The Production Pipeline of 'End of the Line'

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.




This report will discuss and evaluate the production pipeline of the final year research project by Matthew Jackson entitled “End of the Line”, in comparison to the methods used in the professional industry. It will investigate the processes, tools and various techniques throughout the production and creation of a CG (Computer Graphics) enironment, bound for animation, while uncovering the issues encountered in the aforementioned project and their respective solutions.


I wish to acknowledge and thank a few people for their guidance and assistance during the technical creation of the environment and globally throughout the duration of this project. Firstly, my supervisor, Dr. Paul Noble, for his endless feedback over the past 9 months as well as my colleagues and fellow students at Teesside University for providing me with the motivation and mental strength needed to push my project to its completion and beyond.



In recent years the disaster genre has made a significant impact on the film industry, however this success has not yet been fully investigated in other digital media industries such as animation, while similar genres, such as post-apocalypse and zombie horror games have been combined with aspects from the disaster genre and are booming in these very areas. Examples of video games which have recently been on the leading edge of this boom include Last of Us (Naughty Dog, 2013), DayZ (Bohemia Interactive, 2013) and Dying Light (Techland, 2015). These games all demonstrate excellent environmental design and stick to their genre tightly, however, although video games have utilised this dynamic genre, computer animations are yet to make the transition.

The “End of the Line” is an computer graphics (CG) environmental development project which focuses around the objective of producing a distinctive and believeable American subway station environment, suitable for integration into a computer animation, with the addition of drawing upon real world examples and references to enhance its accuracy.

This project aims to expand my knowledge of industry-standard 3D CG pipelines, workflows and practices when undertaking the creation of a CG environment, from concept through to final product for use in a computer animation. In turn I will compare and learn about professional practices at each stage of the pipeline, before implementing a reasoned choice into my own workflow and finally evaluating the cumilative progress, effectiveness and successfulness of my ability to implement these practices.

Other research included in this project which will be discussed in this evaluative report includes the use of physically based rendering practices (including shader nodes and CG lights within Maya 2015 (Autodesk Inc., 2015)), an overview of the architecture designs of the New York City Subway system and colour theory (including their utilisation alongside post-production methods such as colour grading compositing).

This report is subdivided into four further sections, exploring the arguments and research outlined above. These are:

  • Methodology
  • Research
  • Production Process: Problems and Solutions
  • Production Evaluation
  • Conclusion


The industry-standard CG animation pipeline has always been relatively unchanging since the first feature-length animated film, Toy Story (Pixar Animation Studios, 1995), was released, where the process was first standardised, but this pipeline was previously adapted from the Visual Effects (VFX) industry previously established. Since, this pipeline has provided the solid backbone for almost any entirely CG film or short animation, with its three clear defined sections, divided into sub-sections, which represent the various disciplines. Due to the fact that this method has been implemented successfully with few changes for over 20 years, I have utilised this structure for my own project.

The three sections include: pre-production, production and post-production. Pre-production is essentially the planning and research section of the project. Designs, concepts and research are completed as well as the project ideology being secured. Some of the departments primarily identified by the Animation Production Pipeline (Nora Willett, 2010) in the pre-production stage include: storyboarding and concept designs. She then goes onto discuss the production stage which follows the completion of pre-production. This stage includes the core elements of the CG animation creation, including: modelling, texturing, rigging, set dressing (or environmental creation), layout, animation and effects (or simulations) and rendering. The production stage provides a linear, but very flexible platform for the generation of content for the animation. At any time each stage can be revisited if problems are uncovered and modifications made. In essence, this stage represents a constant circle of improvements from all the included departments until the director is satisfied with the final product. Fabio Pellacini, a former pipeline researcher at Pixar Animation Studios describes the vertical hierarchical pipeline at the studio following these core workflows (Fig.1).

For the production stage I followed a very linear version of this structure, however as discussed above I revisited sections to make improvements when bugs or parts of the pipeline clashed with the progress made during the production stage. Although I have not used all of these areas due to the specialised nature of my project, the major have still be included as has the overall workflow.

Finally, I used the full array of sections in the standard CG pipeline. These are: editing, compositing and colour grading. Editing was included to ensure that the shots integrated well together, while the compositing and colour grading were used to erase any stray artefacts and colour management errors present in the final images which were overlooked during the rendering process. Each area of the pipeline will be discussed more in depth in the Production Process section of this report, including problems encountered and solutions therefore found.

Since this project focuses on all areas of the traditional 3D CG pipeline, before its commencement I evaluated my own skills to fully understand the disciplines which I would need to research to equip myself with the relevant skills required to complete the project.

Also prior to the project I consulted prior comparable projects to assess the correct timescale needed, based on my skills to complete the various sections of the pipeline. The result scheduling is shown in Fig. 2.



This research section will talk about the collection of research undertaken for this project, alongside its impact and implementation. I used the research to improve my skills and consolidate any weak competencies, so that a high level of quality was achieved through.

The first area of research centred around investigating physically based shaders, texture node networks and lighting in Maya 2015 (Autodesk Inc., 2015), ultimately linking to rendering through the Mental Ray renderer (Autodesk, 2007).

Physically Accurate Texture Shaders Physically based shaders such as those in Mental Ray’s Architectural and Design Visualisation Library (mia_material_x, mia_material_sss, etc.) differ from the standard Maya shaders because they scatter and react to light in a realistic manner whereas the default Maya shaders do not. These shaders will require little modification in order to reflect real world materials, since in reality all materials have a specular value and RGB value which can be applied to the shader.

As described in Digital Lighting and Rendering (Brin, J., 2014, p.323) a material’s attributes, such as its diffuse colour, specularity, reflection and refractive nature are defined by its bidirectional reflectance distribution function (BRDF). This function defines how much light is reflected and absorbed by the material from various different angles. Mental Ray’s physically accurate mia texure shaders are based on accurate BRDF data whereas their Maya shader counterparts utilise a more generalised and cut down version of this BRDF dataset, therefore reducing their accuracy. The same dataset theory also applies to the illuminating and sub-surface scattering shaders (for use on skin and plant matter).

Following this research I exclusively used shaders from Mental Ray’s Architectural and Design Visualisation Library, so that I could ensure that all objects and thus their mapped textures reflected light in a physically accurate manner.

Physically Based Lighting Physical based lights are CG lights that emit light according to connected IES (Illuminating Engineering Society) Profile data which reflects that of real world brands and models of lightbulbs (Digital Lighting and Rendering, Brin, J., 2014, p35-36). IES profiles, readily available from many sources online modify the standard Maya lights’ attributes into real-world units, for example, lumens and colour temperature. The lumens defines the intensity of the light, while the colour temperature dictates the diffuse colour of the lights. Most standard yellow tungsten light bulbs measure roughly 3200°Kelvin (°K), luminescent white strip lights measure at 4200°K, while nature sunlight rates anyway between 5000-10,000°K depending on the lighting and weather conditions (, What is Colour Temperature, Future Publishing, 2012).

When setting the light’s intensity using lumens, you can easily calculate the value based on the light bulb’s wattage rating. Whereas with standard CG lighting you have to constantly modify the intensity and thus produce a multitude of renders until you find the right settings. With lumens you can set the intensity value to the light bulb’s models correct value and then you are left with only one value to change, the camera’s exposure settings, rather than potentially hundreds. This pays dividends and drastically reduces the amount of time spent setting up the lighting rig, which can ultimately be spent on refining other areas of the pipeline which are much more experimental.

Linear Workflow Using linear workflow is an essential part of the CG pipeline, especially within professional, industry-standard packages such as 3D Studio Max (Autodesk Inc., 2015) and Maya (Autodesk Inc., 2015). When an image is produced from a renderer it has a gamma curve applied to it, however when we view this image on a monitor this gamma correction curve is applied for a second time, therefore significantly skewing the colour information and displaying it incorrectly. In order to correct this to ensure the on-screen colour levels match those displayed on the monitor post-rendering, a linear workflow colour management setup needs to be adhered to. By setting this up in the 3D package, the rendered image is given the inverse gamma curve on output thus resulting in a colour level which is the same as that shown in the package (, Colour Managed Linear Workflow, Autodesk Inc., 2014).

Incorrect usage, setup or disregard for this system workflow will cause unrealistic decay, blown out highlights and physically inaccurate lighting.

Architectural Designs of NY Subway

Colour Theory / Post Production techniques for mood



Toy Story (Pixar Animation Studios, 1995) - INDUSTRY PIPELINE EXAMPLE

Pixar Animation Studios

Physcially accurate shaders

Digital Lighting and Rendering (Brin, J., 2014) - p.323-324

Mia v maya materials example - - colour temperature - Linear workflow