Hi there folks, my name is Finlay Pearston and I am a lecturer in Game Art at Abertay University in Scotland helping aspiring game artists realize their ambitions of working within the games and digital art industries.
Prior to this, I was a student studying animation and went on to work in the VFX industry for a few years. I greatly enjoyed this position as it opened my eyes to all the wonderfully talented people and dedicated work that goes into VFX and digital art. I then transitioned into the game industry where I worked as an environment artist for several years on multiple releases on PC, Playstation 4, and Xbox One. This was the position that solidified my passion for 3D art and most importantly within the game industry. I have always been intrigued by the different creative ways artists can come up with to make beautiful art within, sometimes, tight parameters of time and resources, and the game industry shows it at its best, in my opinion.
Modular Street Pack: Goals
I decided to create a modular street pack because of a few reasons. The first was that I wanted to try distributing my work online as I had been an admirer of the work of organizations such as Dekogon who pride themselves on producing high quality, multi-purpose asset packs for game developers to utilize within Unreal Engine 4.
The other, more important reason, was that I wanted to flex my own creative muscles and create something that not only pushed me for original content but also on a technical platform as this would need to be something that held up to the high standard of work we see all the time on the Unreal Engine Marketplace.
The street pack came about as I had noticed that whilst there existed a market for high fidelity props and lower fidelity buildings there didn’t seem to be many cases of high-quality props and building assets that could be used to make a complete environment. This became the key focus of the project, to produce something that any new artist or game developer could pick up and run with, without having to worry about creating loads of additional art.
Reference was a big part of the project. Although for a lot of the pieces it was a case of just taking a walk and seeing how stuff was built in real life I still needed a big library of images to keep referring back to which I scoured the internet for. The biggest challenge was that there are so many variants of the same sorts of things that finding a variety of images from different angles of a specified manufacturer model was tricky at some point. Without this, the assets would have lacked authenticity and since these are things that we all see in day to day life, attention to details we are all familiar with was key.
I broke down the pieces into a list:
- Buildings (pieces that would be created to construct the main building shapes)
- Details (things like decals and effects that would be added towards the end of the project to make the scene look lived in)
I modeled all the low and high poly versions of the assets within Max and then baked the normal and ambient occlusion maps from those meshes inside of Marmoset Toolbag. This was a great way to work as any changes or problems with things like cage sizes or clipping could be altered very quickly and the outputs could be viewed almost instantaneously.
From here, I moved into Substance Painter where I imported the previously baked normal and AO maps and began the texturing process on all my assets. I am a big fan of discord channels and the DyNusty Empire discord channel in particular which is where I discovered a grading tool ACES LUT for Substance Painter. It allowed me to view my assets as they should appear within UE4.
The tileable textures in the scene were all done using Substance Designer. I am a huge fan of this program as I have a great deal of fun every time I create materials in it. The biggest problem is not knowing when you’ve spent too much time tinkering around rather than just calling it done. I would encourage anyone who is interested in material creation to check it out as I have found it massively superior in creating believable and flexible materials.
The presentation of the project was another big-time expenditure but since part of the reasoning for the pack was to appeal to an online market, it needed the time. I began by constructing a small street area which I would use to demonstrate an example of what the artist could achieve with the pack. This was important because I wanted it to be clear that this pack could stand alone without any additional input of art if that was what the person wanted. I then began placing cameras around the scene in order to capture the street from the most interesting angles. This was difficult as, with any modular scene, you run the risk of being repetitive with what you are showing since the assets are designed to repeat with small variations. This meant in some cases that parts of the street’s composition had to be adjusted in order to make each shot tell something new.
I am a keen enthusiast of lighting and grading within games, films, TV, you name it, and so this became the next task to overcome. Creating a natural scene that we are all familiar with is hard as we are all able to spot blatant errors. This again meant a lot of experimentation by testing out different times of the day and sun positions. I had already decided at the start of the project that I would have an emissive option added to the relevant assets so that this pack could accommodate night scenarios. This was the right decision and paid off - lighting night scenes is a joy since, most of the time, they look much more visually compelling that evenly lit day scenes.
The project took around 6 months to complete part-time since I was still employed full-time as an environment artist at a studio. This meant a lot of creative working hours, but it was well worth it as I knew I was going to get something I would be proud of.
The major challenges of this pack where mainly encompassed by the fact that I wanted this product to be accessible and relevant to whoever purchased it. I get a great kick out of picking up a piece of content just to open it up and learn from what the artist has achieved, potentially applying it to my own practice. This meant that my pack needed a wide variety of features for the artist to be able to learn from clearly but that could also be advanced enough for those more familiar with UE4 workflows. I feel like I struck a good balance of this with the final piece and I am confident that it will be useful to those who have need of it.
The next thing for me will be to try and create more content like this that can benefit those in the development community whilst still making sure I am pushing my own skillset further and further.
Special thanks to Ross McCafferty, a constant source of piercingly great feedback.