Subnautica - Gameplay Trailer

I got this project via Ryan Clark of Crypt of the Necrodancer; This was doubly flattering because I've never worked with him! (Necrodancer's trailers were made by Marlon Wiebe)

This was a challenging project where I had one of my first real big missteps in the creative process. Luckily I had enough time to recover from it, and the trailer didn't suffer for it but I'm not happy about how I started this project off.

I never played Subnautica before getting this gig; all I knew at the time was that it was a very popular scary underwater game in Steam Early Access. I don't typically work on survival-genre games, but the game is so beautiful that I couldn't say no. I didn't know just how big a task I was signing up for.

At the start, I watched the previous trailers for Subnautica to get a sense of the game. I had a long conversation about the trailer's direction with Charlie Cleveland the head of Unknown Worlds. Previous trailers focused heavily on gameplay, but he wanted to highlight the single player story campaign to distinguish it from other survival games. Most survival games I'm aware of have no single player story (hence my general lack of interest) so this excited me. He gave me a rundown of the entire story, and then I asked for materials.

You normally see this shot during the day in the game, but a member of the team recommended capturing during sunset so it'd be prettier.

My first mistake was that I didn't play the game for a significant amount of time before starting work on the trailer. Because of this I had no idea how the story was told in the game. So when I requested files that included voiceover from the games' characters, I had no idea that they're mostly audio logs from events prior to the game, and are never depicted on screen in any way. I was intimidated by the size and scope of the game, so I thought going through raw dialogue files would be more efficient than playing this massive game. This mistake set me down a path I would've known was not worth walking down, had I done my due diligence.

I wasted a lot of time combing through the dialogue files, and constructing a video outline for the trailer that could not be told via the game's visuals. Even worse, the vibe of the video outline was that of a bombastic Hollywood trailer, which is the complete opposite of the feel of the game; this is another thing I would've known just from a few hours of playing the game.

This contains audio log spoilers

I spent the better part of a week making that outline, and only some dialogue remains in the final cut. After presenting it to the team, they immediately said that it wasn't reflective of the game, and they were absolutely right.

This response was completely my fault; it was on me to get it right the next time. So I dove in deep with Subnautica; I watched a LOT of YouTube tutorials about how to get started in the game, the numerous game mechanics, and just what you do in the game. From this I started my own game where I played the game in "Freedom" mode which removes hunger and thirst mechanics, but still leaves all the exploration, resource gathering, crafting and discovery. 

I played enough to understand the core gameplay loop, then used debug console cheats to skip ahead to parts of the game that require hours upon hours of play to unlock. With this new knowledge I started my video outline all over again. One helpful thing I did during the previous cut was to subclip and transcribe all the dialogue into separate bins; this let me find what I needed very quickly.

I threw out all almost all the character dialogue because there was no way to visually support it, but a few lines I held onto because they could potentially work as a last line at the end of the trailer. I ended up strictly using the PDA AI voiceover because by it's mostly game world exposition. 

To add visual interest, it was suggested I recapture this shot while holding a hoverfish!

Next I went through all of the music again to figure out which cues would be more fitting to the tone of the game. There were quite a few very high energy cues that I mostly threw out except for a few that might work for a short and snappy intro sequence. For the main body of the trailer I wanted ethereal, ominous, epic music that gave a sense of awe and wonder.

I decided I wouldn't tell much of the game's story at all. Cutting visuals to the character audio logs would require a massive amount of cheating, to the point that it would misrepresent the game. Also, if I showed visuals from the actual in-game story, I would very quickly get into spoilers. I ended up making a trailer that gave an overview of what you do in the game, with a very brief tease of the greater story.

The escape pod sequence is the first thing that happens in the game; it was featured in previous trailers, but I still wanted to use it so this trailer could exist in a vacuum. I wanted to spice it up a bit to set it apart from previous trailers, so I cheated some footage of gameplay to make it look like the player is running out of the exploding ship before escaping in a life pod. 

My first version of this shot had no other fish or rock formations. The art director suggested I add those in to give a sense of scale.

For the main body of the trailer I made a list of all the biomes in the game, all the tools you use, and all the vehicles that you pilot; I needed to figure out what I would show, and what I would omit, because there's simply too much in the game to show in the trailer. 

The best and worst part of this project was simply how big it was. I just had to do the legwork of exploring every biome enough to know whether I would show it in the trailer, how to show it off best, and why to show it. I explored every wreck, used every tool and researched the different creatures and resources. 

I wanted to be true to the game as possible. For example, it's possible with cheat codes to go anywhere in the game without dying, but certain regions are at depths that require special vehicles, so I wrote down what biomes were at what depths, and what vehicles could go to those depths to make sure my capture was accurate. 

Many of the resources in the game simply disappear when you receive them. I chose this one because it visibly gets cut off.

The one thing I was very worried about was showing off the base building mechanics. I'm simply not someone who plays games where you build bases, forts or really anything; it's not something that I engage with in video games. I asked Unknown Worlds if they had any save game files with impressive looking bases, but surprisingly they didn't.

Unknown Worlds sent out word to the community via Discord and Twitter that they wanted impressive bases to feature in the launch trailer. I ended up with a little over half a dozen submissions, and the final cut used capture from three. Special thanks to: ComicalSkate, Games For Days and Sans the Skelebro for making the trailer even better than I could've hoped to in the time that I had! 

There are TONS of things to fabricate in the game, but I chose this stasis gun because its basic shape is recognizable.

The basic structure of the trailer stayed the same from my second video outline to the final cut:

1. Cold open of ship exploding, and protagonist escaping.
2. Contrast the bombast with a serene sequence that's in keeping with the tone of the game.
3. Gradually introduce game mechanics reflective of the core gameplay loop.
4. Show off the most impressive looking parts of the game
5. Tease the greater story.

We also wanted to incorporate some quotes into the trailer, so I cut in some placeholder cards for spots where I thought they could work. 

This was difficult to recapture because I had to find an empty space, start building it, rush over with the freecam, and find a good angle.

As soon as I had a video outline with this approach I sent it to the team who agreed that this was a much better direction. After I sent my first version with capture it was just a lot of iteration and refinement via notes from the team. Special shoutout to art director Cory Strader who worked with me closely to make sure the game looked as good as possible!

Some examples of changes I made to the capture:

1. Incorporating more footage from brightly lit biomes; my first cut leaned more heavily on dark regions, but the majority of the game is spent in bright areas.
2. Recapturing the reefback to give a better sense of scale via small fish and the environment.
3. Enhancing basic swimming shots with actions like using tools.
4. Shortening the intro to get get to gameplay sooner.
5. Recapturing scenes during a time of day with more flattering light
6. Avoiding animation glitches
7. Recapturing so that aquatic life are featured more prominently in the frame.

There was no 3rd person debug camera for vehicles, so this shot is totally faked. I just swam towards and around it to give it movement.

There were also some debates about quotes, and how many to use. I was a little skeptical about having all the quotes coming from streamers and YouTubers, especially ones I'd never heard of, but if YouTube comments are at all a good measure, the response was overwhelmingly positive.

This was a challenging project, but I'm very proud of the final trailer, and the team members at Unknown Worlds were also very excited about the trailer. One of the most satisfying parts of my job is seeing a team get re-invested in their work after seeing my trailer for their game!