After completing a composited car for my last personal project I decided that I would like to try something totally different during this module. I decided on the Idea of sculpting a stylised head sculpt. I started by researching different styles and ways others have approached creating a stylised head sculpt.
The different kinds of sculpts included:
Sculpts that were very stylised but also had photorealistic traits such as pores on the skin, very detailed eyes and realistic hair grooms.
Sculpts which focused more on being cartoony with sculpted hair and less realistic features.
Exaggerated features/ caricature head sculpt.
Danny Macs workflow
I particularly enjoyed watching this tutorial and also found it rather helpful as Danny runs through the pipeline of how he creates his own characters.
The features that I am interested in focusing on are:
Big detailed eyes
Realistic hair groom (xgen)
Blend shapes allowing different facial expressions
Different versions/ variations of the sculpt etc mesh hair piece/ painted texture/ vs skin normal maps and xgen eyebrows
The software that I am considering using is: Zbrush, Maya and xgen plugin
To start of the process of creating a head sculpt I decided to do a rough concept based on the head sculpts that I have been looking at. I decided that I would like to create a female head sculpt. I focused on having the eyes as the main focus point with the other feature raining stylised but smaller.
I then started to to sculpt out a basic head shape from a base sphere in Zbrush. I used this series of tutroials as Zbrush was relatively new to me when I started my head sculpt. It explained and worked through a useful workflow demonstrating the functions of brushes which worked well for creating a head sculpt.
The screenshot below shows the basic human head with eyes and the main facial features added. I used the masking tool to drag out the ears and nose. I used the the inflatable tool but in reverse to create the holes for the eyes, nose and mouth. Other bushes that I found useful where the dam standard, pinch, smooth, inflate and the move tool. The eyes were just spheres resized to fit within the eye socket.
I started to add some more detail to the features. I also corrected the anatomy so the head looked less alien.
I used the dam standard brush to cut and define the brow, nose, lips and ears.
Side view of the head sculpt.
This screenshot shows that the face and lowers features were to squashed and close together especially the mouth so I spaced them out using the move tool.
This screenshot shows the corrected version with better anatomy.
I looked at reference images as I wanted my head sculpt to have similar proportions to a traditional Female head sculpt. For example a small nose which is quite low on the face, large eyes and a defined straight jaw below the ear.
I used these images as reference:
While online I came across this video what shows snippets of the design process etc while creating characters from inside out.
I also sent images of my head sculpt to our Lecturer Mike who takes Life Drawing. He gave me some feedback on the anatomy of the face. The feedback consisted of the lower face being to pointed and the lips, mouth needing there general shape corrected. He also had concerns about the eyes being to large which would cause the blink to be distracting, if I went on to animate the character.
After I had textured the main head I created hair and a t-shirt. Again I used a sphere which I wrapped around the head using the move tool. I then added detail into the model using the dam standard brush. For the t-shirt I used the extract tool by selecting the area I wanted to create a mesh for then selected the width. I then created a shirt automatically based on the settings I had chose.
After completing all the meshes I needed in Zbrush I used the z remesher tool to reduce the poly count while trying to try and maintain a good topology. I then projected the detail back onto the low mesh to try and recover any detail I lost.
Texturing /UV mapping
To texture I used the polypaint feature built into Zbrush. I started by adding a base skin colour which I then developed by adding areas of blush, freckles and eyeliner. I used the mask tool to create the circles on the eyes I then painted in some detail. I found poly paint easier to use than painting onto a normal flat uv in photoshop.
After creating the textures I created uv maps in Zbrush using the uv master plugin. This allowed me to draw onto the mesh areas I wanted to protect from the seem and others areas I wanted the seem to be placed. This method allowed me to create a single uv map that I didn’t have to stitch together. I was then able to easily export my polypaint texture as png. I then took the texture file and added more detail in Photoshop.
Exporting from Zbrush
After I had finished using Zbrush I exported my models as objs and imported them into Maya. I lined up all the separate pieces of the mesh.
I added my textures through the ai standard preset which allow me to play around with the settings for example the reflectivity of the eyes.
For the t shirt I found a fabric texture which I placed over my texture in Photoshop. I also made a normal map using crazy bump to give the fabric depth. I then plugged the normal map in through the Maya bump map section.
For the hair I added the ai physical hair shader which allowed me to add root and tip colours etc.
Overall I liked the way this render turned out but I would like to develop my hair and skin texture trying to achieve a more realistic character with an actual hair groom and skin texture.
I wanted to experiment with xgen hair. I watched this tutorial which gave me the idea of using my polygon hair as a guide for the xgen hair.
I drew out curves around my hair which I was going to use as guides for the xgen.
I imported the curves into Maya and connected them to the scalp to be used as xgen guides. I selected the appropriate settings but the hair wouldn’t render out following the guides. I tried numerous different settings. I also re imported the curves multiple times but the xgen hair wouldn’t work using my curves.
I looked into creating hair using curves without xgen allowing Arnold to render out the curves but I also couldn’t get this method to work well.
This link covers the process of using curves:
I watched several more tutorials on creating xgen hair without using curves as guides.
I followed this tutorial which worked and allowed me to render xgen hair. I then used the xgen brushes to groom my hair. I added a hair part by freezing one side of the hair and brush it to one side. I also added noise which creates more realistic hair groom. The clumping brush also came in handy as it allowed me to create some clumps of hair for example the strands at the front of the groom.
This was my first result after using the xgen brushes. I had some trouble with the hair colliding with the ears but tried to work around the issue playing around the collide distance sliders.
Freezing the hair to create a part:
I added a cows flick to make the hair styled slightly more.
Test render result: After some playing around with the xgen brushes I started to get the effect that I was aiming for.
I also went and added eye brows and eye lash xgen descriptions as seen below:
The xgen eye lashes clip through the eyeball which I need to fix. They are also exaggerated but I might reduce the effect slightly as you struggle to see the eyelid.
As well as working on xgen hair I wanted to improve the skin texture by adding a normal map and sub surface scattering.
I painted out the skin texture on parts of the face I didn’t want it covering. This is a close up render of the skin.
I also added subsurface scattering which really improved the overall result of the skin.
I played around with the xgen hair and tweaked the skin texture. Here is a test render:
I asked my Lecturer Alec about how to retain the texture detail as when I applied sub surface scattering the texture faded. Also I decided that the normal added to many wrinkles to the face, making the character look older than I wanted. I used the the ai standard skin preset and then re applied sub surface scattering with more realistic values. I then added more freckles and skin details to the texture. I plugged the texture and normal map in through the subsurface scattering tab.
Final Model render:
I decided that I wanted to create some facial expression so I would have some variation for my renders. I used the new blend improved blend shape editor which allowed me to alter the expressions using the sculpting tools without creating duplicates of the model.
Toon Shader Experimentation
During class I saw other groups using the recently released ai toon shader. I decided to give it a go as I have always appreciated the toon shaders in games like Borderlands.
During this semester I have been working on a team-based Studio Guided Project.
My group has been assigned to a Studio called Mashmob were we have been interacting with them in a studio/client like relationship, were they set the brief of creating a trailer for a feature length script. During the process my team has taken the brief and went on to ideate, research, concept and develop over the semester based on the guidance of the studio.
To start the project off we were given Mashmob’s feature length script which we had to shorten down into a 2 minute trailer. We also were told to start some concept art, look at different animation styles, local artists and put together some mood boards to present in our next meeting.
We worked together as a team on a a first draft of the script trying to capture all the main scenes that we thought would work well in a trailer. We got feedback on the script saying it was too long and would show to much of the story.
We researched and collected relevant images covering style, culture and location for our mood boards.
I completed some concepts of what I imaged certain scenes. We particularly liked the idea of adding in ink transitions during the dream sequence. I completed a test in after effects below.
Early Lighting test
Ink transition in After Effects.
We also continued on developing the script based on the feedback we where getting from Mashmob. We drafted 4 scripts in total. We decided to take the approach of making a longer trailer and then two different versions of a shorter teaser trailer.
The general story Includes themes of domestic abuse, alcohol, drugs depression and the Banshee who the main character Alan has nightmares and flashbacks off.
We decided on a shorter 1 min script focusing on the main character’s interaction/relationship with his doctor/psychiatrist while discussing his dream/flashbacks.
SHE FAERIE (DRAFT 4) Final
SCENE 1: INT.
Shot of Doctor’s office, intense lighting, aerial angle, slow, drifting camera pan
Alan and the Doctor are sitting at the desk, in a dimly lit office. Blinds over the window, the only source of illumination in the room.
Alan is tired, slumped posture, but still tense and nervous, fiddling with the ring(?).
Doctor is clinical and professional, detached but not wholly unsympathetic.
Cut to Doctor.
Do you know what triggers these attacks?
Cut to Alan at the desk, who remains silent for a moment. Grim expression on his face. Starts to look down.
(Shaky breath)’I don’t know, it’s just hard to, sometimes… not be…angry’
Cut to POV shot of the ring in his hand. He turns it over and hears echoing, distorted screams.
Cut to a close up of Alan’s profile, intense lighting.
‘It always ends with the same result, a dream…no…a nightmare’
Liquid starts dripping down the walls engulfing the room as Alan’s breathing quickens. Bracing himself, his hand slides along the table, hitting a container of tablets. As they fall slowly, the darkness continues to grow, quickly warping the background. Time speeds up as the sound of pills hitting a forest floor is heard.
We see Alan running, stumbling (this is a recurring shot).
(In no particular order)
Emma and Alan kiss under the cranes, flame burns through the image revealing Stephanie wielding a burning branch. Conor wails. A phone shows no reception, Emma’s back arches off the ground. The granny swings at Alan, terror in her eyes. Alan swings at the camera, knocking the camera round to an empty shot with only the ring, a splash of blood stains the scene from the direction the camera swung from, partially covering the ring. Alan clutches his throat as his eyes roll back. Stephanie stumbles backwards tripping over alcohol bottles. We see a silhouette fall from a cliff, Belfast’s skyline in the background. The figure dissolves into smoke upon impact, only to rise back up into a think skeletal form, which reaches towards the camera and raises its arms, they start to come to together, quickly-
-Sudden cut to a close up of the blinds being opened (in the doctor’s office.) The sound the blinds make when opened coincide with when the banshee’s clap would be heard.
Cut to the doctor, standing next to the window, hand on the cord, flooding the office with light, looking down at Alan at the desk.
That’s enough for today, Alan.
Cut back to Alan, squinting in the light and visibly agitated, who takes a puff on his inhaler and nods. We see the shadow of the crow in the window.
Cut to the view of desk and window, camera slowly zooms in. We see Alan rising to leave and walking past the camera. On the windowsill, outside the glass, a raven caws and flies away. It does a figure of eight before coming directly towards the camera. The wind rushing through the feathers sounds like a muffled scream. Cut to black as the Raven ‘collides’ with the camera.
After a while of working on the script we decided to go with the shorter teaser like script. We thought this would be more achievable in the time frame and with our skill sets.
We then started to work on a rough story board based on the script to get and idea of the scenes and early camera angles etc.
After completing the first storyboard. We went away and created a digital storyboard/animatic. I also made some rough models and did some lighting tests to try capture the atmosphere I imagined.
I also researched some potential lighting and modelling examples:
I edited together the animatic and focused mainly on the office scene shots and camera.
Animatic with sounds
We decided that we were going to use a 3D scene for the office sequence and then transition into a mixture of 2D and 3D elements for the dream sequence.
Molly and Charlotte are going to focus on the dream sequence while Rose and I will focus on the 3D scenes.
The main characters in the trailer will be Alan and the Doctor. The others characters will be Emma, the granny, the banshee and Alan’s dad. They mostly will be created in 2D
We researched for some reference images that we could base the designs on.
Rose then and drew up some concepts for the Doctor and Alan.
Rose decided to take the doctor to model and I focused on Alan.
I decided to model Alan in Zbrush I started with a sphere and used various tools including the dam standard brush, move tool, smooth tool and inflate brush to get the rough shape and details of a male face.
For Alan we had decided that we wanted him to have a heavy brow with sunken eyes. I decided that I would like to make him quite tired looking due to the fact that he has been having nightmares.
This was my first result after sculpting for a while:
After feedback I went and exaggerated Alan’s features even more and changed his face shape slightly. I also used the built in poly paint feature. I used built up base tones for the face then added darker purples and blue round the yes and forehead.
I also sculpted a hair piece starting with a sphere which I then wrapped around the head. I used the dam standard brush to add details. For the eyes I just used spheres which I also textured with poly paint.
Updated sculpt with textures:
After finishing the head I moved onto the rest of the body.
During the trailer Alan will be sitting the whole time with his legs under a table. I decided to put most of the detail into the upper body as this was the part of the model which will be seen.
I sculpted the rest of his body with his legs mainly being stumps as they won’t be seen during the animation.
We had decided early on that Alan would be quite muscular.
I added a t-shirt by using the extract tool in zbrush.
Retopology and projecting detail
After finishing the sculpt I used the z remesher tool to draw guides round parts of the model which needed to have a good topology for animation. I then decimated the model to reduce the polygons while using the guides to try and maintain a good topology.
Once I had my low poly model I used the project tool in Zbrush to project detail back onto the low poly model.
To UV map the model I used the uv master tool with Zbrush which allowed me to draw on were I wanted the seems.
I then exported my high poly polypaint texture using this uv map.
After finishing the sculpt I exported it from Zbrush as an obj and assembled the pieces in Maya. As it was my first time properly rigging I watched some tutorials.
I found the tutorials very useful as a guide on the general workflow when it came to rigging a character. I added a basic skeleton with joints and then bonded the skeleton with the skin. I then used a delta mush option which did a good job for an automatic process. I then added to the skin weights using the skin weight tool. I used the hammer tool and influence joint feature to try and ix the skin weights. I had quite a lot of trouble at first especially with the arm anatomy. Due to time restraints I couldn’t go back and fix the model in zbrush so I tried to add length and fix the arms in Maya.
After completing the skin weights I moved on to adding controllers for the limbs, neck eyes and hands. I also fixed any new skin weight problems I encountered during the process.
While I was ordering the hierarchy I came across some issues which temporally broke the rig. After some research and help from my lecturer and class mates I managed to fix it.
Issues I encountered.
Fixed and completed rig:
After rigging the the main body for Alan I created blend shapes with numerous facial expressions and lip sinking mouth movements.
After finishing the character I started to work on the 3D environments that we needed for the 3D section but also the 2D section.
I started to work on the Doctors office by mocking up the general layout and design of the room. I created some assets such as the desk, blinds and medical equipment hung on the walls. My teams mates also made some assets for example books and filling cabinets which then went and placed around the scene.
I added in the models for reference and scale.
Molly sent me a rough blocked out sky line for one of the 2D backdrops. I decided to use previous assets from my floating city project. I built an outside scene to be used as a back drop for the office.
I also created a backdrop using these assets for the 2D kissing scene.
I then went on to create a cliff and forest floor for the 2D character to run through.
I created the the cliff by extruding faces in Maya. I then added a rock texture, normal map, specular map and a displacement map I made in crazy bump. For the forest floor I used crazy bump again to create maps to plug into Maya. I then used the sculpting tools to move the forest floor mesh around the cliff.
I also started to work on the previs updating it as work was being produced. I edited it together and added the sounds Rose provided. We also got Ryan, Mike, Lauren and Charlotte’s sister to record voice lines for the trailer.
For the 3D doctors office the lighting was important as I wanted it to feel atmospheric and kind of creepy.
I started to light the doctor scene by adding in two main back lights behind the outside skyline. This created a gradient of light. I then added in an atmosphere via the Arnold settings. I then created a spot light focusing on the office windows to create light rays inside the office.
These were my original lighting tests for our previs. I used them as a reference.
Lighting setup in image below:
I created lights in the room at the windows and at the sides to give soft reflections of the walls.
I then setup the scene and got some test renders.
During our meeting with mash Mob Jude said that he liked the previous purple and blue colours as he thought it was more atmospheric.
I then went back and changed the lights based on the feedback
I also then went and created a similar light setup for the Kissing scene for the 2D scene.
Dream Sequence Scenes
For the banshee scene Molly was creating a 2D animated character. I suggested the idea of creating a 3D element using particle effects in Maya. Concept by Molly
This is the tornado swirl effect I created in Maya. I got feedback on the swirl that I could render it out as a 2D plane but it wasn’t a priority.
I also helped Charlotte out with the ring scene by adding a ripple deformer to create a realistic splash effect when the ring hits the blood. I added in some lights, the bottle and a camera movement to try and make the scene more visually interesting.
Ring ripple test
I also added in some lights for the abuse scene Charlotte was working on.
Rose also struggled to completely finish her rig. She had used an automatic rig which had some missing controls and the skin weights weren’t the best. I tried to fix and improve the rig by adding in skin weights and some controllers. I didn’t have time to add blend shapes. It also had a scaling issue which I managed to work around. After I was finished with the rig I placed it in the scene ready to be animated.
I started to animate Alan for the first Doctor scene. I got feedback from Alec numerous times on how I could improve the animation and make it feel more natural. The main piece of animation is the dialogue scene as it involves lip syncing and facial expressions. due to the problem with my characters arms the movement was a bit limited but I tried to work around the issue.
Bottle Fall Test.
I used bullet dynamics to create the falling bottle. I made the table and the floor tiles colliders. I was happy with the general falling motion with this scene but I would like the bottle to fall closer to the the camera.
I edited together another previs making it feel more like a trailer. I added a sound track that I thought suited the trailer. I also added any new work that my team mates provided. Jude also had asked for the title screen and approved screen to be placed at the start to make it feel more like a trailer.
I continued to edit and update the previs. I got feedback from Jude to try and time the music to the changing shots.
Rendering/editing and sound
After all the forest backgrounds had been lined up by Molly she sent me the forest cliff file with a camera. I rendered out the forest scene and then Molly added the 2D animation to the background renders. I setup the renders for the 3D scene and set them to render while Charlotte completed the animation for the doctor’s office scene at the end.
Molly added some FX and I then edited the scenes together. I added some effects for example a fade to the forest in the bottle drop scene and lighting strikes during there running scene. I also colour corrected some scenes to try and maintain the purple colour scene we were going for.
This is the final edit with the renders completed.
After Effects Edit:
During this semester we have been set a personal project. I have decided to look at Compositing and and VFX.
Add 3D elements to a photo or live action background.
Footage can be provided.
- A shot of live action footage or photo with composited 3D art..
- A break down render demonstrating the composited layers and/or post effects
For my project I can consider to look any of the following.
- Scene Continuity *
- 2D & 3D Tracking
- Matching Colours *
- Matching Lighting *
- Contact Lighting *
- Light-Wrap Techniques
- Edge Blend Techniques
- Working with Shadows
- Grain Management
- Lens Distortion & Aberrations
- Match Move
I was interested in the class and talk given by Victor Perez at University. “Victor is a Spanish film director, producer, screenwriter and visual effects artist with more than 20 years of combined experience, and has worked on a number of Hollywood films, including Christopher Nolan’s The Dark Knight Rises, Rogue One: A Star Wars Story and Harry Potter and the Deathly Hallows. Victor started his career in the film industry as a photographer and digital compositing artist.
During the class Victor went over the basics of compositing showing us some Star Wars compositing and a demonstration with a rock character using the software Nuke. He talked about the physics behind light and how taking pictures was a very important part of becoming a compositor.
In the industry this role or job generally is the specialism in lighting, visual effects and compositing. The role would involve production work such as working on 3D models, VFX, lighting and compositing for studio projects. They are responsible for creating the final image by combining layers of previously created material from different departments within a studio. Although it is primarily a 2D role within the 3D world of CGI and VFX, Compositors need a well-rounded understanding of the CG process and would need to know a range of artists skills.
The main software for compositing in the industry is Nuke. I am going to use Maya, the Arnold renderer and then later experiment with Nuke importing my render passes.
Arnold provides the following passes:
direct_diffuse + indirect_diffuse + direct_specular + indirect_specular + refraction + deep_scatter + mid_scatter + shallow_scatter + primary_specular + secondary_specular = Beauty
Each pass contains an important part of an image for example Specular indirect contains reflections. A beauty consists of all the passes merged into one image
Ideas and Inspiration
At first I wanted to composite a King Kong model over Belfast during a sunset inspired by the recent Kong Skull Island poster. Then composite more helicopters over it etc.
I had quite ambitious ideas at the start but sadly due to my lack of skills and understanding of compositing and certain functions within the different software packages I had to keep my idea pretty grounded. My lecturer Alec told me to focus on a still image using AI shadow matte in Arnold.
I moved on to focus on more basic ideas. My main idea was composite the model of a car not a photograph. Many of the car advertisements are actually made using CGI and the car actually isn’t even real. I wanted to recreate a still of a realistic looking car travelling round a scenic road.
This article goes over how some companies make their Car advertisements using a special rig: https://jalopnik.com/how-they-shoot-a-car-commercial-without-the-actual-car-1782499530
Here are some examples of rendered cars:
Due to time a wasn’t going to be able to model a high quality realistic car model so I had to research and try and find a free model.
I recently watched the Audi R8 advertisement and decided that I wanted to use this us this car. If I could find a high quality model for free
I went onto Turbosquid a well known model website and found a model of an Audi R8.
This is the model I downloaded off Turbosquid.
I also looked researched images on google of roads with trees. I Also through my own camera role and picked out a few which I thought might be suitable.
I particularly liked the Autumnal images due to the extra colour present.
Introduction to Shadow Matte
When I was discussing compositing with my lecturer Alec, he told me to look at Shadow Matte within Arnold (Maya) and start to try and create basic compositing pieces using this feature.
“The Shadow Matte is a specific shader, used typically on floor planes to ‘catch’ shadows from lighting within the scene. It is useful for integrating a rendered object onto a photographic background. You can also render out shadows separately for use in a compositing package”
I looked up some tutorials and started to experiment with the shader on some models I had downloaded from Turbo-squid.
The general setup of a scene consisted of making an environment within a camera by selecting an image plane. I then placed the grid so lined up with perspective of the image. I then placed a plane on the grid and imported the Audi R8 model resizing and positing it until it looked right. I then placed some basic lights and applied Arnold’s Shadow Matte shader.
Here are some practice renders using the Shadow Matte shader:
I placed lights trying to match the light coming from the image. The textures on the car were not all working so I tried to re texture it using ai standard. This allowed me to work with some reflections while learning how to place lights and match the light from the image.
Experimenting / Test compositing
I then took the image and edited it in Photoshop and Lightroom:
This was my result from my first test composite. I added rain and I edited the colours.
I also went on to try and practice using my own photograph of The North coast and composite a dinosaur creating havoc on the beach. I used the same techniques practicing, trying to get the light to look real and believable.
T rex and Car composted. I turned the car and placed its door beside the reck to make it look more realistic.
I then went on to edit footsteps and smoke over the composite.
Final Scene Setup
I completed the same steps by creating an environment using an image as a background. I have chosen this image as my background for my final composite.
I re textured the Audi R8 model using ai standard shaders for example car paint etc.
The lights that I will need will be:
a sky-dome with an HDRI image of trees acting as a light source. It also will provide reflections.
I will have a general directional light providing shadows and reflections.
I will have 3 area lights providing light from above and behind replicating the general light within the image.
I also will have projections maps providing light projecting on the colours of the leaves as reflections. I will have 3 projection planes.
“Places the texture on a planar surface and projects it onto the object. Places the texture inside a sphere and projects it onto the object. … Defines the projection surface as a box. Maya places images on each plane and projects them onto the object.”
I used Shadow Matte on a plane to capture my shadows.
This is the HDRI map that I used. It has misty lighting which I need and it will provide reflections of branches onto the trees.
Projection maps and scene setup
Render Passes (AOVs)
After my scene was setup and I had completed some corrections I rendered out my passes using Arnold and then saved them as an EXR file. I then went on to import it into Nuke
Spillting up the EXR into the different render passes.
After Importing the EXR file I made shuffle nodes and applied different Render passes to them. I connected the passes with dot nodes and then went on to use plus nodes to combine them. I had problems importing the background image (backplate image) as it kept displaying as the wrong size. This therefore meant that the other render passes didn’t match up with the image. I messaged Alec who told me to add a reformat node to fix the image size. This worked but then the final image still wasn’t displaying correctly as the shadow and car displayed brighter than the original render.
Problem in image below:
Script of nodes:
I attached the view node to the shuffle node before the backplate and the composite seemed to display correctly.
Final Original Composite
This is a video of the render passes. I ran out of time so I don’t have break down video and I also was having problems in Nuke.
While researching local animation studios for placement, mainly based in Belfast I kept in mind that I wanted to use the placement year as an opportunity to work in a number of different mediums. This would help me gain experience and understanding on the aspects of animation we don’t cover on our University course. I am particular interested in working in 3D and hope to find a placement which will give me the opportunity to improve my 3D skills. In school, I secured work experience with JAM Media, Belfast. During the week of experience, I gained an insight into how the studio operated and the techniques they used. I also was given the opportunity to explore some Adobe products and create some simple 2D animations. After experiencing how a studio approached 2D animation projects I would like to find a studio which produces 3D work.
During class I got talking to a Masters student who started talking about a Studio called Streetmonkey who created work for the BBC etc. This sparked an interest as I went on to research the studio. While researching I found out that Streetmonkey is a production and post production company in the heart of Belfast’s Cathedral Quarter. They have been producing award-winning animation, motion graphics, VFX and live action for TV, film and online since 2002. I went through their portfolio and show reels and saw some work which I immediately recognised from TV. The BBC NI duck sting (Ident) and the BBC Pro 12 Rugby animated advert. I also was interested in their live action work and their VFX and compositing. I have chosen to write my cover letter to Streetmonkey. After viewing their show reels and discovering that they make a considerable amount of content for BBC NI and other well know companies that operate in Northern Ireland. I would love the opportunity to learn new skills working in a range of mediums while helping the animation industry grow in Northern Ireland.
Role that interests me
A role that currently interests me in the industry is Compositing and VFX. This role or job generally is the specialism in lighting, visual effects and compositing. The role would involve production work such as working on 3D models, VFX, lighting and compositing for studio projects. They are responsible for creating the final image by combining layers of previously created material from different departments within a studio. Although it is primarily a 2D role within the 3D world of CGI and VFX, Compositors need a well-rounded understanding of the CG process and would need to know a range of artists skills.
I have chosen compositing for my personal project as I am particularly interested in lighting and the layering of images. So far, I have been using Arnold shadow matte to composite models such as cars and dinosaurs into Maya environment scenes. I plan on starting to try and learn how to use the software Nuke which is industry level compositing software.
Animation Industry talks
The Animation Industry talks on Friday the 8th of December were very interesting and allowed us to see the range of potentially different jobs within the industry. I was interested in the class and talk given by Victor Perez. “Victor is a Spanish film director, producer, screenwriter and visual effects artist with more than 20 years of combined experience, and has worked on a number of Hollywood films, including Christopher Nolan’s The Dark Knight Rises, Rogue One: A Star Wars Story and Harry Potter and the Deathly Hallows. Victor started his career in the film industry as a photographer and digital compositing artist. He then works on collaborations as 2D technical director and consultant where he developed a combination of both artistic and technical understanding of film postproduction and visual effects from a photorealistic point of view.” During the class Victor went over the basics of compositing showing us some Star Wars Compositing and a demonstration with a rock character using the software Nuke. He talked about the physics behind light and how taking pictures was a very important part of becoming a compositor. He seemed very down to earth and his class was helpful. During his talk Victor showed us his latest short film Echo. He went on to the describe the process in which it was made using pioneering, never-seen-before motion control technology and was shot in just five long takes. There were many other talks but I found Victors the most helpful as I am currently looking at compositing for my personal Project.
So our next assignment after finishing the 3D environment project is to make a 30 second short based around the theme of bravery. We have been placed in groups by our Lecturer. I am working with James, Michael and Lauren.
Initial Ideas based around the theme of Bravery.
These are the ideas we came up with during our initial brainstorming session:
Tough guy gets beat up by tiny spider. As it is spider season and lots of people are scared of spiders we thought it would be funny if instead of the spider getting squished. The spider would get his own back on the humans.
Tiny character ends up being the tough guy. Developing on the spider idea. We discussed how the character who you would least expect would end up being the tough guy and save the day etc.
Goat and Chupacabra. Another idea that we had was about a mountain goat and a creature called Chupacabra.
Mountain Comes alive. The mountain could come alive
Decided Idea- Goat and Chupacabra
The Story Idea we decided to center around is about a goat character and their run in with the mythical ‘Chupacabra’ a creature said to hunt them. (Translates as goat sucker) We like the idea of the goat trying to stand up to the creature and possibly killing it accidentally.
It started around the idea of certain breeds of goat have a genetic problem that they faint when startled, we liked the idea that the mountain goat would just freeze up and fall over at the sight of danger. We like the idea that the goat would be brave but because of its natural behaviour couldn’t help but fainting.
The Chupacabra will be more sinister coming to eat the goat but may be in for a surprise when/if he gets shown up by this small goat.
At this stage we briefly looked into the myth of the Chupacabra:
- Originated in Puerto Rico in the mid-90s
- Attacked and killed local livestock
- Has been reported across Mexico and the southern U.S
- Often described as a canine or reptile like creature
We might design our own take on the chupacabra, maybe using elements from the reports. The Chupacabra most likely is some for of wild dog but reports and imagery describe the Chupacabra as being dog like and even as far as being alien. (Wide range of different description.)
Here are some images of different versions of Chupacabras:
We have been looking at Maurice Noble for background ideas.
- Maurice worked for Warner brothers in the 1950s.
- Known for his background art he produced for Warner Brothers and MGM
- He worked on the original Duck Dodgers cartoon as well as What’s Opera, doc and the Road runner series.
- His style fits with the one we are going for.
Walt Peregoy: Scooby Doo
The background art for the old Scooby Doo series has been another style we have been considering.
- Created by Walt Peregoy for Hannah Barbera
- The painted style and dark colours would make a good base for our environment
- The shows monsters would also tie into our theme
We are also going to look into Looney Tunes for comedy references and staging as it is very close to what we are going for. For example when the Coyote holds up the sign.
We worked as team to come up with a rough storyboard Jame then took it and sketched it up for our first animatic.
Since our last presentation we have been re-working parts of the story
-We removed some panels such as the rock crushing the monster
– Sound is something we discussed and want to be more prevalent in the story telling
-Finally we needed to emphasise the joke at the end more
We worked on the animatic and made a digital version each painting different parts of it on Photoshop. We presented this digital animatic and got feedback from our lecturer and class mates.
The ending doesn’t really make sense when the goat faints so we decided to take it out of the animation and just end it when the Chupacabra gets knocked off the mountain.
I edited it and added rough sounds to try and help it read.
We worked on the story and the started to make a 3D Previs. We made some basic models and animated some camera angles.
The previs didn’t read as-well as we original thought so we re animated some cameras and scenes. Lauren then took the files and worked on the previs while Michael, James and I started to model for the actual animation.
I decided that I would design the Mountain Shack (Goats House)
I researched houses with very angular sections I looked at haunted and run down houses from Tim Burton’s Films.
I like the roof and chimney on this house but I feel like it is to square.
I really like this house it is a great size and the roof is what I imagined.
I liked the porch on this house and the circular window up stairs.
Shack Model Development
This is my first mock up model of the shack trying to take parts from the photos I researched. I used simples shapes and played with the vertices to try and make a wooden plank effect. I also misplaced tiles and added nailed down boards to give it a run down look. In the image below I added some rough colours.
I decided to make an upstairs level and a slight side to the roof to make it more visually interesting. I continued to move and planks to make them all different shapes. I added spouting and a barrel to catch water as it was a mountain shack.
I thought it would be cool to add in a weather vane but instead of a Cockerel on top I made a goat with arrows. I also added in tin cans and I modelled torn fabric based on fabric found tied to bridges and tents in Nepal. I thought this would be cool as I could add cloth simulations to it to demonstrate that the shack is getting battered by the wind.
I was testing some lights as I thought it would be cool if the wooden planks allowed light through them so you could still see the basic shape of the shack even in darker scenes.
I also added in a side roof covering barrels and buckets. I tiled it and added a slant to make it look like it was dropping with age
I also modelled some other assets like a barrel, stove for inside the shack and buckets etc
Chupacabra Concept Model
I wanted to give sculpting in zbrush ago so I decided to sculpt a Chupacabra concept. I Focused more on a dog like creature as I preferred that idea rather to the alien ideas. I used the drag tools in zBrush to get the rough shape then used carving and clay tools to build up more detail. The arms and legs didn’t turn out that well but I was pretty happy with it being my first attempt at zbrush.
I made a very basic rig for it so I could pose it beside the shack for scale and renders.
We decided that we wanted our textures to look quite cartoony and resemble a painted style. I recently have been playing the game Fortnite which has a cool style and textures are painted. I suggested this and thought it would make a good reference for our textures. Michael suggested using the program 3D coat which turned out to be very easy to pick up and really useful for texturing models in the style we wanted. It was quite like Photoshop just on 3D models.
Example texture I painted onto my barrel:
Final Textured shack
I went on to add ncloth and wind to the torn fabric. I played around with the settings for a while reducing the dampness etc until the cloth looked right.
I also wanted to add leaves tumbling around outside as Lauren was making some tree models so I thought it would be another way to make the animation feel more real. It also will be useful as a establishing shot for the shack and a build up.
I had high poly leaves and low poly leaves I assigned the high poly to the low poly so they would follow each other exactly when I moved either the high or low. I then added the cloth to the low poly leaves. This sped up the cloth allowing it to be edited in real time. I added a turbulence box and air columns to the keep the leaves moving naturally and to prevent them from sticking. I played around with numerous settings until the leaves looked natural.
I used the free leaf model provided from this Plural Site Creative Tutorial. Then made low poly and high poly models from that.
I took it upon myself to do the lighting for the scene. I particularly like lighting so it was a no brainer for me. I discussed with the group and originally we had thought that the animation would be set at night. I proposed that it would be set at dusk with an sunset through the mountains. We all agree on this and so I began working on the lights.
I created an environment to create light rays and a slight fog which pushed up the render time but allowed me to play with the light in different ways. (god rays). It also made the mountains seem distant which was nice. I primarily used a spot light placed behind the mountains to act as a sun. I then placed area lights to make a gradient of colour behind the spot light. I placed four more area lights to light the mountain range and shack from different angles so the detail of the models can be seen and they aren’t just silhouettes. I also then placed two directional lights to light the shack and general scene. The shack also has lights in side to shine through the windows.
First Light test. I used an image plane as the background
I moved on a used the most recent scene which I put together. First light test with atmosphere.
Back light added producing more colour.
More Back lights again.
Darker versions. This version looked cool but wouldn’t work for because you can’t see very much detail.
Final Scene imported and constructed. Light setup from above.
Maya viewport version.
Final lights through valley view.
During the animation process Lauren, Michael and James used the rigs to animate the scenes. I setup the ncloth for the leaves and fabric blowing outside the house. I began to setup the scene and import the ncloth . We used blend shapes in both the rigs.
I mainly focused on the scene set up while my other team members where animating using the rigs. I placed the trees and ncloth. I also then went to place mountains to benefit from the lights to give a nice backdrop. Once the animation had been imported I placed the textures in the correct file and applied textures. It was then sent around to have the remaining textures applied.
Michael and I carried out the renders. We ran into numerous problems with the render settings and the colour correction settings. We also had problems with certain animation and cameras. We sat for two days and fixed the problems and then went on to finally render. We used the settings 4/2/1/2/2/2
Problems we fixed included:
Rigs separating from models
Ncloth not colliding.
Our lecturer Alec set us the task of downloading the Jack and Jill body mechanic models and use them to create poses. This task allowed us to start to learn how to use them before we try and animate the models. The image below shows the basic controls of the rig.
Here are some poses:
Boxing and Performer
I then went on to try and animate a walk cycle and body mechanic movement using the rigs.
Before we started to try and animate a walk cycle we were shown a demonstration in class on how to approach it. I went on from there to try and animate the jack rig in class. I found it rather difficult yet great fun as you could create whatever walk you wanted.
The hardest thing about the walk cycle was making it look smooth and getting rid of the popping knees.
This is an early attempt on a walk cycle. I removed the arms and the head so I could focus on the leg movement.
The main problems were the knees kept popping.
Here is a video that explains the process and helped me during the task.
I also found this channel which provides reference walks and other movements. I used the male walk as a reference.
The most useful reference material I used was from Richard Williams Animation Survival Guide. I imported the images into Maya and used them as a rough guide for timings and keyframes.
After I tried a basic walk. I went onto to try and get a basic run. It turned out very rough but was great practise before I went onto animating my final walk cycle.
I then went on to use the Jill rig to try and animate a female walk. I used the images above as a rough reference.
I also got Jack Ellison – https://jackjellison.wordpress.com to video me attempting a girly walk.
I looked at other examples of female walk cycles on youtube to get a feel for the differences between male and and female movements. (Mainly in the hips)
This is a play blast mid way through animating the female walk. I tried to animate the arms at the same time as the legs which proved to be quite hard.
I went back and used the animators survival guide for reference for the arms.
I found the placement of the arms extremely useful as it was one of the hardest parts of the walk to get right. I adjusted the hands out slightly to give a slightly more feminine look. (Over exaggerating)
Here is a render of my final walk cycle. I rendered it out with a light to give the impression that the model is actually making distance. overall the effect worked yet the shadow makes the ground flicker slightly.
For my body mechanics I wanted to focus on some kind of extreme sport. I thought it would be cool to try out a free style move. like a backflip.
To practice before starting my final body mechanic I decided to try and animate the rig carrying out a upper cut boxing move.
I used endlessreference’s video to base my upper cut on.
I blocked out the major movements of the character based on the reference then went back to fill in some keyframes to make the animation slightly more fluid.
The animation didn’t turn out as well as I wanted and I felt like a punch didn’t really show as much movement as I wanted.
I then moved onto my backflip. I used a reference video of a backflip.
During animating the backflip I decided to turn it into a dive implementing the backflip movement into a a dive through a hoop.
Here is my final render:
We have been set the task to create a 3D interior/exterior scene that conveys a narrative. We have split into groups. I am in a group with Holly and Conor.
Storyline: We started by discussing some ideas and how we could deliver a narrative through visuals and sound.
- Boxing Arena (1940s-50s) / Pool table/ aftermath of sporting event
- Old person’s living room/ up as reference
- D day Beach/ Saving Private Ryan
- Holding Cells (which have been escaped)/ The Dark Knight
- Strange/evil bakery
- A murderer’s house/detective office (room with clues to their evil character)
- Kung Fu master’s place of meditation and training
- Stage show/eerie magic show
- 1920s moonshining / Speakeasy (L.A. NOIRE inspo)
We picked our top ideas and discussed what we could do with them before deciding on a final idea to go with.
Old Person house/Living Room
Boxing Arena/ Aftermath
We decided quite early on that we wanted to go with the boxing idea. We discussed different settings/ narratives that we could explore. We particularly liked the idea that the Mafia would be involved and there has been some sort of scandal.
Boxing Arena –
- time period could be old or new.
- Two fighters.
- Could be a scandal involved/cheating
- Old vs new boxers (end of an era)
- Possible accident/taken too far (the referee could be bad or a cheater)
- Props – Banners/flyers/newspapers
- Sound effects/music to explain story too – shouting/screaming/ whistles/ booing/ dollar bills
- Possible camera technique- newspaper/flyer flies towards camera to start story then it blows away to go towards stadium
- Animal vs man
We have decided to set the scene between the 1940s-1950s giving us a nice style and plenty of assets to make based on the 1940s.
Concepts and Imagery
Photo mashed concept
Tunnel into arena idea
Older boxing sketch idea/ based on Sherlock Holmes
Story Behind the scene
The narrative we wanted to explore was that the Mafia have been involved in a scandal during the big fight. We discussed potential ideas and did some research on the Mafia and boxers during the time period. Holly found an article about a boxer nicknamed the Raging bull (LaMotta) who deliberately a lost a fight for loyalty to the Mafia.
“We continue looking at the darker side of the boxing industry. Former middleweight champion Jake LaMotta spent five years trying to avoid the trappings of the Mafia. Known for his grit and toughness in the ring, LaMotta was called “Raging Bull.” Finally, through his brother, LaMotta was offered a deal for the championship, in exchange for loyalty to the Mafia. LaMotta signed with the devil and intentionally lost a bout.”
We decided to base our narrative on LaMotta with the Mafia being involved with our match. We decided that we would make assets/props suggesting the Mafia were involved (at the match) guns/ fedoras, money briefcases and shaded glasses.
To start we decided to make some posters based on the LaMotta fight against Billy fox
Holly came up with the main style for the posters. We then made one poster each.
Modelling the Room
I decided that I would like to come up with general room design. So I began modelling what I imagined based of reference images and input from my team. Before I started on the room I made a rough boxing ring so I could have a centre piece to base the room around for scale.
I also tried out some lighting.
I began to place shapes to create the curved arena I imagined.
I then went on and made the arena stands as rough shapes and started to make a large stand in the background.
I added in railings, stairs, temporary floor boards and some other models just to scale the scene.
Final Room Design
This is a render of the final room model and its layout.
Room modelled by myself, assets modelled by Holly, Connor and Myself.
We also discussed the idea of placing silhouettes of Mafia members which could cast shadows over the scene. This idea wasn’t used in the end due to use of the occlusion texture.
At first UV mapping seemed daunting but I gradually got the hang of it. The images below show some uv maps on some of my assets. The video helped me learn how to UV map.
As well with making the room I took my other group member’s assets and placed them round the scene. Here are some renders of our scene with our assets placed.We focused on the placement of the Mafia assets to give subtle hints. For example the money coming out of the briefcase. My main goal while making the room and placing the assets was to make it feel lived in. Like a boxing fight had just taken place.
Final Rendered Video