Major Project 12 - Drone Animation

Now that the textures are all completed, I can move on to finishing animation and rendering. The few shots that I had already animated before needed to have their animation copied onto the latest version of the model because the UV had changed. This was an extra hassle due to my haste to get onto animating while waiting for Dan to finish the drone. It wasted quite a bit of time, which was slightly irritating. There were only 5 shots where I had to do this so it wasn't too bad. The new animation on following shots is easy enough, as it is only the drones flying which involved just a handful of keyframes so it is going at a very quick pace at the moment.

The longest shot to complete so far has been 41 where the first drone explodes. I had experimented with some dynamic physics simulations but they weren't working well. The issue was that I wanted to be in control of where the pieces landed. Also, because all these piece on the body of the drone are so tightly packed, they weren't reacting naturally when becoming soft bodies, they just flew off in really unrealistic ways. In the end, while keyframing the pieces took a long time, I believe it gave me much more control and the fiddling I would have had to do to a simulation to get it how I wanted would have been just as much an effort. I am really happy with the result of this shot so I don't mind that it was a time consuming task.

LoopDeLoop Piggy 3 - Animating (Part 2)

Animating the pigs proved more complicated that I expected, not on a technical level, but for the creation of the loop. I had considered using a simulation technique to create a dynamic scene, however this would surrender the control of where the balloons landed and so I wouldn't be able to make everything loop as nicely, also it wouldn't really be animation.

So I worked one at a time, bouncing balloons of the walls and onto other walls. To start with I did a 2 balloon loop. This meant having balloon 1 end in the exact same place that balloon 2 started from, and balloon 2 would end where the 1st started. 

However as I went on, this became much more complex to comprehend. As I created looped involving many more balloons, which then bounced off of other in midair, I had to plan much more meticulously where the balloons were going. There is a loop of 4 balloons which go around in the middle in a very complex loop. If you follow 1 of these balloons around, it actually takes 4 loops of the whole clip for it to end in the place it started, however there is also always another balloon starting in that same place on the first frame of every loop. So there are kind of 4 versions of the life of one balloon loop happening at once. This was what was getting very confusing. I was taking many notes while going through this process, trying to keep it all straight in my head.

Eventually, when happy with the animation, I started on rendering. I started with a AIStandardSurface balloon preset, however it wasn't giving me the translucency and light glowing through that I wanted. After a lot of fiddling with more presets and tweaking the settings a lot, I came up with something I was really happy with. It is actually somewhere in between balloon and glass materials, but it gives me a light glow on the walls and floor that I am really happy with.


After 9 computers rendering simultaneously overnight... this was the final result. I am very happy with it, and excited to play with more lighting and materials in my final film project. 

https://vimeo.com/326158442

LoopDeLoop Piggy 2 - Animating (Part 1)

After finishing the pig head balloon model, I moved onto actually creating the animation. I recorded some footage of the hallway at university. Stepping side to side allowed the footage to be easily repeatable or reversed to create the perfect loop. However once I started tracking the footage and cutting takes into loops, I decided against this idea.

It was hard to make a version that was looping nicely as well as the HDRI hadn't come out successfully to create realistic lighting. I think this was due to it being a small space so the stitching process was struggling. This would have worked with a 360° camera and would have given me realistic lighting, however I didn't have access to this tool. 

I instead then decided to create a fully CG hallway within Maya. This made everything look cleaner and also felt more in line with the brief being an 'animation' challenge, it felt weird to have real life background plates. 

So then I finally got onto actually animating, and this was were the real challenge began.

Major Project 11 - Eye Effect Final

While I was waiting for the drone to be finished I moved onto a few other effects that I was going to leave until later. From the tests I did much earlier in the project, I had decided on the way I wanted the bionic eye to look. I created the graphic circles and animated them spinning around. I was mindful to make sure they looked like they were doing something rather than just aimlessly spinning. 

I did many colour tests and after getting some feedback, decided on somewhere between blue and teal. Now this effect is very easily copied to other shots where the bionic eye is visible. However unless it is a closer shot, I won't bother with all the detail, I'll just have the glow and lens flare as you wouldn't be able to see the spinning circles in wide shots anyway. This will be a low priority effect though, I am focusing on everything that tells the story first then coming back to these additional things for the non critical shots. 

Major Project 10 - Lasers and Time Ramps

I have started working on effects that change the length of a shot. This is especially relevant to shots with time ramps or any slow motion. The shot where the Wanderer dodges the laser blast and throws the energy grenade is one where I have made it slow motion only for the middle section of the shot to show him spinning and to focus on the fact that he has pulled out a grenade. I have created the effect as a whole in a 60fps After Effects document. This then allows the effect itself to be at a correct speed throughout, but then I can apply the speed ramp afterwards in Premiere, giving me more flexibility about timings and amount of slow motion. 

The other main shots like this are later on in the big climax moment when he is running through the field. I want this to only go into slow motion when lasers are fired and explosions go off. This should create a very dynamic and exciting look. This example, I will do after I have the drone animation done because I don't know where or when the lasers would be happening. 

Major Project 9 - Starting Drone Animation

Dan sent me the finished drone model so that I could start animating. I got to work straight away, working on the first shots in the film. I am animating chronologically through all the drone shots in order to keep everything clear and simple for me. 

The shots that I've animated are very simple, just the drones flying towards the house. Using the rig and controls set up by Dan has made this very simple to work with. The most complex of the shots I've done so far was when the 2 drones turn to look at the house after the middle drone was destroyed. Their shields then smash together as they charge towards the house. I wanted to really feel the weight of these hitting against each other and slightly bouncing. This accentuates the speed at which they react and turn their shields to protect themselves. I have also made a slight difference in reaction time between the Mrk IV drone and the Mrk II so that the Mrk IV reacts much quicker to show how it is a newer improved model.



I am however going to be putting this drone animation on hold for now. I will wait until Dan has finished the textures just in case there are any tweaks that may cause issues with what I have already animated. Hopefully this shouldn't take long so that I can get back to animating and back on schedule. 

Major Project 8 - Refining the Edit

In an attempt to cut the film down to a more concise running time, I went on a trimming spree. I watched through the edit multiple times, taking frames away from the beginning and ends of shots. 

I remember hearing about this technique in an interview with the directors of Avengers: Infinity War. They learned when they worked in TV that once they were happy with the structure of the film, they would go through and literally just pull out frames to bring down the total run time.

https://www.youtube.com/watch?v=tFuM1j1t7k0 36 minutes in.

Doing this in my edit, was successful in trimming the film down by around 20 seconds, however I still wanted to trim it further. So I went back through again and was much more aggressive with what I was cutting out. I cut shots that lingered for a bit too long, or just could be shorter and still serve the same purpose. I then found myself cutting a couple of lines which I realised did nothing for the story. I could have not had them in the script in the first place because they weren't really necessary if I had have combed through the script more thoroughly. This helped save me a lot more time, and bring the film to just under 10 minutes.

I am now going to go through and rough out any shot which has effects which might dictate the length of a shot, for example when there are time ramps and laser blast where the speed would change how much the shot needs.

Major Project 7 - Editing

The editing process was a fun challenge. This is the biggest project I have ever undertaken, with over 200gb of footage to sort through and pick out the best shots.

Dialogue was something that I was excited about trying out, part of the reason I added so much of it to the script. I think that this was definitely the greatest challenge about the edit. Choosing the best takes that fit with each other took a long time because there were so many takes. I knew when filming that I should take a lot of takes because the opening dialogue scene is very important for setting up the whole film. I could have probably worked a bit quicker on deciding takes as I spent so much time just watching over and over for the tiniest of differences that really didn't make much of a difference. 

There was an interview with the lead editor of Amazon Prime's Counterpart where he would just do his first edit using the last (or second to last) take, then if something isn't working, he runs back down the line of takes until something works. This lets him move a lot quicker and get a first edit quicker because the later takes are usually the best and most refined, it then becomes much easier to tweak from there once there is a rough version to watch. 

https://www.youtube.com/watch?v=7ojMQlP1cR8

This would have been a good way for me to work quicker to get my first rough cut. Instead it took a bit longer than I expected however I am still on schedule. The first rough cut is longer than the animatic was, coming in at around 11 minutes verses the animatic's 7 minute run. This is due to more pauses and ebbs and flows that make everything feel more natural, I didn't account for this in the animatic. I will trim this a little more to make everything a little more concise.