Concept
Before the semester started, my friends and I discussed doing the group project together. We knew this year was going to be vital for our qualifications so we wanted to get a headstart on it.
Rhys mentioned that he wanted to do 3D animation; if we didn’t have 3D animation in the project he would move to another group, which was fair since that was the area he wanted to go into in the future. We all still wanted to work together so we made sure to incorporate both 3D and 2D into most of the ideas, since most of us preferred 2D animation.
After deciding to work together, we discussed ideas. I came up with several ideas before narrowing them down to one or two I thought could be feasible. Some of my early ideas were as follows:
- A comedic parody of gangster movies starring a group of children playing in their back gardens. The plot would involve the main character swearing loyalty to the kingpin of the underground candy trade in order to avoid being sent to time-out. The film would parody various scenes from mafia and yakuza movies. It would end in a car chase in Little Tikes cars and scooters with the characters shooting at each other with Nerf guns and water pistols. It would have been primarily hand-drawn 2D animation, with some use of 3D for the vehicles and more complicated shots (either using a 2D-style shader or traced over). This idea was fun, but would have been too complicated for our timeframe.
- A surreal horror short about an urban explorer discovering an elevator in an abandoned building that goes down to the centre of the earth. He would have either ended up in a Hell-like landscape, or a surreal alien-like world (based on the conspiracy theory that there’s a utopian land populated by ancient beings below the surface of the earth, Agartha). The short would have used primarily 3D animation with a somewhat realistic style for the first sections (when the main character was exploring the building and finding the elevator) and an overwhelming multimedia style for the earth’s core. It would have been shot in the main character’s POV, as if we were watching footage recovered from a GoPro. This one would have also been a bit complicated and was too 3D-heavy for a group with mostly 2D artists.
- A cute and comedic short titled “Hubble Bubble” about a witch brewing a potion in a giant cauldron. Every time she grabbed a new potion ingredient, we’d see a clip of how she got that ingredient. Each ingredient would be from a different country and the clip would show her interacting with a folklore/mythological character from that region (the artstyle would also change to be reminiscent of the traditional art from the area). At the end we’d see the cauldron start to bubble up with menacing music, before cutting to the witch hopping in the cauldron with her rubber duck and a glass of wine. The title would appear again but extra text would fade in spelling “Hubble Bubblebath”. The short would have primarily used 2D hand drawn animation with some of the cutaway sections utilising other methods (depending on which countries and art styles we chose). This was one of the strongest contenders but overall we preferred Jess’ idea.
- This idea got far along enough for me to work on some concept art. I was inspired by cartoon depictions of witches, my main inspiration being the YouTube animation Witches On Tinder by Worthikids
- A short loosely based on the story and characters of Journey To The West. It would have featured a group of university students based on the original characters going on a “quest” to do something mundane like returning a library book or picking up food. The short would have focused on the comedic potential of the characters and what the archetypes would look like in a realistic, non-magical, modern setting. This idea wasn’t discussed much so we didn’t pick an animation method or pipeline for it.
All of us made PowerPoints with our initial ideas which we shared on a group call.
We discussed them all and made a list of pros and cons for each idea, including discussing the technical aspects of each project and how long they would take. We all voted on our favourite ideas and ended up choosing one of Jess’ ideas.
It was a parody of reality TV, but with a team of ghost hunters going to an old house to capture the ghosts on camera. The ghosts would give dramatic interviews and act petty like the Kardashians. It would have over dramatic reality TV music and sound effects.
Over time this idea changed somewhat. We kept the reality TV parody characters, but we developed more of a plot and added some more emotional scenes. This was because during our pitch presentation, we were told that our characters lacked motivation or a reason for the audience to care about them.
Character Design
At first we weren’t sure what era the ghosts would be from, so I designed a character with a classic “haunted house” aesthetic with Victorian/Edwardian clothing. Since we were taking inspiration from reality TV I thought it would be funny to give her a typical overfilled, over-botoxed look that a lot of reality stars have, but make it fit the time period by saying she died from bee stings that made her face puff up.
Jess later said that she pictured the family being the ghosts of dead 90s/early 2000s celebrities. I liked the exaggerated proportions from my design, so I reworked her to fit Jess’ vision.
I drew inspiration for the new design from several sources. I wanted to give her an overexaggerated glamorous yet shallow look – a Hollywood socialite, reality star, and runway queen, only concerned with her looks and the attention they bring her. I named her Berlin PremierInn as a play on Paris Hilton (Miss PremierInn is married, but she kept her maiden name as it’s part of her brand). Since I was going for an over-the-top look, I drew inspiration from drag queens and their caricatures of feminine glam. Trixie Mattel was one of my main sources, her big hair and overlined lips fit the bill perfectly.
Via Wikipedia: By dvsross – Rupaul Dragcon 2018-103, CC BY 2.0, https://commons.wikimedia.org/w/index.php?curid=75468433
I also looked at female celebrities like Dolly Parton, with her stunning outfits, big hair, and curvaceous figure.
Via nyhistory.org
Berlin’s face was designed to look like she’s had a lot of work done (although she’s OBVIOUSLY all-natural, babe) and I took inspiration from influencers, celebrities, and models with heavy cosmetic surgery to design it. I gave her a tiny nose, big cheeks, and even bigger lips as these are common features of heavily injected faces.
For her outfit, I chose something that was glamorous but not too complicated so it would still be easy to animate. I took inspiration from the Spice Girls, they’re from the right era and I love their fun and cute outfits. I drew a few different designs, but in the end I gave her a simple mini-dress which was girly and glamorous but still easy to draw.
Via Harper’s Bazaar
At first I wanted to give her a handbag chihuahua like Paris Hilton, I thought it would be funny for the dog to be a skeleton to match the ghost theme. I drew this in one of her concept designs, but ultimately decided that this would just make animating her more complicated.
At first we were going to colour the ghosts with fairly standard palettes of skin tones and colourful clothes, but we soon realised that this looked strange with their tails – especially Michelangelo, since he was shirtless. We decided to colour them in a monochromatic palette. Grey or white would have been too bland, so we preferred a ghostly green-blue hue. We also decided to make the ghosts semi-transparent so they would look more ethereal, plus it would help them mesh with the 3D background and not look out of place.
I tried out a few options for the ghosts’ colours. I coloured Berlin in a few variations of the colour scheme we had in mind and put her in front of a dark-toned background, then turned down the opacity like how she would appear in the final film. We chose a palette based on how well it would contrast against a dark background. I also made a 2D art style guide for Dan and Jess to use when designing the other two ghosts to make sure we had a cohesive style.
Originally we thought it would be funny for Berlin’s implants to be visible under her skin, since they’re not actually part of her. I drew a design for this but we ended up not using it as it would have complicated the colouring process. In ghost hunting videos they often talk about “orbs” being visible on camera, little balls of light that are supposed to be spirits present in the room. We wrote a joke about “two large orbs” being visible on the ghost hunter’s camera, we ended up keeping the joke even after scrapping this design element as it could just be taken as a misunderstanding.
I also thought about how she should move when animated, I noticed that people who’ve had Botox and filler injected often don’t move their facial muscles much. I still wanted her to be expressive, but I added a note to her design sheet to say that the outside of her face shouldn’t deform much, but her eyes should still be expressive. I wanted to give the impression that her lips were so big that she couldn’t really move them properly, so I decided that her mouth should move within the outline of the lips – the lips themselves don’t move, but the opening of her mouth does.
I included a height for her in heads and instructed the others to do the same with their reference sheets so we’d know how tall to make each character in relation to each other.
Dan, Jess, and I also did some designs for the ghost hunters. Up until this point, the three of us were able to split up design work evenly with the three ghosts, but since Rhys is more comfortable with 3D work we had to decide which of us would be doing the two additional characters. We thought the fairest solution would be for all of us to come up with our own concepts, then choose our favourites based on external feedback.
I mostly based my designs off two TV characters; Gus from Truth Seekers (played by Nick Frost) and Andy from Detectorists (played by Mackenzie Crook). Truth Seekers is a series about a group of amateur ghost hunters, which was exactly the vibe we were going for. Detectorists is about metal detecting but I took inspiration from the characters carrying around large pieces of equipment and wearing big earphones. It reminded me of the kind of improvised equipment a ghost hunter might have.
Via The Guardian
Via The Guardian
Our tutors advised us to cut one of the ghost hunters since we didn’t really need two on-screen. I scrapped the shorter one and kept the tall and scruffy one as he fit the personality for the ghost hunter that we were envisioning.
I tried out a few colour schemes for him and chose one based on my group’s feedback.
I made a turnaround sheet for him like I did with Berlin and included some expressions and a height reference. I didn’t add any animation notes because Rhys is more knowledgeable about rigging and 3D animating than me, so I thought I’d leave those decisions up to him.
We wanted Frankie to have a t-shirt with a logo for his ghost-hunting “company”, the Paranormal Investigators and Supernatural Seekers (P.I.S.S for short). I doodled a few designs then made the best one from the bunch.
With the model sheets done, I animated a short test clip to be used to test our production pipeline in Krita. Dan coloured it using the provided palette.
Storyboards and Animatic
Jess was in charge of the script since this story was her idea, we reviewed it as a group and began storyboarding once we were happy. At first we split the script into sections for me, Jess, and Dan. We handled a few pages each, then we recorded a reading of the script and Rhys edited it into a rough animatic. After showing this to our tutors, we received some feedback on the script. Jess wrote another few drafts, then I re-did all of the storyboards to fit the new script.
The last two boards were for a post-credits gag about the parent ghosts discussing how they think the “recording session” went, but we ended up cutting it to reduce our workload.
We recorded our voice lines for the new script. We did more takes and put in more effort in this version since we planned on using it for our final audio and didn’t want to do another version, since it would interfere with our timing and lip syncing. We wanted the audio quality to be consistent between the characters, so we agreed to meet in-person to record. At first we looked for recording studios but we couldn’t find an affordable one. Instead we borrowed a microphone from the university and found an empty classroom without much background noise or echo. To dampen the echo even further, we took turns recording our lines with a winter coat over our heads. It worked as a crude version of sound insulation.
I edited all the characters’ audio tracks together and added some free use sound effects where it was needed for timing – for example when the fire ignites. We planned on adding music and more sound effects later.
The new audio was combined with the boards to make a basic animatic.
https://drive.google.com/file/d/15Ktsup701ugy02elhZJuD8MuhOIBGsZc/view?usp=drive_link
This was used to make another version of the animatic, which I then split into clips for each shot.
Organisation
During my placement at JAM Media, one of my roles was as the assistant to the production manager. This meant I had experience with production management, so I was in charge of that aspect of the project.
I based my organisation on what I learned at JAM Media and adapted that to suit a small-scale project. At JAM, we had several large spreadsheets to track the status of every shot in the film. The sheets were split into scenes with each shot in the scene listed in a column alongside who it was assigned to, and columns for each step in the production pipeline (line, ink and paint, shading, ect). There were drop-down lists for each category with options like “in progress”, “for review”, “approved”, and so on. This allowed the production manager to be able to see what stage any given scene was at by looking at its shots in the spreadsheet.
I adapted this to our project by creating several spreadsheets. My first attempt was a spreadsheet for our pre-production assets, I used a simple colour-coding system to show the status of each asset but quickly realised this was a mistake.
One of my group members was colourblind. I knew this and checked to make sure he’d be able to see the difference between the colours by using a colour-blindness simulation site, but it didn’t occur to me that even if he could tell the difference between the colours he wouldn’t know what each of them signified.
He suggested making a key to show which colour meant what status, but I soon realised that it would make more sense to just make a drop-down list with text so we could add more status options (such as “needs revision”).
I learned how to make drop-down options in Google sheets and made a revised spreadsheet. I also made drop-down options for the “assigned” column so we could easily change the spreadsheet if responsibilities changed.
I then started working on the other sheets. For the 3D models, Rhys and Dan made a list of the models we’d need, and Jess sorted them by priority (as in which models were most important to the script and which would need to be finished first). I made a sheet based on this, with the models sorted top-to-bottom by priority. I colour-coded the list to show where each level of priority started, but the colour-blind group member would be working on the highest priority models which were at the top of the list anyways, so he understood.
At first I had separate sheets for each stage of the models’ pipeline, but I ended up combining them all into one spreadsheet with columns for each step.
I also created shared folders in Google Drive based on the file directories I worked in at JAM, with sub-folders for different parts of the production.
It’s important for projects to have consistent naming conventions so assets can be easily found and replaced if needed, so I wrote a naming convention guide for everyone to follow. I included a breakdown of what each part of the name meant, so it would be easy to understand.
We used this naming convention to break our storyboard down into shots, and once those had been counted, I created spreadsheets for each step in the 2D shots’ pipeline. I did the same for the shots with 3D aspects.
Since there would be 3 of us animating, we were a little worried about small inconsistencies in the art style between shots. To avoid this, we decided that the person who designed each character would be the one animating them – I would animate Berlin, Dan would animate Michelangelo, Jess would animate Madison, and Rhys would animate Frankie. We stuck to this as a general rule, but not entirely. Some characters appeared in more shots than others, so some of their shots were assigned to different group members. For example, Madison appears in a lot of shots, so I did some of them rather than Jess.
Animation
After all the preparation, we were ready to animate! Jess sent me a render of the sofa to use as a base to animate over, she included a still frame of Madison to use as height reference. We had included the characters’ heights in heads on their reference sheets, so I knew to make Berlin one head taller than Madison.
I cut the animatic’s audio up into sections for each shot. At first I saved these clips as .mp3 files, but I found out that the version of Krita I was using had problems with that format as the imported audio would stutter or not work at all. To fix this, I converted all the clips to .wav, which worked just fine.
This was my pipeline for rough animation:
- Open the template and rename it according to our naming conventions
- Import the appropriate wav file
- Record myself doing the motions I want to animate
- Add the first few keyframes and send in for feedback
- Add some in-betweens, asking for feedback regularly
- Add the lip sync once we were happy with the timing
I tended to keep my drawings loose and sketchy at this stage, and sometimes went off-model. I found this to be the best method for me personally because it allowed me to do more drawings without becoming burnt out or wasting time on polishing frames that could be erased or replaced based on feedback. I planned on being more accurate and on-model during the inking phase, we had agreed that I’d be doing the ink and paint for Berlin so I didn’t worry too much about my rough frames being easy for another animator to interpret. If someone else on the team was assigned to ink my roughs, I would have been more precise to make it easier for them.
The lips were done on another layer so they could be edited without affecting the rest of the body. When I was done I’d send the shot to the rest of the group and to our tutor for feedback. I’d edit it based on their advice, then move on to the next shot in my list.
As we got further into production, it became clear that we likely wouldn’t get everything done by the hand-in date. We were honest about this with our tutors and were told that this wasn’t uncommon (many other groups were in the same situation). We got some conflicting feedback from the tutors: one told us to finish all of the rough animation and just have one or two fully polished shots as a proof of concept. Another tutor told us to fully finish as much of the film as possible, and cut the rest. We debated this a bit, but we decided to finish all our rough animation before moving on to lining. Jess had worked really hard on the script and we were all enthusiastic about the characters we had created, so we didn’t want to throw all of that away. We looked through our shots and chose two that showed off our art style the best – one shot that contained all 4 characters with some fun acting, and one shot that showed a 2D character interacting directly with a 3D model. With that decided, we could finish up the roughs and move on to the next stage.
When I was lining my shots, I was careful not to go too far off-model. I had the character reference sheets open and had a layer with screenshots of them that I could trace over. I partially traced some of the key poses when a character would turn around (which helped me keep their proportions consistent) and drew the in-betweens based on those poses. I usually traced the head and torso, then filled in the rest of the pose.
If a character only moved slightly between frames, I’d use the lasso tool to move parts of the body (like tilting the head or moving an arm). This kept things more consistent, but I tried not to do this excessively as it can make some movements look a little stiff. Like before, I did the lips on a separate layer. Sometimes a character’s lips would be moving but their body would be static, so having a separate layer made things easier.
The next step was colouring. This was the easiest step, I just colour-picked from my reference sheet and coloured in a new layer under the line art.
I also worked on some of the 2D assets, namely the photos of the ghosts when they were alive, and the big oil painting in their sitting room.
Once these were done, I rendered each shot as a PNG sequence. This allows for transparency and doesn’t compress the quality too much. The original plan was to take the PNG sequences directly into After Effects for comp, but Jess realised that this created far too many files to conveniently upload and download from the Google Drive. We still needed the 2D animation to have transparent backgrounds so we could overlay them with the 3D assets. QuickTime videos allow for transparency, so we took the PNG sequences into After Effects and rendered them out again as .mov files.
After sending these over to Jess, we encountered more issues. Dan’s videos were very out of sync and after a bit of discussion back and forth, we figured out that his After Effects was set to ignore hold frames. Meanwhile my videos were far too fast, I found out that this was because my After Effects was set to interpret PNG sequences as 30fps (while our project was in 24fps). I changed the setting and exported them again, which fixed the issue.
As well as my animation obligations, I had also taken on the role of Production Manager (as can be seen with the spreadsheets, naming conventions, and file management I did). This role also included conflict resolution. When members of the group would disagree on methods, argue with each other, or worry about our timeline, they would usually message me. I tried to figure out these issues by acting as a middleman between group members to come to an agreement on problems. I also delegated some of the roles within the project and tried to do so in a diplomatic manner, basing it on each group member’s strengths, weaknesses, and availability.
These roles often changed based on problems that would pop up – for example if there was suddenly a problem with lighting, someone would have to fix it, meaning that their workload increased and some of their assigned work may have needed to be reassigned to keep the workloads balanced.
Posters
We needed a logo for our poster and for the intro for the film, which I handled. I based the font of the “keeping up with” part on the font used for Keeping Up With The Kardashians. For the “apparitions” part, I wanted to make it look like a blood splatter/blood-soaked font to incorporate the supernatural element. I did a few variations with different shading and we picked one for the film.
In preparation for the end of year show, we came up with some poster concepts. I sketched out a few ideas, including these:
Berlin’s manicured nails dragging down like monster claws, tearing the paper.
A parody of The Exorcist’s poster with the silhouettes of the ghosts in the window.
Via Amazon
A poster that showed off all of our characters
The other group members came up with some poster ideas too and we picked our favourite from the bunch.
We decided on an idea that could merge the 3D and 2D elements of our project without adding too much to our workload. Jess came up with the idea of a poster that looked like a conspiracy theorist’s corkboard, with photos of the characters connected with a red string. We’d be able to re-use some 2D assets from the film on the poster; I edited my drawings of the 2D characters to look like magazine cut-outs. Rhys was able to make a few simple renders of the Frankie rig to add too, and we used the logo design I’d made earlier. Jess modelled and textured a simple cork board design to match the dimensions needed for the poster, and added our images as textures onto pieces of paper she’d modelled. She added the string, then added lighting and rendered it out to be printed. Her design related to our project, was unique, showed off all our characters, and didn’t take too much time away from animating.
Reflection
If we were to do this project again I think we would cut down on the 3D assets a little, plus leave more time at the end for rendering and comp. We also encountered some technical issues along the way, but now we know how to fix those issues if they occur in future projects. Another big issue was communication, I think my conflict resolution skills were somewhat lacking. If we redid this project I’d be clearer and stricter about delegating tasks so there’s no arguments about who’s doing what. I also think we tried too hard to follow everyone’s feedback, even if it was contradictory or not what we wanted for the project. In 1st and 2nd year you’re expected to follow any feedback your tutor gives you, but we didn’t realise that it’s different in final year. If we redid this, I think we would be more selective about which feedback we act on.
Overall, I think this project went OK. We were a little overly ambitious, but we were honest about that with the tutors and were able to negotiate what we needed to have ready for hand-in. I’m quite proud of my organisational work and I think this has been a valuable experience for my future career: I’m thinking of going into production management. The film also looks close to how we’d envisioned it, I’m happy with how my 2D animation looks and I think we were able to pull off the hybrid style to a decent enough degree. The dialogue and character acting got some laughs out of the people we showed it to, so I think the film serves its purpose as a comedy/drama. I got quite attached to the characters during production, and I hope the viewers will like them too!