Well I finally got around to posting and sharing my experience with Phoenix Point. I went through some big life changes and this post had been placed on the back burner for a good amount of time. But now that the dust has settled a bit, here is my experience with their Art Test.
Not to long ago [ Mid-january. Told you, it has been a while. ] I was given an art test and instructions to create a small environment in less than a week. I needed to create heavily modular buildings that would fit a top down tactical game while dwelling in the universe of Phoenix Point. Focusing on a specific faction I had to make at least three buildings with various ground elements that could be arranged into a small environment and rendered in their preferred engine, Unity. I also had to utilize PBR and the metallic workflow. Deliverables included obj files, textures in Targa format, and a rendered fly through camera of the build. [That should have been my first red flag.]
Now given the allotted time, this is more than enough time to generate a small amount of assets to create a good base environment. Something just pass a blockout mesh or simple stand-ins. What actually happed however is that I unwillingly turned it into my own game jam. Going 20 hours straight to produce three buildings the day of the turn in.
As expected, it did not go so well. I highly recommend never to rush Art Tests.
But before we go into details and showcase my result. Lets go through and breakdown the art style. Including the faction design I was tasked with recreating and a general “rule-of-thumb” for top down modularity and asset specs.
All images of Phoenix Point used in this blog can be found in their press kit, various social media sites, and Trello pages. Nothing presented here was contained in the Art Test Package.
What’s the Purpose of an Art Test?
Now when you are given an art test, it’s usually done so to gauge whether you can work with little to no guidance and to determine what your work process is. If you’ve gotten an art test, it generally means they like your portfolio. So give yourself a pat on the back!
The art test will generally gauge how fast you work, how you pace yourself, and how you prioritize your work. Essentially they are testing how well you can perform without any micromanagement. Most art tests come with tight deadlines, which is there just to gauge if you can meet deadlines. So if you don’t finish an art test, at least submit what you have.
“Every shot not taken, is a shot missed.”Some old saying my parents taught me.
Google credits hockey player, Wayne Gretzky
Lastly, don’t be shy to ask for compensation for your time. [ Something that I myself don’t do much of. ] For freelancers, dedicating time towards art tests means taking time away from paying gigs. Or if you work in a different industry, taking time off work to focus on the art test. [ Which I have done while working at TRU.] Your time is valuable and you should be compensated for it. Especially if the test requires you to submit source files. NEVER submit source files. [ For this case I only handed in three out of twelve assets, as they wanted source files to “review” my topology and modeling. Why do you they need all of it? ] One way or another the company will use it in a stage of production. If source files are required, politely ask for compensation. Until then just send rendered watermarked images. In my own experience I did a lot of art tests when I graduated college. Pouring hours of work and complete weekends [ had a regular 9-5 retail job at the time ] into art tests to never hear anything back.
Lastly, always ask for feedback. Most companies will ghost you, but there are a handful that will provided reasons and breakdowns as to why. One case it was simply another candidate had better traditional art skills than me. My art test was great and well done, but the other candidate had a better traditional foundation. [ Yes, that did hurt a little. ]
Art Style Background
My art test focused on the Disciples of Anu, one of many Phoenix Point factions. Now according to their Trello and Website, the Disciples are an impoverished zealous group that worships a sun like god. Demonstrated by a pristine and imposing temple amongst the slum like dwellings. This sets up some great contrasting themes with the rigid, hard edge, and structural tones of the temple to the organic, lumpy, and filth of the worshipers.
Per usual, I run the concept images through some color generators, and create a simple color pallet to base my texturing off of. So the art test should contain these color elements within three buildings that demonstrate organic modular shapes. These buildings also need to invoke a feeling of filth and squalor. Simple to do with the darker colors in our pallet. As a bonus building, we should also include a temple to completely incapsulate what the Anu faction is.
Again, the genre that this game is aiming for is a top down turn based strategy game. Which falls in line with games like XCOM, Into the Breach, Banner Saga, and Battletech. So taking a look at how those games are built along with provided concept art will help establish a solid conceptual foundation to create our modular assets.
So first thing first, we want to knock out the easy tasks. Allowing us to focusing most of our time on assets and components that will be seen by the player. For example, trash models can be done quickly as the player wont spend much to any time thoroughly analyzing the garbage textures and mesh.
Before we even begin modeling we need to determine the relative size of the assets. Most turn based games rely on a grid system. As shown below. So most if not all 3D assets need fit within a certain grid space. Whether it be a 2×4 or 5×3. We don’t want to create abnormally large assets like 16×8 grid spaces, as that will take away the modular use of the asset. With tactical games there is also a lot of elements that allow for stealth and cover, as well as vertical travel, so we also need to incorporate that into the design of the village.
In order to create buildings that would fit into their world we need to match their scale and determine the grid spacing. For Phoenix Point it’s very easy to see that scale in their more human environments. By looking at their tiles, we can approximate their grid spacing. We can determine rough height and width ratios, as well as the complexity of meshes and transfer that info into our own representation of the Anu.
This link is to the XCOM 2 Art Dump that has great examples of how the props and components where broken up and made.
We need to first create a set of Modular assets for the Anu dwellings. By either creating smaller unique buildings or environment elements like guard towers or houses. The other option is to create individual components like walls, floors, and barricades. We need to prioritize less important assets [ which generally are the quicker ones to create ] over more crucial assets that take up more visual real-estate.
A list of tasking would look something like this.
- Background and ground elements.
- Something that can be generated very quickly and is then hidden by our created buildings.
- Modular Building Components.
- Simple Wall, Floor, and Roof components.
At this point we don’t have to worry about texturing. We just need to quickly generate meshes and get them into the scene to determine silhouette and composition. This pass would focus on the Anu Dwellings at the time, given that was the first requirement on the Art Test. Focusing on the temple components last. Once everything is in the scene, we did a quick lighting pass, and a good base is generated, we can then focus on textures.
- Simple Props.
- This would be assets like trash, small decorative components [ like a string of lights or potted plants ] and other components that helps break up the silhouette and composition of the building but isn’t paid much attention to by the player.
- Larger Props.
- This would include buildings and features that are more recognizable, used frequently by the player, or complex. This would include barricades, level defining props, or any intractable components.
- Clean Up.
- Then lastly we would want to clean up and get some good lighting done and set up our fly through camera.
Now that we went over how to do it, here is an example of how NOT to do it. Now this was a pretty bad submission. My priorities and tasking were backwards due to very poor planning and not much forethought for the outcome. I created three texture sets for the whole thing, attempted to create a ground level, and cobbled together three building from twelve static meshes. [ It’s one of those “do as I say and not as a I do” kind of hypocritical moments. Really should listen to myself more. ]
I reached out after submission [ as I was put in direct contact with a member of the hiring team ] and inquired of any feedback or comments on my submission. As of April haven’t heard back yet. [ And pretty sure I won’t ]
So what could I have done better with my test?
- Don’t cram 7 days of work into a continuous 20 hour sprint.
- You overwork yourself, don’t get a break from it, and really loose track of how things are going.
- Good Lighting
- This is pretty bad lighting. Doesn’t set a mood, doesn’t evoke a feeling. It’s pretty bland.
- Nothing screams “filthy”
- Need to push more on the silhouette of these buildings. They are too rigid and clean, for the “hovels” of the Anu.
- Silhouette & Composition
- Again, this obviously screams “rushed production”
To many errors were made on this one that culminated into a final product that was other than what was needed. Doing this has allowed me to identify my mistakes, and as soon as the dust finishes settling I will come back to this and make it a proper portfolio piece.