Off The Map. Dream to Reality.

So this post covers the process Solstice Studios went through from start to finish for Off the map, a project run by the British Library and GameCity.

The theme is Shakespeare, with a choice of the plays; a Midsummers night Dream, Macbeth and the Tempest. EndFragment

But, first, lets introduce the team.

• Charlotte Watson: I'm the teams Technical and Environment artist, working mostly on creating assets, terrain, blueprints and optimising the level. I also worked a lot on the level design, taking 2D maps and making them 3D, as well as placing the assets in the scene to create a visual narrative. ​

• Brandon Lee Welch: I'm an aspiring Concept Artist/Character designer originally from South Africa, I'm serving as a Character Artist/Concept Artist for our Off the Map Project

• Julie Thomas: I'm originally from North Wales, With a love for the colour green. I'm taking a main role in the projects Asset design and creation for our off the map project.

• Hannah Brown: I'm one of the Environment Artists in our team, I have been working alongside Becky, developing the look and atmosphere of our Midsummers night dream level.

• Becky Green: I am one of the environment artists along with Hannah; I work on the concepts for our level, as well as creating 3D environmental assets. I aim to express the story of midsummer’s nights dream using environmental storytelling, not just dialogue and exposition."

• Charlie Wallace, I'm working as a character artist and level designer on the game. I've worked primarily on the character of Titania but also had a hand in storyboarding and creating the level layout and functionality.

And for more information on the process please see the rest of the blog, as well as our individual blog sites which can be found below:

Charlotte: http://atomicartist42.wix.com/atomicartist42

Becky: http://thestudentsketchbook.wix.com/gameart

Brandon: http://kline621.wix.com/brandonwelch621art

Hannah: http://hannahbrowndmuga2.blogspot.co.uk/

Julie: http://julesgameart.wix.com/julesgameart

Charlie: http://powerdohnut.wix.com/artofwallace

Jumping straight in:

The first thing we did was call a group meeting. In this we wanted to cover three important decisions; which play we wanted to work on, what style we wanted to go for, and what kind of game we wanted to make.

This involved a lot of brainstorming, and so we decided to let everyone go away and come back with their own pitches for what they wanted to do.

The first thing we decided upon was that we wanted to focus more on creating an immersive environment than a complicated game. This is because none of us are programmers, so we felt it was best to focus on our strengths, but of course push ourselves to get some game-play into the level. Instantly then we started to focus in on games such as Dear Esther, Everybody's gone to the rapture, and Journey, as these games emulate what we wanted to achieve.

This decision also helped us to choose which play we wanted to make. A Midsummer's night dream. We felt it worked best as a walking simulator due it's natural beauty as an environment and it's strong imagery and poignant themes of mystery and comedy.

We then started researching the play and putting together mood-boards to come up with an art direction early on. Charlotte began work on a style guide, and everyone was tasked with reading the original play so that we could stay as true to the story as possible, and capture it's essence in our work.

The Idea behind the project

The play has so many characters and different story-lines running through it, that we decided it would be best if we alluded to it's main plot points using visual cues, such as Bottoms fountain, and the purple flowers. We initially split the play into four main acts, which are represented as clearings. Throughout the level we also initially wanted the player to slowly realise how fake everything was in a dream like way, as their world becomes a theatre set.

However this had to be cut down due to time constraints, so we focused on two clearings, leaving the ending very open, so that it could be continued in the future. The first clearing is you waking up in the forest with no memory of the night before (much like what happens to our young lovers), you're introduced to Puck, and see the Lion Statue (a homage to the band of performers), as you go down the path you come across Puck again and again, coming across a small alcove with the purple flowers he uses to cause his mischief.

The final clearing is Titania asleep, with her Bottom (the statue). Our main goal shifted a bit from the start of this project to become more about capturing the atmosphere and main story of the play, as opposed to more specific plot points. We also ended up scrapping the forest becoming more like a theatre set, (except for our prop in the first clearing), as we felt this didn't add much, and so we brought comedy in with Puck and his various quips throughout the level.

We decided to make it first person so that we could have the environment as this visual unfolding of a story that anyone can get to experience. Puck made sense as a narrator for our level, as he's the main architect of the mischief that runs the story, and he's a great tool for adding some comedy to the level.

Before getting any further into our creative process, a more tedious yet incredibly necessary task needed to be tackled first;

Planning

We used a Gantt chart to keep track of the project milestones as a whole, then had a weekly time-table to make sure individuals were on task.

We also had an asset list which was divided into clearings, and had check points to show where people were at with completing their assets. One of these checks was to make sure they’d been checked by the team leader to make sure style was kept consistent throughout.

A technical document was also created to make sure assets were created sensibly and would not affect our games frame-rate.

Style

From the get-go we all agreed on going for a semi-stylised environment. Taking inspiration from games such as Dishonoured and Dragon Age. However we realised we were being too vague on this, and so as the level progressed we ended up going for a more realistic approach, choosing to stylise the level through colour and visual effects, as opposed to shape and detail. Post Processing played a big part in this:

(shot of level with and without).

Becky and Hannah were in charge of coming up with our levels overall look, as well as any details we wanted to work out early on.

They did a fantastic job, and whilst unfortunately we didn't get to create everything they concepted, we owe a lot of our levels final mood and look to them.

Concepts

To ensure we maintained a consistent style throughout the project we made sure to continuously post on this blog with what we were doing and give each other feedback on the work, we also made sure that people who worked on similar tasks such as foliage creation, made sure to check their work with each other, before calling it complete.

Level Design

We decided on our levels layout relatively early on, thanks to the collaborative work of Charlie and Charlotte (and of course the whole teams ideas and feedback). Charlie designed the path, and Charlotte visualised it into a white-box to get a better sense for scale, and to start working out how we were going to tackle the task of creating a 3D exterior environment. The environments layout and terrain evolved a lot throughout the project like so:

Charlie's Maps/initial Paint-overs:

Maps

Initial Story-board

Early Engine White-box

Developed in Max

Replaced with Terrain in UE4 4.8.3

Simple block-outs to get a feel for creating a forest

Getting the terrain finalised was a huge leap forwards in the project. With this down, all we needed to do was populate our world. Of course as mentioned in planning, we have an asset list, Julie, Hannah and Charlotte all worked hard to create assets to populate our level, whilst Becky was our main foliage girl, and a great job she did of it too.

Developing the environment also involved a lot of step backs and re-evaluating what we had so far, and deciding if we wanted to change/add/remove anything. One of the ways we did this was by taking screenshots of our level and doing paint-overs. As well as continuously doing block-outs within the level to work out if we wanted to change the composition, add anything in etc. a good example of this is how the final clearing developed using simple cubes to start working out a new composition for it, this also allowed us to get team feedback on an idea, before wasting time creating it only to have it scrapped.

Paint-over of a bank to visualise populating the area.

Initial paint-overs of clearing 1 and 2, again to help visualise populating the forest.

Finally a demonstration of how much some foliage and assets can transform an area from a dull bank to a lively clearing:

A splash of Field Research

During this project we took a few trips to different natural parks to gather research and textures. These trips were very useful, and were especially useful in creating convincing terrain textures and foliage for our forest.

Here are some pictures of our trips to Bradgate and Abbey Park:

Models

When it came to deciding what to make for our level we turned to the original play quite a lot. For example we included assets such as the wild purple pansies, since these are the flower; love in idleness, which Puck uses to spread mayhem throughout the forest.

A 3-legged stool because it’s mentioned that Puck uses one to cause mischief, and the lion statues because the lion is an important character in the play, and play within a play, as a reference to the mechanicals (actors), and the Lion that scares away Thisbe from the wall.

Modelling went on from week 2 all the way until the final week. Models were created in a multitude of different ways depending on the asset.

For the foliage we used master sheets, grouping different bushes onto one texture sheet to save memory. We were also careful to avoid overdraw with our alpha's, as this would impact frame-rate hugely.

The main programmes we used to create our models were 3Ds Max, Z-brush, Photoshop, X-normal and Substance Painter. Some assets were concepted before being created, and some were just made or concepted during creation.

For example our lantern and tree-stumps:

Here are some of the models we created in this project:

Throughout the project our two character artists were working hard to concept and develop Puck and Titania. Brandon worked on Puck and Charlie worked on Titania

Both guys started out by concepting, using this character mood-board Charlotte put together to help them out:

This was to get them to stay within a certain style, and we all agreed that we loved Brian Froud's impish depiction of the fair folk, plus we felt it worked well with the classical nature of the play. We also all agreed that the fairies should both be designed with an Indian Influence to their designs, as the fairies in the play are mentioned as coming from India originally.

Brandons Concepts:

He then moved on to sculpting Puck:

Before creating his low poly:

And then applying textures, rigging him, and putting him in game with some idle animations

Charlie's Process was very similar as all character artists tend to follow a similar pipeline.

He also initially concepted Oberon, however with this character scrapped he turned his full attention to Titania.

Charlotte provided an initial concept for Oberon to get him started:

He then started concepting him:

Before moving onto Titania:

Sculpting her:

And texturing her

Before also rigging and giving her a sleep animation for in game:

UI

Whilst the game-play in our level seems relatively simple, it took a lot of hard-work and scrapped ideas before we ended up with our final product.

Becky, Hannah and Charlie all worked on adding pieces of UI to our level, whilst Charlotte implemented it.

We were also lucky enough to have a voice actor, Hannah's fiancée, who provided us with the audio, which Brandon edited for us. We were really pleased to have it in our level as we think it gives it that extra bit of life.

Game-play

We wanted to use game-play as an aid to make the player interact with the world and move them through the environment, rather than have it as the main focus. Therefore we kept it mainly audio based, with simple instructions for the player to follow. It was really cool letting other people play the game and seeing what came intuitively and what needed a bit more pushing.

For example we originally only had Puck give one instruction for the player to find a script. However people kept going back to him expecting a second, therefore in response we added the second instruction or statement rather, which was to let them know they could know exit the clearing.

Just to give an example of how complicated engine blueprints can get, here is a breakdown of our simple, 'highlight script, pick it up and leave the clearing' game-play became.

Check to see if actor has collided with sphere collision around character:

Creates material instance on object (creates highlight material):

Detects if you are looking at the object:

Fades out highlight the further away you are:

Pick up the object and create widget for it:

Detects if you have picked up object and if so triggers matinee event:

And the material for any object that will use this blueprint:

Lighting the Level

Feedback on lighting was vital for pushing our level from mediocre to impressive. Lighting makes a huge difference to the overall look and feel of a level, and we all feel like Charlotte did a great job with it. She struggled initially to balance mysterious, night-time and shadowy with welcoming and most of all, visible, however after lots of iteration, we all feel like she achieved a good balance.

We initially were going to have our level set at different times of day, however we scrapped this due to a level of complexity we weren't prepared enough for, but also due to the fact that we felt sticking to midnight made more sense in capturing the plays charming night-time atmosphere.

Overall she used many different type of lights in the level to achieve the effect you see in the final version. Along with different light functions to get the water caustic and dappled light effects you see in the first clearing and by the pool.

She went from this:

To this:

To this:

Post Processing was also super important for getting the final atmosphere right for our level, it also became a lot of the lighting once Charlotte had finalised it without post processing, which can clearly be seen with this comparison:

It took our level from bland to magical. As our tutor said, “The environment should be like a normal forest x100 on the magical scale, that means saturation, colours, and inviting lights/music.”

Optimising

Optimising a level is something that is important to think about from the get-go. Which is why from the very beginning we made sure to share texture sheets where we could, channel pack maps, reduce over-draw and generally let Charlotte work her engine magic to keep everything running at at least 30 fps.

Something that we’re all very glad we added to the level are the more specialised materials and effects Charlotte worked on. We think they add a bit more magic to the environment. We chose to add in the stream as felt it would make for a more magical location for Titania to sleep by, and we could use the waterfall to create a composition that led your eye towards her.

Extras

After much thought we decided on Solstice Studio’s as our team name, and after some iteration:

We ended up with our final logo:

Becky did a great job coming up with an image for our game’s name, SummerTide, and all in all I'm glad to see our project take on a more professional appearance.

Trailer

Finally we needed to re-do our trailer to meet with the competition’s requirements. Charlotte storyboarded this:

And then edited the trailer so that we ended up with this:

Hopefully it shows our game off to its greatest potential, we’re proud of it regardless.

All in all it’s been a wild journey, but we’re all really pleased with the outcome. The work didn’t end at submission for DMU though, we polished a few things afterwards, and came up with our Team Logo and Game name for the competition. Of course we hope we win but we learnt so much with this project, that at the end of the day, we’re all just proud to have made something we’re excited to share with friends and family.

And finally some screenshots from the final level: