Art - Post Mortem
Introduction/Overview
Elliott and I have created an environment pack for a near-future slum with a European/British Slant, the overall theme for this a large city/settlement built into the side of a rock face to escape rising sea levels. The project’s focus was assets on the scale of a multiplayer first-person shooter. The environment pack was broken down into smaller asset packs such as: decal pack, clutter pack, large prop pack, building pack and vista pack.
Parallel to this we also developed open source tools that will be included in the environment pack. The tools include a modular building system, a random asset location system, spline based cable system all created using Blueprint within UE4.
Similar existing tools are only available via proprietary engines or through payment e.g. Ubisoft Anvil, UE4 Marketplace. We want to make our tools free and open source bringing this functionality to a wider audience.
Project Details
For my section of the project I needed to focus a great deal on making industry standard, reusable and modular assets. This means that I produced a select number of assets which can be interchanged, adjusted and reused as much as possible. I was also in charge of the Art Direction, something that I feel I struggled a great deal with and will discuss further later on. Project Duration
The project started on the 12th January and finished on the 22nd May being approximately 4 and a half months in total. The original time of 12 January - 24th April was broken down into a development schedule which was constantly referenced and gave us a good understanding of where we should be at any given time. Unfortunately due to various reasons the actual development of the project became somewhat more fluid as the project progressed and the extra time was assigned to various tasks when it was announced.
Successes of the Project
This project was very much a project based on an idea for a system Elliott and I had come up with to create a large environment quickly for a specific type of location we had developed. The system was quite theoretical to begin with and because of that both of us were still learning a great deal about the idea, how to implement it and how to best create assets for the idea. This learning process and teamwork was definitely one of the major successes of the project. We could discuss both sides of the project (tech and art) with each other without things getting lost in translation; we had our specialisations but both of us advised the other on their section of the project. This meant that throughout the project we were getting reliable feedback and ideas on how to further our section of the project.
The project started out with a quite different artistic direction and went through multiple overhauls to improve upon the idea. Each time these overhauls significantly improved the direction of the display level and honed in the asset list. They were mostly sparked by my own frustration with the way the display level was coming together but were sometimes pushed by tutors during presentations were it was clear I was struggling to make the visuals work together. This is constant reflection, critical assessment and vision was incredibly important to the success of the project as without constant iteration the artistic direction of the project it may not have come together as well as it did. That being said the Art Direction itself is probably the thing that was the least successful, constant assessment and willingness to change and rework helped to reduced the problems caused by this, I will discuss that more later on.
The project being split into multiple packs of assets, and also being thought as more of a collection of assets, was a really good as it allowed me to spend longer on individual assets to make them as good as possible. Each asset had to be of a certain standard and I can’t hide mistakes for the sake of the environment as a whole, as each asset will be judged as an individual asset. I feel that I have improved greatly in many different areas because of the time and effort I put into individual assets. Particularly successful assets are the Road Barriers, Traffic Cones, Sofa and Builders Bags. I feel my time assessing the forms and materials is best seen with these assets. I spent a great deal of time at the start of the project studying materials which meant that by the end of the project texturing objects was a great deal quicker. Also my time studying the way dirt and grime falls and builds up really helped me in my interview with Playground Games, where I was commended on the Road Barriers and Traffic Cones.
During one of the overhauls I realised that the lighting in my test level was very bitty, noisey and uninteresting and so I spent a few weeks working with lighting as my focus. Fortunately during this overhaul I had come up with a better layout for the environment which meant I could spend more time purely focusing on the lighting. Until this point I was quite certain that I wanted the environment to be set during the day as you can see from my concepts but because of the intensive lighting focus I experimented with some drastically different lighting. I found that an early evening/night time setting could provide some much more interesting lighting. I think this is a choice that has worked out pretty well for the environment. I still feel that I have a lot to improve regarding my lighting ability and would have prefered to learn some more about it and continue experimenting.
Problems with the Project
This project began with quite an epic concept in mind, a large wide open vista with a city sprawling out in front of the viewer. This is an environment that I had been concepting and thinking about for a while and such I thought I could quite easily realise these thoughts. Because of the level of detail I wanted to go to with each individual asset and because I wanted the environment to tell a story the concept developed and the idea continued to become more and more refined. The major problem with this was that I came into the project with an entire world, storyline and vision that I wanted to portray and it was too large a scale to feasibly do in the time and to the standard that I wanted to do it. I also found as I started to work on the first whitebox that however much I worked into it and reworked my ideas the environment wasn’t feeling right. The sense of story wasn’t coming across. After realising this I looked back at my reference and found lots of reference of seaside towns like Dover to try and create a layout that felt as if people lived in this space. I created a scene that resembled a residential British road I feel like this direction had potential and really helped to come up with ideas that I used in the final iteration of the environment unfortunately again my direction wasn’t refined enough and so I was stretched quite thinly between which assets to produce and the environment lost momentum and became a noisey mess.
A lot of these Art Direction issues came from coming into the project thinking of and idea rather than an image, ideas can be reconsidered in a different way every 5 minutes where images are solid and unchanging. After realising quite how wrong I was going about this I discussed with tutors and peers and started looking into British Dockland areas and trying to find reference from these various towns that I could piece together into an image to recreate. I was still concepting at the end of april one month from the final hand-in. I have produced numerous different iterations of concepts for the environment and the asset pack I feel like these concepts were very much an idea generation and expression tool and we never wanted to exactly reproduce these but use them as a springboard this worked as a process but I don’t feel that it was the most effective method.
As the project progressed I feel that the focus became more and more about the close up assets, with VR being a possible method to show the work off at degree show and the amount of detail I wanted to achieve with each asset, this unfortunately meant that much of the vista pack isn’t to a standard that I am happy with. The project’s focus changed for the better but this does mean that unfortunately the vista pack suffered from a lack of time. This is definitely the weakest section of the asset pack but it is still functional as a vista pack of assets to be used from a great distance to give the sense of world beyond the buildings. I had big ideas for this pack of assets but feel my time was better spent both for the project’s sake and for my employability in industry to focus more on the close up assets.
Changes I Would Make
The project was quite a learning curve and if I were able to tackle a similar project in the future I feel I have definitely learnt many things that I would bring to the table. First and foremost I feel the project was a success but the speed of production was very changeable as either I was struggling to achieve the mood and atmosphere I wanted in the test level or I was working hard to achieve the level of finish that I wanted on individual assets. I really enjoyed working on a project that had a bit more research and development in the pipeline but I feel if I were to do this again I would approach this element in a different way. Having a solid key image either from a third party concept artist or from a photo as an aim would have really helped the project to flow smoother. This is because I was quite stretched between various different elements of the project. Conclusion The project was a struggle but came together fairly successfully the environment feels grounded in Britain and has a high level of polish. The research element of the project and the work with Elliott’s tools I feel worked very well in the end, I would like to have longer to experiment with what I can do with them but am content with how they have been used in the display environment. I am looking forward to optimising the environment for Virtual Reality in preparation for degree show and I look forward to making some small one shot scale environments to practice lighting and show the speed of producing environments. I really want to do a video showing the process of creating an environment, hopefully I will get this done this weekend and be able to post it on here when it’s complete. This will be really good evidence of the speed of creating an environment.
The project really helped to speed up my asset production speed and has been quite a struggle to keep up with the schedule for. Going in I thought I was more certain of the end result than I actually was and I feel that this is the main thing I have learnt from the project. Unexpected changes and uncertainties are things that have changed the schedule of the project dramatically so I hope to counter these things with future projects. Either know exactly what I’m going to produce or give myself specific extra time for elements that are not concrete.
Final Display Level Renders
Quality not Quantity Work Smarter not Harder
Fail Faster
