top of page

Portfolio piece: Curfew

First person, Sneak 'em up level made in Unreal Engine inspired by the Dishonored series.

My goals with this project:

  • focus on multiple paths

  • explore verticality 

  • learn from the Dishonored series that inspires me

  • work with restrictions in assets

Curfew Overview

Level production

Research

Researching and gathering information.

Dishonored is one of my favourite series and I wanted to learn from their impressive level design.

Pre production

ideation and iteration

Paperdesign

I spent a lot of time on the paper design of the level, redrawing it several times, going back an forth between pen and paper and blocking in Unreal. The most important thing here was to define each area and giving them a reason to exist.

Blockout

This was day 4 of pre-production, a quick blockout using brushes.

The start area

The next day I defined the more special areas. This is the framing of the goal from a basement in which the player starts.

The gate

This is area 2, where the player is introduced to guards. The mannequins served both as guard placeholders and size references.

Residential district

Area 3. Moving in and out of the buildings here was very important to set up early as I wanted it to flow nicely.

The hotel

The last area, difficult to get right. I experimented a lot with which windows in the hotel building was best as the goal. I wanted it to be high up but not unrealisticly high up based on the surrounding buildings. 

The guard

With some extra time on my hands I gave the guard AI some features that made it more recognizable, something a lot of playtesters enjoyed. The lantern's light was also a good foreshadow for example when the guard was about to round a corner.

Video: guard behavior

Here is a video showing the guard's behavior. It goes from idle to chasing you, when you turn the corner it walks up to the last place it saw you and then searches at a random point in range before going back to it's original position.

Base material and lights

I wanted to place some basic guiding lights in each area but because the "detail lighting" setting in Unreal makes everything grey, including lights, I decided to apply a grey material to the brushes. This way it was more gentle on the eyes and the lights' colors were visible. 

Alpha

Pickups

My goal was to have all core mechanics to be somewhat up and running in alpha, and I found I had some extra time to implement pickups, something I otherwise would have kept on the wishlist. But this added another layer of exploration for players, so I am glad I added it.

Extra meshes

I also created some simple meshes that I felt was missing from the free asset pack that I chose to work with, so I for example created this safe - making sure to always have references and the mannequin for scale.

Interiors

For alpha I made sure to have all interiors somewhat meshed out as it felt easier adjusting the outside to the inside than the opposite. As the player would weave in and out of the buildings I wanted to make sure it felt natural and flowed well.

Alpha blockout complete

Towards the end of alpha I had more or less everything important in place, things were going according to plan.

Video: alpha playthrough

Here is a video I recorded, a full playthrough of the level during the end of alpha. During the recording I noticed some errors that I later corrected. 

The full video is 3 minutes long, sped up slightly.

Level wireframe

The first thing I wanted to do during beta was to refine my paper design into a digital overview. A top-down wireframe image was best to keep the scale correct, and made it easy to draw over in Photoshop.

Beta

Digital overview

This is a drawover of the wireframe image.

I believe making overviews to present a project like this to others is as important as the level itself. It was a challenge to make it readable and clean while having all necessary information in it. 

Area 1 meshed out

During beta I also set up a goal to have all four areas of the level meshed out in four days, one day per area, four hours per day. 

Area 2 meshed out

Before I set this goal I have to admit, I thought it was a waste of time to mesh out a level like this. But in hindsight I am very glad I stuck to it, it made my world much more believable and allowed me to guide the player in less obvious ways with props that I did not think of before.

Area 3 meshed out

The most difficult thing to mesh out in this level was the ground. With slopes and different heights it took a lot of time to get right, so I always started ground up in each area. In doing the most painful things first it made the rest feel like a breeze.

Area 4 meshed out

The last area was trickier than anticipated. I redesigned the hotel building a bit as I realized I had not used a reference for it before so it naturally felt a bit off. Going back to the drawing board here was a good change of pace, and in the end made the building more believable.

A useless landmark

In the first area of the level I had deliberately placed a factory building behind this alley - something I thought served as an important landmark for the player. But while playtesting I realized that no one noticed it or needed it to find their bearings, so I removed it.

Overview refinement

During this time I also experimented more with my overview, trying to make it feel like it was part of a bigger map. I also went back and forth on how to show the verticality and insides of the buildings, I will talk more about this further down.

Backdrop

For gold I spent some time making the area feel bigger in-game by layering buildings as a backdrop, especially to the western side as it was supposed to be a coastal city. It was very rewarding to hear "this looks huge" from the last wave of playtesters. 

Gold

Final overview

I settled on having dark grey signifying a lower level and lighter grey being higher up, something that felt like such a small detail but it did a lot in aiding the readability.

And for the paths, it was difficult to not draw too many lines so I treated it more as a hint to the possible paths instead.

Video: gold playthrough

A similar playthrough video as the video in the alpha stage. This was recorded before I had edited the post process (final result in the image below). 

Around 7 minutes long, no need to watch all of it.

bottom of page