S E T T I N G  |  F I N A L  D E S I G N
​

Testing Projections 

​I knew at the very beginning I wanted there to be a contrast between practical and digital elements of the film to reflect the concept of the virus and radio infecting a real place. Real textures surface mapped onto the CG 'caves' with neon lighting was close but still a bit too digital and stylised but were good points of reference for describing how I wanted the film to look and how I would construct the setting practically. Last semester I done some hologram style tests using cinema 4D and After Effects to get a sense of how the electrical/digital inclusion of the radio transmissions would fit into the tactile cave environment. I borrowed some of Calum's mod roc from his set to test out how the glowing digital animations looked once projected onto realistic textures and suddenly felt like I could finally see the method coming together. 
​

Lit some of the backgrounds which will be used to extend the 'set' of the cave. Really happy with the effect of the projected glowing spiders. Below is an example comped up and edited version of the cave backgrounds which I think works well for this wide shot but is too explicit for the whole film, I want it be close and dark for the whole first section.

​Making the Caves

As most of the shot angles would be worked out at the final stages it made the most sense to focus on the backgrounds and plotting out the action. I knew most of the film would be in darkness and full of close ups/flickers so I started with trying to work out the complicated sections with the most animation first. Which was anything involving moving through cave tunnels and multiple glowing bugs in the same shot. I tested this out using card to figure out how many layers I would need and scanned them into After Effects to experiment with depth and parallax using 3D camera layers. Once knowing this worked, I recreated them bigger and covered with tinfoil and mod roc to get the rocky cave texture. ​I then took images under the rostrum using real light and coloured gels

​

​Testing Projections Again

I tested projecting again once I had animated a bit more to work out if other materials such as tracing paper or polystyrene worked better for the spider projections (as mod roc or heavily texture surfaces distorted the images too much) as much as the these gave a cool ghostly/glowing effect, I decided to stick with mod roc for consistency and make sure that flat-ish sections were used for 'character' animation and textured/layered/bumpy sections would be mapped onto using masks to enhance the area with projected light. 

projection tests on roc and flat textures - glow effects working how I want although some detail is lost during projection which is why the spider silhouette is ideal as it is a recognisable shape. Textured surface is good for mapping onto the surface or basic light effects but might distort some of the animation and effect the 'tunnel' sequences where the spiders need to move through the background so it's best to work this out first. 

I also tried out colour and effects here which was helpful for my workflow as it became apparent that compositing of scenes and glitch effects (which would normally be done in post production) in After Effects would need to be done pre-set up of the practical shot and that projection was the final stage. For this reason I assumed that it would be wise to get all the animation done in complete first and project everything in the last couple of weeks (also due to limited space) however this ended up not working as planned as I needed to test everything and couldn't risk leaving it this long.

For the four spiders crawling towards the camera, I animated one then duplicated it and altered each one slightly differently in After Effects, this was useful for working out how I would do the scenes with multiple spiders.

Couldn’t quite get the mapping to work so I had to rethink
Glitchy birds eye view spider was the first thing I animated - realised shot doesn't work

​
Thinking about the workflow 

'Projections in the forest' works by taking photos from the camera position, animating within a mask and then projecting the animation back onto the surface. 'Mamoon' was created in a similar way by taking photos of the sets and using these as reference for the animation which is projected back onto the surface. Buck's Tinder advert combines practical sets with projected lighting and comped in CG animation. I'm using a combination of these. A mix of textured 'rock' like backgrounds for projection mapping, real imagery, 3D animation and some comping in after effects. 

Attended Move Summit in February where Blue Zoo spoke about their process for 'Mamoon' which involved creating the full set and then using photos of it to block out the animation. I wanted my projections to feel integrated into the practical environment instead of floating on the surface so also researched different projection mapping techniques, which used specific software and masking out sections of photos. I knew I needed to come up with a simple and personalised workflow for my project and do a lot of testing to get the method to work and that this trial and error time would limit time spent animating.

1. Work out logistics of shot
2. Background/set build
3. Take photo from ideal projection position to use for reference
4. Animation/masking
5. Effects (flicker, warp, glitch etc.) in After Effects
6. Project back onto background and film

Changing Shots 

Ended up holding off introducing the swarm of bugs until nearer the end and just following one up until this point. As well as changing the spiders to crawl away from the camera towards the light at the end instead of towards the viewer. This made more sense visually as you're being led through the environment by the bugs.
​
​

A N I M A T I O N  |  S P I D E R S

I had been using a free spider rig for all my tests last semester and at the beginning of this semester. Somewhere along the line I realised I really liked this and wanted to use spiders in a few shots. My film was becoming overly complicated so I decided to use just one type of bug and keep it simple. I had a go at modelling one myself but it just wasn't giving the same effect. So I made the decision to use the free rig and modify it how I wanted. This meant animation would be easier as animating creatures with 8 legs is difficult enough and I wouldn't have to spend ages rigging when it isn't something I am interested in. Through this process I've learned so much about Maya and can feel myself becoming better at animating (something I didn't think was my strong point) ​

I made life easier for myself by choosing spiders and a ready made rig which I just modified and was able to concentrate on the rendering and animating.

Problems

Animating the spiders was a huge learning curve as although I had a bit of experience, I didn't expect to be using it for most of my film. Getting a clean walk cycle was the hardest part and took various tries. A lot of them looked glitchy which was what I was going for and didn't really matter for shots of them quickly scurrying away but I needed them to drag themselves forward and move along a path. 

Couldn't get him to walk
Rig and file corruption (having to copy and past key frames)
Different versions of maya (having to copy and past key frames)

​

Spider Reference 

Tried to animate the spiders without reference for a long time and then could only look at animated or tiny spider reference. I finally sucked it up and looked at proper reference of brown recluse spiders and my walk cycle slowly started to improve and I think I cured my fear in the process. Success.

Vector Rendering & Cave

​To get the spiders looking digital and electrical, I used Maya vector to render the movement as a wireframe png image sequence. Which would then be placed into After Effects and positioned for projection. Rendering usually did not take long as there were no textures or camera moves as I was doing that in camera however I had to render it out from various different angles most of the time. For shots with the spider interacting with the cave (e.g crawling over the edge) I had to render it with the cave, without the cave and then mask out the spider in After Effects along the outline of the cave.

​Messed around with different render settings - ‘hidden edges’ gave it a quality I wanted but shots from behind and infant of the spider didn’t work as it creepily looked through the whole spider

Mike helped to project my textures onto a cave in maya so that I can map out the action for projection for the tunnel sequences where the animation has to interact with the set. The other shots I am able to cheat a bit with and rely on close up dark backgrounds or compositing tricks. A shot towards the end also involves a large amount of spiders populating the scene so I've been experimenting between populating a full shot in maya or comping in after effects and changing the animation slightly/dropping frames - I want the spiders to appear kind of a like a 'hive mind' where they all act similar so realism is not super important. Something that I really dislike in CG animation is that it takes so long to look not too smooth. I wanted more a top motion aesthetic so dropped frames and adding slight shaking. 

​I’ve been doing everything at 24fps and focusing on making the spider animation as legit as possible I had a few people tell me I was being too perfects with it and then when integrated Into the scene you wouldn’t notice if some of the legs were out. But this was the first time I’v properly used Maya and I wanted them to move as close to reality as possible as I felt this added to the creepiness of it.

R A D I O  |  T E X T  |  T I T L E S

​

Like rest of after effects comps my text has ‘glitch’ effects as adjustments layers so I can change the text
I've went back and forth with the radio element quite a bit and decided against using dialogue to spell out what was happening too much and only use it to signpost the beginning and end. I want the radio screen to filter in as the screen fades to black to split up the film into sections and also illustrate the passage of time. The image on the right is my reference for the text, I used cinema 4D to create a 3D 'nixie clock' style clock. I'm happy with how this and the animation is looking independently but unsure of how to best comp it into the film as I don't want it to look out of place or too graphic/mograph.  

​P R O J E C T I O N  |  C O M P O S I T I N G

set up, learnt from tests, projector issues, working it out as you go, spider surface projection, texture projection mapping

Creating the light projections - 

This meant each shot required a lot of ‘testing’ before completing it but getting through shots was quite quick as I had previously finished the animation, tested the positioning and effects and was simply just capturing the frames. I arguably made more progress when I was under the lights and testing and animating rather than trying to formulate shots and sequences and wish I’d spent more time trusting this way of working rather than trying to avoid every possible failure and issue

Normally I or other people would make sure the animation blocking is solid before embellishing but as the mood and design was so important it meant shots had to be comped and flicker effects added in completion before projection first and the only editing would be colour grading and darkening - therefore I ran out of time towards the end and had to compromise on the animation quality for the ‘look’ I knew early on that the realism of the spiders wasn’t really important as I was glitching them, dropping frames, skipping positioning and things but I still wasted quite a lot of time on trying to make the loops/cycles of the spiders as smooth and perfect as a possible. This probably drew from my own unconfidentce in my animations skills. I knew I could pull something together that arguably looked good but was worried about my animation, animating the most Maya I’ve ever done and it narratively making sense.

S P I D E R   P R O J E C T I O N

The projection workflow ended up getting a lot more complicated than I anticipated. For each spider shot I had to animate in Maya, render it out as a vector png image sequence from the ideal angle, importing it into after effects for editing (matching up to scene, masking, glitch effects etc.), rendering it out as a video and projecting. If it did not fit or match I would either reposition the projector (couldn't scale it up bigger or smaller on the projector I have, only adjust the focus) or have to adjust and re-render in after effects. Some sections had up to 13-14 renders/tries and some had to be altered slightly in Maya again (I've included all of these in my digital USB submission). Animating and projecting actually went very quickly, most of the time was taken up working out the shot and getting it to match up correctly. But I suppose this time for other people is taken up during colouring and comping, whereas this method is kind of backwards. Once the sequence is captured on dragon-frame that's it complete except for some light colour grading and masking and any other issues that I have outlined below.

​Some shots were very close to the edge due to the angle and others I needed to mask out areas of the cave in ‘post’ and move it up to match.

​As I had a lot of backgrounds 'done' already and feared the spiders could distort too much for some more complicated shots, tested out trying to comp an early projection test onto a different cave, ignoring the obvious colour issues on the whole scene and just zooming into the section around the spider.

I discovered that small in scale animations distorted and blurred too much during projection (the spider hanging down on a thread), so needed to project big then scale down the footage later. This sometimes caused problems as I was still trying to achieve the 'glowing' and 'loss of detail' effect that you get when you film projections (as seen in the earlier projection tests of the green spiders) as opposed to capturing frame by frame. I found that the scenes of the spider crawling over the rock towards the camera were too detailed that they looked digitally comped. I've considered re-shooting some of these scenes but each one took so long. So I might re-project the original animation on flat and then overlay the glow rather than having to re-set up each shot.

Spider Projection - Crawling over layered surfaces

problem 1 - animation is hard
problem 2 - have to cripple the legs to make it look right from the angle
problem 3 - when rendered without cave I don't know how to get the limbs to wrap around the surface

ended up being more complicated as had to render out each test from after effects, test position and repeat - some shots had around 12 test renders before it worked.

Anything from the front looked a bit weird to me and although it kind of added to the creepy eery-ness I wanted to use focus on like hints of the spiders e.g scurrying past really quickly and the gangliness of the legs - for this shot I wanted the legs to be bigger/slower and more menacing so I cut it shot and and shot the one slow leg bigger

P R O J E C T I O N   M A P P I N G

Projection Mapping Workflow - Bark/Texture Projection

take high quality photo of surface
create mask in photoshop using magic wand tool
import mask into after effects and use as track matte for light animation, distort and glitch
project back onto surface in darkness and capture frame by frame 

For the first test I captured random frames whilst it was projecting to get a 'ghosting' effect however this was way too flickery. Frame by frame meant I could still keep the details of the animation and control it better.

​
Cave Lighting Projection - Learnt from first tests, learnt from bark tests, too much light, does it fit in, felt it lit up too much detail and also wasn't clear what it was, used vector blue to soften the projection and ended up with water then a texture that looks a little like spider webs, will continue to use this for further texture projections


Like with the tests on the black card, I took RAW images of the layers of mod roc separately in order to layer them on multi-plane on 3D layers in After Effects to extend the set and environment. This method worked a bit better than what I previously tried which was just projecting light onto the surface and not actually plotting out the action. However I think there is 'too much' projected light going on for these scenes - the effect needed to be more subtle and less flickery.

Wanted to blur the projections slightly so that they blended in with the more digital looking cave backgrounds using vector blur in After Effects (which accidentally creates a cool water effect) and ended up really happy with the result as it looks a lot more like electricity/spider webs. However I need to be careful as this could look too computer generated which is what I was trying to avoid 
​

Spider Webs - I felt the presence of the spiders had to be built up slowly so that you're kind of anticipating them before they're seen. I wanted the light to pan over the webs during eery/glitchy parts of the sound to emphasise this. The above projected light on the cave edges is supposed to also slightly look like electrical webs but I wanted the initial beginning shots to feel more realistic. They took quite a few tries to get right. The first shots I done looked really good on Dragonframe and in shot but as the lines were so thin it didn't pick up very well and looked too dark amongst the other shots. It kind of just looked out of focus and out of place rather than being used to illuminate the tunnels of the cave like I wanted. I tried again with a more 'wireframe' style web to match the spiders however this looked a bit too digital and clean. I finally managed to get it right in the blue version, just need to do a bit of editing with the loop to integrate it with the other shots.

Plant Texture Projection - Failed, too much light/projection, doesn't fit in and is not really essential or add much

P U T T I N G    E V E R Y T H I N G    T O G E T H E R

---

Soundscape

I created a 'sound scape' style version with some of the projected loops all playing together with the sound. This was helpful in sensing how the mood of everything was matching up. I actually like this more and think it works better than the film.

Final Film

Current working version, still hoping to add in some scenes, re shoot some scenes, colour grade a bit cleaner and add to the audio (rising base track and radio) before degree show and film-house. I don't feel like this version is much of a film at the moment due to it's length as I had to cut out a few scenes to make it work with the soundscape. I am trying not to beat myself up about it and carry on adding the scenes I planned. ​A lot of my work isn't even included in this version, mainly a lot of the backgrounds, as I felt it didn't fit. Although there is not much of a narrative it has to visually make sense that you are travelling coherently through a cave with these spiders and I had to think specifically about line of vision and moving left to right. I probably overthought this in hindsight, but I didn't want to just shove everything into this cut when I didn't feel it was reading well.

I also didn't include the orange radio text animations as I don't have dialogue mixed into the sound yet and it felt out of place. The text element appearing takes you out of the world so I think if I do a version including the radio dialogue then it should simply just be included in the soundscape rather than something we also see visually.

​I think it needs a lot of re-editing as a film. The visuals and sound a definitely the strongest element I just don't think both match at the moment. I think maybe some of the scenes aren't integrated enough. My next stage is to fill in some of the gaps in the film including integrated some more of the backgrounds, re-work the sound and have more resolved ending where the spiders crawl over something that reveals the end of the human world and effects of the virus. This version currently only has two shots of the group of spiders as it took so long to mock up and then ends quite abruptly. So I want to have more of the infestation of the spiders chaotically. 

I plan to add more to the visuals to expand the illusion of the ‘infected world’ and make the build-up to the spiders tenser. I do not currently resolve how I planned (with evidence of humans being wiped out by the virus) so I intend to revisit this before the degree show. I also intend to add radio dialogue to the beginning and end mixed in with the soundscape. However, I made the conscious decision to make sure this version stands on its own and not overcomplicate it before making any changes or adding to it.

​Poster Design & Professional Practice

Initial mock-ups for potential poster design layouts in Photoshop and Illustrator. Considered using some imagery of spiders or evidence of radio disturbance to allude more to the film but decided to go for something simplistic and ambiguous. As the film is more atmosphere and aesthetic driven than narrative it makes more sense to hold back any information that might ruin it or set it up for more than it is and I feel these designs suit me too. I like the middle one the best and will probably add some 'white noise' tv fuzz to the text or background.
​

Reflective Statement

Overall, I'm proud of the work I've accomplished this year, how much I've learned and how hard I've pushed myself to do something a bit different and believe that production is going well. However I don't feel my project is successful at the current moment as a 'film' or at the standard I would have hoped for the hand in (which I intend to achieve by degree show and the Filmhouse screening) and itt has been quite difficult to get my head round narratively and as a production method. 

Despite the theme and concept remaining constant since the beginning of fourth year, I feel my ‘film’ has changed a lot over the course of the semester and year. I decided to stay true to the atmosphere and mood of it instead of trying to push a character driven narrative into it that it did not need. I began by animating spider sequences quite early on as I anticipated these would be the hardest shots to complete (as it required a lot of Maya) during this process I actually realised that the film could work alone with the spiders being ‘the bugs’ and lead the viewer through the environment towards the ‘reveal’ and that I actually really enjoy animating in Maya. This refined my concept more as all the animation is digital projections with physical and ‘real’ backgrounds and adds to the idea of a virus infecting a space.

My workflow has been a little complicated and difficult to pin down which meant there was a lot of time spent testing. As I was working to a locked soundscape, it has taken a while to pin down the exact shots needed to make this make sense as a ‘film’ and I’ve struggled a bit with finding the balance between planning/organising and spontaneously letting the animation unfold as I move through the sound. It feels that it was going to either be a film or a more abstract projection piece and I’ve always wanted it to fall somewhere in between. Overall, I made things quite difficult for myself in terms of technique. However, I believe this has paid off and has encouraged me to progress more with projections in the future.

The final film in its’ current state needs work in my opinion. The complicated workflow set production back a fair bit and although I am pleased with how it’s looking so far, I’ve had to make a shorter, simpler version that is stripped right back. If I could do anything differently in regard to the film, I’d realise that it never needed a fully characterised narrative and focus on making sure the shots were locked before animating to ensure continuity. As although shots needed to be worked out on set, a more finalised animatic would have saved time.
​
I have a bit of a lack of confidence in my own work which I think freezes me up a bit and think I would probably have progressed further at this stage if I trusted what I was doing was good. And think that I maybe lost sight of how it flows together as I was concentrating on making each scene perfect. That being said I think this is the most ambitious and successful work I’ve ever made and is in stand with the goals I set at the beginning of the year to make something a bit dark, a mix of practical and digital techniques. And overall feel it's looking quite professional. I plan to use sequences in my show reel and any job applications and hopefully enter it into festivals.