The Endeavour - Harnessing The power Of Collaboration

In this article we're joined by the team behind The Endeavour, a lushious ship wreckage environment created for a Beyond Extent monthly challenge. We pick their brains on how the project came together, how they collaborated effectively as a team and other more technical aspects such as their material workflow using tiling materials, how they created their foliage and shooting the cinematic. Strap yourselves in because there’s a ton of knowledge here!

Intro

Adam: Hi, I’m Adam Rzatkowski, I’m a Junior 3D Artist at Dovetail Games. I’ve been working in the industry for just over 6 months now, and it’s been a pretty wild ride every day. I am a self taught artist and most of what I know comes from YT, forums, and reading software manuals which is as boring as it sounds. Lately I joined the Beyond Extent community and that really expanded my horizons in terms of env art and put me in touch with some amazing artists.

Scot: Hey, I’m Scot Daniel Burns, I’m from Scotland, UK and I’m currently trying to get my first environment art position in Games. I've been improving and creating new environments in my free time and doing group projects like this one since studying Computer Games Art and Animation in University. I really enjoy trying to create a believable world and trying to tell a story through the environments I make and challenges like this one are a great opportunity to make a nice environment in a small space of time.

Dylan: Hello! I’m Dylan Burgess, a Junior Environment Artist at Raven Software.  I’m new in the industry, with an academic background in studio art from my BFA at the University of Florida.  Around school I taught myself game art using online resources and developed my 3D environment art portfolio.  I joined the Beyond Extent community to get more connected to other artists in the games industry and to gain some experience through the team challenges they organize.

Cristian: Hey Everyone! My name is Cristian Bercea and I'm currently working as a 3D artist at Dynamight Studios on Fractured MMO. I'm originally from Romania, but I currently live in Tuscany, Italy. I got into the world of 3D thanks to my university path, Industrial Design in Florence. After that I started learning more and more about the game industry and took more specialised courses in this direction. Thanks to Patreon I discovered great communities and artists and besides learning and getting a lot of good tips at the same time you can support their work. This is my second team challenge  on the BeyondExtent community and I love how it keeps getting better and growing in an organic healthy way!

Guillem: Hi! I’m Guillem Pavón Falcón, I’m from Girona, a city near Barcelona, and currently work as a 3D Artist at EGM Studio, a company focused on automotive marketing. I have a degree in video games art done at ENTI-UB and I love to create environments. I didn’t like to study and didn’t know what to do with my life, until I found that awesome degree that helped me develop my creative side and explore the cgi world. Once in the correct road, I’m trying to learn as much as I can to be able to do anything required to create immersive and astonishing environments.

What Inspired The Scene?

Adam: I believe we started with one of the suggestions provided in the challenge references. I’ve not done a proper sci-fi scene before, and immediately gravitated to the idea of a crashed spaceship that has been repurposed by some shady local people. We voted on the theming in regards to time (past/present/future), theming, and biome, and in the end the team seemed quite aligned in our vision. Researching concepts and ideas we found that most ships were crashed on deserts, so I was keen to take the idea to a different, colder and more green area.

Scot: The initial theme for the teams for the Beyond Extent challenge was “Shady Reststop”, with that in mind we gathered references individually then presented our refs/ideas to each other in a group call. The group call then sparked some additional changes to the refs which were then quickly photo bashed together to represent the general idea we wanted to aim for.

Dylan: The concepts that our team was pulled towards most were similar to the environments found in Death Stranding (the rocky Icelandic fields with a cooler color palette). We spun our piece off of that angle, and developed some compositions that could give the landscape some presence and emphasize the isolation of our crashed ship. Personally, I found the theme we settled on as a good chance to develop some reusable landscape assets to push the scope as far as we could with our relatively short deadline.

Planning The Scene and Assets

Scot: Initially we broke down the scene into about 4 areas after we had decided on a rough concept. The Interior, grill area, landscape elements and the main structure. Early on we had some team changes, this prompted me to organise our Trello boards and create cards for each element of the scene that had to be made, allowing team members to assign themselves to assets and areas of the scene.

Adam: Personally I just focused on making sure the engine is big and detailed enough to pass as a building. I did not think about expanding beyond that until I had the engine locked in. Seeing the other people progress with their areas was very inspiring and jogged my imagination a lot, I am probably to blame for expanding the interior WAY beyond the scope.

Dylan: Everyone in the team picked a part of the scene that they’d be interested in developing after discussing it as a group. We ended up structuring our areas of interest individually while communicating our ideas throughout the process to make sure they meshed together properly. I personally focused on foliage and rocks, some of the landscape layout, and some of the small props for scatter.

Given only a month for the project, how did you refine your scope, were there any big cuts?

Scot: After the initial blockout, each team had a meeting with the Beyond Extent’s challenge moderators, Timothy Dries, Lloyd James and Cairo Goodbrand. This was just an initial check in with the teams, we spoke about the scope at that stage and removed all the interior elements to the scene. Funnily enough, the scene at this point was probably near half the size of the final scene which has slowly grown over the month in a more organic way.

Dylan: Our scope started out somewhat close to our final scene while initially concepting, but to make sure we would have a proper deliverable at the end we cut back on that scope considerably. We set some stretch goals that eventually led to us reaching where we ended up.

Planning a Colour Scheme

Adam: I think that was established in the early stages of concepting? The general idea was cold environment with a cold metallic structure, and warm lights to draw in the attention, but everything is cold and grey when you’re making an engine!

Scot: Adding to what Adam has said, there was a heavy re-use of tilables and trim materials with minor colour variation. The idea was the metal could have all been salvaged from the wreck potentially so a similar look was definitely on purpose across the assets from a storytelling

perspective and also a time-saving one. At the end of the month we had the discussion that it was a bit monotone which is why some color was injected with paint materials and coloured graffiti.

What was the team composition and how did you play to each other's strengths to finish this project?

Scot: I think pretty organically Myself, Adam and Dylan settled into what we knew we could contribute best to early on, Dylan pretty much handled the rocks and foliage early on and added to the props I had already created to really fill the scene towards the end of the project. Adam established a lot of technical aspects pretty early too with his Master Material and planning out a trim sheet to use around the scene, primarily on his engine and my ship hull.

Pre-Production

Scot: I did some initial photobash images for the initial rough ideas for the scene, these were mostly for a broad idea to pitch to the team.

During the challenge I also did a couple sketches when trying to get some ideas, however most of the “sketching” was iterations of various 3D designs rather than traditional sketches.

Dylan: For individual assets, we created reference boards to guide any decisions we made. In the case of rocks and foliage, I looked at references from Iceland itself for inspiration and then spun off of those references in my own way since we were going for something a bit more alien.

Materials and Textures

Adam: I am a huge fan of a good trim sheet, and knowing that we would be working on sci-fi assets, we agreed that going (at least somewhat) the POM/mesh decal route was reasonable considering the size of the assets and time we had. It also meant that if one person made a solid trim or a decal set, it would open up options for the whole team to create more assets. With that in mind I got together with Scot early on and we had a look at some jet engines and references that broke down the spaceship and engine into some iconic and recognizable shapes and characteristics. I then put that together in the form of a master trim sheet and a tileable material that it would blend with. I was aiming for fairly plain and realistic steel, pipes, and cables not to get too far away from what is familiar to the average person. The early planning really paid off, as the trim saved us literally on the last day as we were able to drop in some large assets to fill out the space very quickly. I think in the end there are very few unique texture sets save for small clutter. 

Dylan: When it came to the foliage, two techniques were used for the textures. For the smaller clumps of plants and grass I used unique bakes for cards. For the larger plants in the incubators I created an atlas containing individual leaf variations and the stem texture, which allowed reuse of each leaf throughout the plant.

Average Week Working On The Project

Adam: For me, at least initially, it was a matter of laser focus on the parts I took on: setting up the tech, shaders, pipeline etc, and moving onto the engine blockout, modeling and assembly. I’d finish work and sit down for 2-3h in the evening on the daily just chipping away at those tasks, weekends were a full on work day on the project.

Scot: I had a lot of free time during this project so I took on a lot of the level layout throughout, so as the project went and assets were needed I would create a barebones version to have as placeholder in the scene which was later improved by myself or remade by my teammates. My main task the first week or so was creating the design of the crashed ship and finishing that, with tweaks throughout the project. Other than that I spent a lot of time in-engine trying to make the scene feel believable as a real space that could exist and adjusting the scene accordingly. 

Other than that, chipping away at the asset list we had on Trello was always something I could do during the week.

Assigning Assets To Team Members

Scot: Everyone was present when we broke the scene into chunks and then each person said what they would prefer to work on with a first and second choice, which they were then assigned as best we could to each area. We had decided breaking down the scene into areas would be best to stop any asset overlap, this changed as a member of the team had to leave early, and we all had a range of how much time we could commit so the areas were assigned according to that as best we could. Later it became more organic as mentioned previously with each scene element having its own trello card and members could assign themselves if they had the time to tackle something.

Material Pipeline

Adam: This was something I tackled first, as I didn’t want people to do too much work before we had a solid pipeline established. Bearing in mind we would have a lot of trim and tiling, as well as POM and mesh decals, I set up the textures to be Base Colour with Alpha, AO/Roughness/Metallic/Height packed into RGBA and then Normal map. With these in mind I built a master shader that had a base layer with all the typical controls over mapping, roughness, opacity masking, hue and normal power, and further 3 layers of entirely separate materials that could be painted on top of it. The materials would be vertex painted in, and would utilise a dedicated RGB mask that could be used either as transition mask for height lerp, or as dedicated mask that behaved almost like a decal. It took me a good couple of days to figure it out, but the end result was that any given asset could have up to 4 materials on it, layered and painted with full freedom, which allowed for heaps of customisation to just plain metal plates and struts. My aim here was to make sure everyone could adapt it to their own style of asset creation, be it dedicated textures, tiling ones, or a mix of any other techniques. That material was then adopted to Dylan’s rocks with added top moss cover and bottom dirt blend.

Foliage Creation

Dylan: I used two techniques for the foliage based on the scale of the plants. For the small foliage, I modeled the entire plant in blender as a high poly and then textured those plants in Substance Painter. Once I had the textured high poly created, I baked alpha cards from those elements for the environment.

For the large plants in the incubator, I broke apart the plants into individual leaves and the stem. I modeled the basic leaf shape and the sculpted details in Zbrush. For the leaf, I created a gradient of versions from living to dead to use throughout the plant. When I had a good collection of sculpted leaves, I laid them out and baked them onto a texture atlas. I included a tiling stem trim on the side of the atlas. Once I had textured the leaf altas in Substance, I used it to build the plant up with bend cards in Blender.

Defining The Art Direction

Scot: The art direction from the start was to have a semi-real appearance due to the time constraints and scale of the project. The colour palette was inspired by Halo and to some extent The Outer Worlds. Other than that we had the general idea that the scene should feel like it built from metal and parts of the main spaceship so that influenced some designs, for example the grill behind the bar is made from a small engine, the floor metal is partly made from old ship panels and the most visible one is the “Endeavour” sign which is made from what looks like a ripped of panel from the ship with the Endeavour branding from the original ship still on the panel behind the neon retrofit.

How did the team create the best practices and ensure everyone was able to follow them?

Scot: Based on Adams initial thoughts on his master material, guidelines for texture packing, required maps and texel density were added to the team's Trello.

Were there any reviewing processes?

Dylan: Due to the nature of the challenge and the schedules of the participants, we had a few meetings to review progress that were organically organized. Generally this happened on the weekend when people were available. Otherwise, any critique notes were jotted down in the groups chat for people to address as they came up.

Scot: Like Dylan said there were weekly meetings that were decided on based on polls most of the time, we also had a trello board where we could mark our work for feedback in which others could comment and leave some feedback.

Did team leverage master materials/shaders to improve production?

Scot: The majority of the materials used a master material Adam created early on in the project which he tweaked throughout the project to accommodate the needs of the project. The material could be used for unique assets and also for pretty advanced vertex paint control over the assets using an RGB mask.

Lighting Process

The lighting for this scene is pretty standard, since it's an outdoor scene. We have a single directional light which is responsible for the majority of the scene lighting. Inside the settlement we have a handful of coloured point lights which simulate the neon and emergency lights at the back of the interior, on top of those there are a few larger spotlights with shadow casting turned off which serve as fill lights to brighten up some really dark areas. In addition the directional light has a material function attached which simulates rolling cloud shadows in the scene which can be animated but for our renders are static so we could position the clouds strategically in the composition. On an FX level we have volumetric fog in the scene to give distance between the ship and the mountains and also manually placed fog cards to emphasise the depth even more to try and achieve a grand scale around the crashed ship by creating a clear atmospheric perspective in the scene.

Planning and Shooting The Cinematic

Adam: I have to give credit to Scot for a lot of what’s in the cinematic. He laid out the main shots of the scenes, which defined clear areas of interest. I have then just grabbed a few camera cuts and panned over the areas and angles he outlined to show off some of the light and material work we did in motion and with some light FX touches. I am by no means knowledgeable in cinematography, so I stuck to some basics I’ve read ages ago: linear camera motion, level framing, keeping the movement direction as consistent as possible between the shots. Other than that I tried to introduce the scene in the order I thought the player would explore the area in a game setting, from the outside in and back out. Music kind of aligned by a happy accident (we were really pushed for time at this point haha).

Were there any ideas from pre/early-production that didn’t pan out or needed to be changed?

Scot: Initially I think we planned to have a more barren landscape, the general goal was an icelandic look where its very rocky ground, this changed overtime to a more foliage dense landscape mostly for visual interest. In some ways the earlier idea might have been more of a challenge to execute well anyway. One element I wanted to add that never made it for the deadline was a ship in the sky that would either take off from the background or flyby above the scene, this was one of the kind of stretchgoals I never got to.

Likewise, what were some parts of production or plans that worked better than the team expected?

Dylan: The scope of our landscape was an area that many people inside and out of our team were concerned about.  To address this concern I focused on it early on and made plans to construct it using as few assets as possible, focusing on reusability.  This allowed us to create a large area relatively quickly so that more attention could be spent on the ship area near the end of production.  Having one person dedicated to landscape assets also freed up the rest of the team to focus more primarily on the things in the area of focus.

Was there anything the team felt they all learned while working on this project that would influence future projects?

Adam: I think I just solidified my belief that a good trim is half the work, and that having a solid pipeline and naming/folder structure makes working much easier, especially in a team, yes I was the naming convention police.

Scot: I think one thing I felt was enforced by this project was that working from the big picture then slowly working down to the details is a really good approach. If you get the broad picture looking good you at least have a fall back if you don't get time to flesh out the smallest of details.

Dylan: This project really showed the importance of breaking down your asset list ahead of time so that you can properly plan all of the things you want to do.  With a plan set up early, you can better determine what to prioritize for one off assets and what you can get away with as trims, tilables, and modular or reusable assets.  We saved a lot of time in the end just having spent a bit of time breaking the piece down and having discussions about our priorities.

Guillem: This project helped me, with the help of my team, to understand how the actual game workflow is done. I come from working in marketing and would like someday to make the change and get into the video games industry. When you work on marketing, you can throw 8k textures at everything to make it look pretty, but I’ve seen that in games you need to have excellent optimization skills. The most important things learned that I’ll implement in my workflow for sure are breaking down models in modular pieces, and the use of trim-sheets.

How did this improve each of you as an artist?

Adam: I think that I’ve improved on the story telling and believability layer of art. Seeing others spin their own little stories into the environment opened me up to ideas I would not have had on my own and made it easier to get creative and bounce off of them. It was also a good reminder that sometimes all you need is one common note flowing through the whole area to tie it together, like trash and graffiti, it’s incredible how much this brought the scene to life and consolidated assets made by different people.

Scot: I think this project really forced me to consider where I had to spend my time and juggling a lot at once and getting it all done on time. It was also really insightful watching the others tackle their assets and their approach on a deadline.

Cristian: This type of challenge probably is the closest that simulates the working environment and the improvements that comes with this experience are huge in every aspect, from art skills to communication skills, accepting other’s thoughts, the whole package. Even you won’t see it right away but definitely is a good progress for your growth, especially if you want to work in the future with others as a team. This may be a small step to get you more comfortable and ready when you get a chance to get into your dream studio.

Dylan: The biggest benefit from this challenge for me was the experience working in a team.  Up until this I worked primarily individually and was lacking in cooperative projects.  This was a great experience for organization and communications skills, as well as learning to divide the labor and make sure everyone on the team is getting a hand in what they want to work on.

Guillem: Seeing talented people work always helps to understand why they do what they do, and it’s a good way to learn and improve. I feel that knowing now how’s the correct workflow will allow me to start creating video game oriented environments.

If anyone on the team struggled or had problems, how did the team pull them back in and push through production?

Guillem: Well I struggled for sure. Never worked before with a source control and I was a bit overwhelmed at the beginning seeing the knowledge and skills everyone had. Scot helped me a lot, guiding me through the source control and making sure I was on the right way. 

Didn’t have a lot of time to work on this, and wanted to do a lot of things, and after talking to the team we saw that it was better to focus on a single prop. It actually helped because I had more time to work on it, and to learn and practice the workflow.

Would the team like to speak about their roles in the project and some of the growth and learning that they had while working on this project?

Adam: I’ve learned to trust and lean more on my co-creators. Working mostly on personal solo projects it’s easy to forget how important it is to trust others to do their job right, and how valuable a 2nd point of view is. Letting go of that complete control over how the scene is progressing was a challenge, but in the end I am certain I could not even get close to the end result without all the contributions and different angles everyone brought.

Scot: I was responsible for a lot of the level layout and set dressing the scene with props and decals. I had a lot of time available to me during the challenge which led to me doing a lot of different things. I also did a handful of props and worked a lot on the larger architecture of the level such as the spaceship hull, living area platforms and the bar section of the space.  One of the main things I took away from the challenge was a better understanding of time priority when it came to which assets needed priority over others and tackling each of them systematically. I also took on organisational tasks so it was interesting trying to get the team together or helping with figuring out asset workflows with them.

Cristian: With fewer hours available per week, from the beginning I chose to start with the speeder prop, and to get more out of this it should be possible to modify the mesh and reuse it in different situations in the scene eg. destroyed, spare parts, alternative colors for texture. In this way I’ve got the chance to make a little contribution to the scene and at the same time keep working on hard surface modeling.

Dylan: My primary role in this project was foliage and rock creation for the landscape, and I played a supporting role in the creation of small props and scatter. The entire process helped to strengthen some fundamentals that I've been working on, and I really got to push my ideas of reusability and working efficiently. The deadline and division of labor really allowed me to turn my focus towards a specific discipline and hone it instead of being pulled in all directions like creating a scene on my own would have me doing.

Would you be able to break down any of the renders and explain how you planned the composition of the shot?

Scot: The main consideration we had pretty early on was to create a good blockout composition, the main render never changed a huge amount throughout the project. Our primary considerations were the rule of thirds trying to get the focal point to work with that and also a consideration throughout the project was leaving enough rest space vs detail, I think we achieved this pretty well for the main shot. Another aspect was layering the scene, the more depth separation we could create the more large scale the world would look.

How did you keep in mind color composition and theory with the project as a whole?

Scot: I don't think any of us would claim to be masters of colour, my main consideration with the scene colours was keeping the scene more vibrant and colourful, I wanted a similar feel to Halo with the vibrant blue skies and green meadows. There was a stage in the project that was much more overcast and fitting to perhaps a more grounded realistic style but it just wasn't as appealing for this project so I pitched the brighter, warmer, saturated scene and the team were happy with it in the end.

Would you be able to paint over of your use of color and light in a one or a few of the renders?

Scot: Most of our scene elements were made from cold metal colours, this was balanced out by the colourful foliage surrounding the crashed ship and the colourful spray paint which covered the wreckage. The use of shadows in the scene was intentional on my part, we initially had moving cloud shadows but they were really hard to control in some way to help composition. Due to this the clouds shadows were just set to static and manually positioned to give a good balance and use them to help the composition as seen in the image. I'm sure there are proper terms for what I've tried to explain in the image but a lot of this just comes instinctually to improve on the scene.

Would you be able to break down the lighting for this project and show how to have it play well with the models and materials?

Scot: The lighting for the scene is as minimal as we could get it. The scene is using dynamic lights so there's no baking for this scene. The scene initially had a much colder feel to match the sort of icelandic landscape we were looking to achieve as a base, however a cold, overcast lighting setup doesn't show off models and roughness values particularly well and the scene was already lacking saturation/colour with the metal structures etc. The solution? We changed the lighting to a warmer tone and gave it more direction to let the metal values pop more and let it warm the scene up. The scene is lit by the skylight firstly giving a good base light value so everything is visible. A warm directional light was added with enough directionality that the engine casing cast a shadow on the platform allowing us to light that area with colourful fill lights to fake the colourful light cast from the neon signs which wouldn't have been very noticable if the directional light was cast into the engine lighting the interior.

The only other note to try get the models to look as nice as possible is to try your best to get a varied roughness detail across the full scene, sometimes you need to exaggerate the roughness values and give them higher contrast than you think they need to allow them to really pop as much as they possibly can.

What did you take into consideration in terms of composition when creating each shot?

Scot: The composition was initially planned for our main wide shot angle and once we thought that view was appealing additional camera compositions were made to show off smaller sections of the scene. Essentially treating each camera angle as its own composition whilst still trying to maintain an overall cohesiveness between them, meaning if you move a table in one camera angle to look nice it might ruin another camera shot so you have to keep in mind all the cameras at the same time and try double check to ensure you don't accidently ruin a shot in sacrifice for another.

What should we be on the look out for from you and the team in the future?

Adam: I thought I needed a break after this one, but off I went resuming my Resident Evil inspired victorian corridor, so at some point that’s coming out and unsurprisingly, it is trim sheet heavy and will have some of the tech I made for Endeavour brought in, but otherwise it’s very woody and carpety and wallpapery, none of the cold steel ;) Beyond that, I’ll be on the lookout for future BE challenges, big and small, they really push you out of your comfort zone and make you grow in ways you’d not expect.
Scot: Immediately after the challenge I made a start on a solo one, I’m currently doing a London underground environment with a hint of Metro thrown in for good measure. I’ll definitely be taking part in future team challenges and working on my portfolio, I want to show off more unique assets and lean into storytelling on my environments and props going forward! 

Dylan: I am planning on doing some more focused work on individual assets and asset kits in the future, although the timing of that is unknown since I am going through a lot of changes in life with a big move coming up.  I hope to do a lot of high quality material and prop work with a focus in sculpting when I find a bit of time.  Where possible I will be breaking these processes down for people to dig into.

Guillem: My goal is to land on an environment artist position, so I’ll keep working until I get it, no matter how long it gets me!

Tips & Advice

Adam: Collaborate! Learning from other artists directly, and sharing the things you are good at already is almost like a hack in terms of how quickly you’ll grow. Not to mention that when you do land an industry job, having that mindset and experience will really stand out and make your
life a lot easier!
Dylan: The biggest thing for me has always been finishing the things you start!  If you have a piece that you have gotten 50% of the way there, try to push through to the end.  And if you have finished something, get some feedback on it and return to it in the future for some polish.  There is always a strong impulse to finish up and move onto the next thing, but even a couple days of polish work after some feedback can really push a project over the edge.

Scot: Learn from others around you, no matter how much or little you know you will always pick up helpful things from other artists on projects or while speaking to each other online. Another really important one is always asking for help if you need it and feedback, it might be painful to hear something doesn't look the best initially, but it all helps the end result so seek out the issues and use them to improve going forward.

Guillem: Do not hesitate to ask everything you don’t know, people will want to help you more than you think!

Outro

Adam: Overall I'm super happy with how the project turned out. I developed new skills in the areas I wanted to, expanded my circle of interests towards tech art, and made some great friends on the way. Whether you are  self taught or formally educated in game dev, jumping into a project with complete strangers is bound to give you at least one "aha!" moment and make future professional engagements that much less scary.

Guillem: It’s my first ever time doing this so thanks for giving us this opportunity and thanks to the Beyond Extent community for organizing this amazing challenge and for always being so kind!

Scot: Thanks to all my teammates for being good sports through the challenge month and thanks to all the organisers over at Beyond Extent for this challenge and all the others!
Thanks for reading through our article, I hope you found something useful here!
If anyone is interested, these challenges happen around every 3 months and depending on when you're reading this the signups could be open so go check them out at Beyond Extent! 

Dylan: Challenges like these are a great opportunity to get some hands-on experience in a production-lite environment. From learning some basic source control, communication inside a small team, prioritizing tasks, and working towards a deadline, the entire experience is incredibly valuable. The added bonus is, of course, a great portfolio piece and stories that you can tell potential employers during interviews. I hope that anyone reading this article was able to get some good information and inspiration for their next project as well.