The magic of salt and smoke. Recreating the techniques used for Sea of Thieves at an Indie price.

Hi Everyone, I’m Richard, a baby real time VFX artist. I’ve been building effects for game engines for about 5 months now. I’ve got a good grasp of all the fundamentals but found that I’m at a point now where I need to step up my game. That is why I’m starting this blog. From now until I inevitably forget this exits and stop updating it I’m going to be delving into various effects I think are cool and deconstructing them.

Sea of Thieves is a magnificent piece of work and is easily one of my favourite games of the year. It’s not often that a game comes out where you can steal a boat, kidnap two people who barely speak any English and force them to summon a giant fish. There is so much that is worth writing about however I want to focus on the behind the scenes video they released early October. Inn-side story #32. Rare lightly discusses a couple of methods they use to build some of the effects. This week I’m going to be taking a deep dive into what they discuss and use them to create some effects. In UE4 I’ll be creating a smoke effect, and a water splash effect. Since I’m purely focusing on the process rather than the end effects, I’m going to ban myself from looking at the game from reference. Anything I make must come from reference I gather myself.

The topics of this week’s blog are:

1. Using this blog as an excuse to go to the beach

2. Using a macro lens to capture smoke up close

1. Using this blog as an excuse to go to the beach:

So admittedly I was very excited to write this blog as it gave me to opportunity to grab my two wonderful assistants, Burgundy and Montague, and spend a day playing around on the beach. We live very far north compared to the artists at Rare, so we couldn’t get to Cornwall beach where they gathered their reference footage. We had somewhere better though, the fine northern beach Redcar.

The place seemed much greyer in person. While it doesn’t have crisp blue skies, heavy waves crashing against rocks or water warm enough to step in without shoes. It did have a large rock pool area, plenty of currents to take footage of, A wilkos nearby to buy masking tape, and a crew mate who didn’t realise his boots weren’t waterproof.

It was a genuinely fun experience, and all we needed to gather the mountain of footage we captured was, a way to get to a beach and our phone cameras. It won’t get us anything fancy like underwater shots, but we got what we needed and it’s a great excuse to get out of the office. To be around nature and remind yourself how things really behave not how things on a screen behaves. It’s very hard to not notice a component of an effect when the splash hits you in the face. We were able to capture the waves, slow motion footage of us throwing things into the water, and currents moving beside rocks. As an independent creator there is an extra benefit to going out in the world to shoot footage as well. Everything we captured can then be sold onto stock footage websites for some extra passive income.

Before we went out, we made sure to make a list of footage to capture. One of our goals was to capture something coming out of the water. This is where the nearby wilkos came in handy. We needed to get something that we could use to hoist a rock out of the water. We settled on using masking tape to make a makeshift crane. It took us three attempts to make a sturdy harness, but we eventually managed it and was able to record what we need.

So how does capturing all this footage really help us out? Well I have a scene with a box of water and a tentacle emerging. I’m going to use the rock footage about to make an effect to compliment the tentacle and give it’s emerging more Oomf. Now instinctively I would do a water simulation in Houdini and translate it to UE4 for the base of my particle effect. However, I am following Rares process and in the video they said that they used the video footage they captured from Cornwall to make the textures for the splash effect, that means that not only are we not allowed to use Houdini sims we also need to use the video to form the base textures that we use. Which makes things a little trickier but not impossible.

First thing to notice with the way the tentacle emerges is that there are two separate points where the tentacle breaches the water. One of which is slightly earlier than the second one. This presents an interesting option. We can either have two smaller effects spawn when the tentacle emerges or one big effect. This option is heavily dictated by your model, as well as the scale. When I first envisioned this effect, I was thinking of those pictures of whales breaching the water where mist and water spray flies everywhere. However, our models’ girth and the size of the checkerboard underneath the tentacle implies that the poor thing doesn’t have a lot of mass and won’t make a massively powerful breach, which means we are going to have to go with the two separate effects spawning. The question now is what this breach should look like. That is where our masking taped rock comes in handy.

What we can observe from the video is that when it breaches there will be a large solid white mass of water that gets bigger and then erodes, water particles that spray upwards, and water dripping off the rock once it’s breached.

For the emerging effect this is what I came up with:

I used that video above to make the base texture that drove the two solid areas of the effect.

The effect has 4 emitters, 2 emitters using that texture, both of which growing using size over life to give the impression of something coming out the water. 1 Emitter uses a splash texture growing to convey some force and the final emitter shoots out tiny droplets.

The dripping water effect uses two GPU emitters, both of which use the same material. They are driven by a constant acceleration and a size by life module. One emitter has a shorter lifespan and starts larger. Both emitters also use collision so the droplets drip down the tentacle.

I used sockets in the skeletal mesh to position where the effects are going to spawn and triggered them in a blueprint. This is what it looks like all together.

Realistically thinking, what even is the technique I used to make this effect? Going to the beach and filming real world reference? It’s one of those things that seems so obvious, but so is gathering reference and that is a step that I constantly used to skip because in my inexperience I never considered it important. Why get 20 videos of something when I can just look at this one video. Getting away from a computer adds an extra degree of separation between you and the effect. Reference gathering stops being clicking on a video and becomes actually interacting with an environment. The mindset changes from “this does this” to “if I do this it does this, which means when this is doing this then this acts this way”. What worked great with this technique was that I didn’t need anything expensive to make this effect. All I needed was a phone camera, a beach, and some good old-fashioned creative thinking. If I didn’t have a beach, I would use a river, or a lake or a bowl of water on the street. I won’t be able to go out and get footage of lightning bolts or explosions this way but for innocuous every day things like water, rain, leaves and wind this separation from a desk helps deliver much better results.

2. Using a Macro lens to make stylised Smoke:

Our next trick was to use macro lenses to film match stick smoke up close and use it to create a stylised streaky smoke effect. I don’t have a high-end slow-motion camera with various lenses to capture the smoke itself. What I do have is a phone camera that shoots 2 seconds of slow-motion footage and a £10 set of Macro phone lenses from an online retailer, and at that was all I really needed to make the effect that I wanted to make.

The filming itself is a multiple person job if you don’t have a tripod to hold the camera. I tried to do it myself in my back garden, but I didn’t have the speed to light the match and position the camera before my phones arbitrary amount of slow motion ran out. So I grabbed my wonderful assistant’s Burgundy and Montague. We went behind the back of an old abandoned building I used my coat as a back drop, we used blue tac to hold up the matches and some slabs of concrete to keep the camera steady. We didn’t have equipment that was made for using Macro lenses, so we had to get extremely close to the matches for the camera to focus. After burning through an entire box of matches we were able to get this.

It was a little bit blurry, but it gave me the shapes I needed. Once I got home, I put this into Adobe After Effects to clean it up. I cropped the video, now I was just seeing that top right corner where the smoke is, cut away most of the frames giving me 49 of the best frames to work with. Used an unsharp mask filter to emphasize the shapes, made the video greyscale from using the saturation controls and played around with the levels to make the values pop more. Finally, I added a Mask layer with a multiply blend to clamp each frame to a shape.

When I was happy, I exported it into an image sequence, pieced it all together in Substance Designer adding in some extra overlay blending and the final flip book looked like this:

In UE4 I made a material that uses the flipbook as an alpha between a light green and a dark green, I also added a bit of UV distortion and an Alpha erode to finish off the look. My particle system has 2 emitters. The first emitter is our flip book emitter. I use a curve to go through the frames because I want to ease the smoke in and out. I also gave each particle a unique rotation rate to stop them looking uniform. The second emitter is a mist emitter I made for the tentacle section above. I just repurposed it to enhance my flipbook.

The final effect looks like the below.

The candle mesh was made by Burgundy (Red Freeman)

What worked great about this technique is that I was able to get a stylised smoke look in a fraction of the time it would of taken me to make the texture or simulation manually. On the day it took us roughly 45 minutes to film about 6 or 7 videos of us burning a cluster of four matches. When you compare this to the length of time it would take to either hand draw the smoke or tweak a simulation its an insane saving on resources. The main downside to this however is that the smoke doesn’t look as sharp or as defined as I could get it through drawing or simulations. That isn’t a good argument against this technique however because our methods for filming this was heavily freestyled. If instead of me grabbing my assistants and us heading straight to a field/abandoned building somewhere we sat down and planned out a set. Maybe used a black shoebox as a back drop and picked a day that wasn’t insanely windy. We could have gotten a lot of extra quality in our video. The processing after would be just the same as if you were processing footage of a simulation so no extra time was used up in that department and the process of setting this up in UE4 would be similar with all three methods.

What I wanted to convey the most in this blog post is how small and indie studios can achieve the results that larger studios can while only having a fraction of the resources. There’s actually a great scene in the black pearl where Jack and will are talking about this exact thing.

I can’t take a 5+ hour drive to Cornwall, I can take a 30-minute drive to Redcar. I can’t use a high-end scenic camera to film waves smashing against rocks, I can use masking tape to construct the environmental effects I want to capture and film it on my phone camera. I can’t film smoke on a high-end slow-motion camera, but I can use a phone with a macro lens. All that’s needed is a little bit of creative thinking.

I want to take the opportunity to give a huge thank you to my assistants Burgundy (Red Freeman) and Montague (James Montague) for helping me out with this issue. I could not of done it without you guys.

Hopefully whether you are brand new RT VFX artist, or you’ve been doing it for several years there was something here that helped you out. If you do find any creative uses for the things, I’ve talked about here please email me a gif at – RichardVFXfanmail@gmail.com or tweet me @stokes_richard. I’d love to see what you can create. Also check back in a couple weeks where I blow up my final year project.