this scene:

We now knew that we could achieve the look we were going for . Below is the final composition process that we go through , with the resepctive render layers which we used as the standard model from this point on in the project:

OK. So everything was in place to start animating. The scenes were built, the characters were rigged, the asssets made, the composite tests were done, we knew that we could achieve the look that we wanted.

The only thing that we needed to do now was to create neurons' indivual sparks. These have always been envisaged in the same way as the animatic: their movement defines how they look really - fluid but snappy.

Below is how the sparks were created:

The following is a render test for when the spark was moving (this is not an animation test so does not move as the sparks will)

The rush of sparks which were needed would have been unfeasible to have animated each one individually so dynamics were used to recreated the swarm. This was quite simple effect to achieve. Here is a first test we did:

The final look for the sparks and how they would effect the environment was one of the first shots animated as it was quite difficult. Using one of the shot comps above, and the dynamics, this final scene was created:

Backgrounds

Whilst we have these scenes, we had to create a few generic backgrounds for the pieces. The an example of some of the backgrounds we used can be seen below. They had to be quite depthy. We did find though that the desaturation when composited bleached out and looked a bit muddy. Some had to be gone back over to resaturate:

There were many things which now needed to be done in order to make head way with the project. We needed set designs and sets being built, character rigs. The characters also needed to be tested for movement and facial expressions.

The Environment:

To create the the environments, we didn't want to create blocky, chunky scenery so resorted to the building-block speed painting way of composing a scene. We thought by creating individudal cell like structures and intermingled them with a variety of orbs it would create plenty of variation in the scenes. 

Before we went ahead though, we had to ensure that the layering system worked as many cells would be layered on top of each other. Below are some test renders:

These worked quite niceley although the movement of the cells looked a bit too underwatery so we decided just to have static scenery. 

Below is a break down of the process behind creating the indivudal cells to compose the environments:

Orbs:

The orbs were not as concepted as the rest of the elements in our piece as they arrived later but we had a good idea of how we wanted them to look. To reiterate, the orbs were the visual representation of the mechanical, electrical side of the brain which we were trying to balance with the cellular, biological side. The orbs were were the white sparks were born from. Whilst still being mechanical they still grew from the environment as if they were organic plant like. 

Below is some further concept for them before they were designed and created:

Below is a break down  of how the orbs were created in Maya:

The character:

We did some character sheets initially so we could get some perspective on their manouvability, facial expressions ranges, limits of their movement and actually HOW they would move considering their strange body mass and construction. 

Their bodys were both composed of an outer "membrane" which encased a "cell-body" in side which served as the head. In order to keep things as simple as possible, we knew that we would have to try and get as much expression out of simple facial features as possible. These facial features initially included eyes, mouth and eyebrows, as seen in much of the concept and animatics but it was an added extra we tried to do with out. Below is a sheet of facial expressions which were done to try out this simple approach:

We thought these were working well which was good as now all the head would require were eyes and mouth - much easier and faster to animate but this test showed that it was still feasible in terms of stylistic choice.

We had an interesting path ahead of us as these creatures were completely made up and so to be believable we had to create their whole physicology and how they move accoding to this. Their head inside the membrane we always intended to be free to move of its own accord. Yet despite the "other-worldly" creature, we were still bound by the guidelines of animation, in terms of posing etc. The below sheet is one look at posing options and how the body, limbs and head would all work in union to achieve solid, strong and believable poses:

When we were sure that the characters would actually work pose-wise, giving us the best chance at potentially solid animation, the rigs were now to be made. 

Before the final shape and rigging of the character was made, we needed to be sure we could replicate the softness of the photoshop images which we liked a lot. Below are some tests for this:

The tests went well. We would be using the same settings on both characters just one would be blue and one would be red. 

Below is a break down of the rigging process for the Jay character:

When the rig was made, we did some final lighting and shading tests to make sure that everything retained the soft feel:

We now began to really play with there movement. There were parameters which were defined by the physicallity of the neuron and the constraints of the rig. We now had to find a balance with these two factors, getting the best compromise. 

Below are some thumbnails for a potential Jay's movement:

Unfortunately this still wasnt quite enough so at the expense of looking a fool, we shot some reference, all in the name of art:

Eventually the movement was worked out and a render test can be seen below. The quirkiness we found was very appealing:

This ending was better and the idea we were much happier with. It was the beginning which now drew our attention. The split screen, whilst completely confident with to begin with, was beginning to look a bit out of place, especially as the piece was becoming more nuanced and understated.

 

Before we started ripping that to pieces just yet, we began playing with adding a little something before the beginning, establishing the characters and their emotions and the environment more. Below is a couple of beginning tests we did to put at the beginning of the current animatic:

 

 

Nope, still wasnt working. A rehaul was in order. This time we thought we may introduce some of the white sparks in the beginning which would later bring Jay and Kay together at the end. We wanted to establish the idea, not just that the neurons were sad that they were disconnected but that they were saddened and actively searching to reconcile the situation. Below are two alternate beginnings which would replace the current split screen situation and join the rest of the piece when they look over the edge of the gap:

 

It was getting better but there was still something not quite right. We needed to try and establish what each character felt for the other very early on, i.e. that they were dear enough to each other that they would become uplifted at the sight of them and painfully dejected at the thought of them not being there.In the one below we changed the beginning, trying to isolate them in frame and converying the isolation they were facing in their situation. Also we had them get excited by what they think to be the radiance from the other (before they actually saw each other) only to realise it wasnt:

 

The holes in the environment that blue was looking through didnt feel in congruence with the rest of the scenery so another change was in order.  It was at this point that we began talking about adding another element to the scenery. Up until now, we had only cells and a few wiry objects. But when thinking about how we could get blue and red pseudo spotting each other we thought it would be good to use some kind of distorted reflection. 

 

The reflections tho were like nothing we had thought of before and would suggest something quite metallic. It was quite fitting tho when we thought of the van der graff generators - the metallic balls which which held a static charge - they seemed the perfect sort of object and was completely congruent. Concept for these orbs can be seen in Januarys posts but for now we were fairly confident this more industrial style for the brain was appealing and would work a treat. Here are a couple of changes made to the animatic where blue spots red in the reflection of these orbs and vice versa:

 

 

 

Here is the final animatic that we handed in for the January summative. We used this as the template for getting the timing for our animation, yet were still open to changes and which we did change small bits in the end. 

Now that we were happy with the animation, it was time to finalise the characters and begin more previz tests, in terms of colours, textures and just generally how we were going to achieve the look we wanted. This would stretch in to much of February and March. 

We began to look at how we could achieve a cellular look, both for our characters and the scenery. We always knew we didn't want the look of our film to be typically CG so trying to aim for even a low level of realism was something to be avoided. This would be something we had to work upon. Below is the sum of a few trials are getting a specific celllular look. 

It didnt look quite right, more like they were marbles. They were also too realistic. The characters are good examples of the 2d-3d mix. They were to be 3d created but intended to look 2d somehow. Many test had to be done as were done throughout february and march.

However, the characters by the end of January were also pretty much final. Here are some final concept drawings for them before they began to be modelled and rigged in February: