373 Metro Final Scene Project

Metro 2033 Game Scene Final Project Post



For this project my team and I have been working to recreate the level "D6" from Metro 2033. I liked this level because it had a lot of effects, interactables, and a mood an aesthetic I really like. It's dark and moody, dead, creepy, and intimidating which I think are really cool emotions to evoke from a player. Below I will talk about my work on the project, some of the problems we ran into, as well as the things I have learned from this project as a whole. 





My main focus for this project, besides being a lead, was to be the level designer. For me, that means assembling all of the models, making sure they have appropriate colliders, and piecing them together to make the level. What I also added to this was to work as the VFX artist for the level as well simply because I have experience doing it and I'd know how the effects should be implemented in the level. While my modelers were making the final versions of all the props, interactables, and terrain, I made blockouts for the various sections of the map as well as the VFX that would be used. Above you can see not only how the level came together but also the three main VFX I created. The first is a miasma effect that spawns around the player when they first spawn into the level. This is an effect that's present in the main game and adds to the deathly feel of the room. The second is triggered when the player turns on the fans and sucks up all that miasma into the central fan. The third is for the generator the player finds at the end and is meant to make the generator feel unstable and dangerous.



As far as building out the level there were a couple of things that I needed to change with our build because of how our pieces fit together and the limitations we had with what we were able to do in time. The first concerns the dolly car the player rides into the silo. In the game, the player is able to ride it from the top, down to the first story a quarter turn, then down to the first section another half turn. Because we had so many issues trying to get the sizing on this just right, we ended up making the track system only move a quarter turn down from the first story into the silo and back. This just made more sense with what we were able to make but also with the pathing we wanted the player to take. The second big change I added was this staircase on the right side of the tracks. I added it because the space felt particularly empty on this side and it was somewhere the player could theoretically get to, so I wanted to add a little more interest to the space and another pathway for the player to fully explore the space.


I have to say that the biggest challenge of this build by far has to be the helix track in the center of it. It was a focal point for the whole level and it was such a hassle trying to get it right. Trying to get he size of it lined up with the cart but also lined up in the level so that the cart goes up and down to the correct places. This ended up needing a lot of iterations on my part, even eventually changing the carts path to not be 3 levels but actually just two. We also abandoned the idea of getting the tracks to line up perfectly with the cart animation, as trying to get both the angle of the cart as well as the movement up and down to line up perfectly was more of a core than it was worth in this instance.

Overall, I'm very happy with how this whole project went. I got more experience leading a team and I think I did pretty well with this one. I was able to get everyone on the same page of what we were looking at, develop a priority for what tasks I wanted done first and I was able to keep everyone busy without making anybody have to crunch (well, I might have a little putting the level together). I got more more acquainted with the process of bring a model from Maya into Unity and learning everything that has to happen to that object before it becomes usable in a level, as well as some annoying limitations Unity has with applying colliders. I also learned a number of new tricks that can be done programming-wise from my programmer team member. I hope to be able to use both his tricks as well as my better knowledge of the modelling pipeline to even better lead a team in the future.

Sprint 5 Review

Sprint Statistics

Tasks Assigned: 38
Tasks Moved to To Verify: 29 
Tasks Moved to Complete: 29



For this sprint, I finally started getting into implementing all of the models we've been getting! First step was assembling and applying colliders to some of the pieces. The stairs were interesting because I had to go through a couple iterations so that they connected smoothly. The first issue is that Unity doesn't have triangle colliders, and you can't rotate the colliders. To get around this, I had to make empties that had colliders on it that I could rotate to make the pitch of the stairs. At first I wanted to get the top and bottom stairs to still be square, but then the player kept tripping on it, so I ended up lining it up with the bottoms of the stairs so that they had a smooth transition point. The silo ring I ended up making it out of box colliders because the player's main interaction is going to be walking on top of it. The dome needed to be a mesh collider because the player interacts closely with it. I also had to add the fans as well as the top fan to the prefab to make it happen.


Another couple of tasks I completed was created some vapor effects. The first I finished was a miasma effect. This is something that will be surrounding the players during the entire playthrough. It uses world space for the particles and they move very slightly to give them an airy feel but still feel stale. The second is a venting system effect the player will be able to activate where the fans will kick on and suck up all the miasma. For that I was able to use the same particles just adjusted in a way to move upwards and be bigger.




Lastly, is the progress I've made putting it models to the level. It's looking WAY better with all the intended models and textures and it makes me excited. The layout is going to have to change from the blockout so that it works with the models we have, but I think it'll look cohesive and proper. I'm really excited to get the last models in there and get all the grey box blockout pieces out of there.

Sprint 4 Review

Sprint Statistics

Tasks Assigned: 30
Tasks Moved to To Verify: 27
Tasks Moved to Complete: 27


Unfortunately, this sprint has been light for me. Had some issues at home and got sick along the way and so I got less done than I would've hoped. Nonetheless, we carry on! I got three of the boxes that the player can open working and animating on triggering with the player entering them. This will change once the interaction system is worked out, and then the player will be able to walk up and press "E" to open these containers. The two single hinge boxes were pretty simple and I'll probably speed them up more later to be more consistent with the gameplay we're mimicking. The third gif shows the toolbox and that one is a bit more interesting because it has a latch and a lid. For animating that I made sure to have the latch immediately move down and for the lid opening to wait just a little bit for the latch to move out of the way.


The last thing I did this sprint was to make a sparking VFX. Using Unity's Visual Effects Graph, I had all the control I needed to make the sparks, their trails, size them and give the velocities that made sense, as well as making all the needed variables public to access so that we can adjust it more once it's in the space. I also added a point light, and using my teammates' light flickering script, made it to flicker along with the sputtering of the sparks. Really happy with how this came out, but I might make the sparking a little bit bigger, more dramatic once we get all the props and appropriate scene lighting in.




Sprint 3 Review

Sprint Statistics

Tasks Assigned: 31
Tasks Moved to To Verify: 26
Tasks Moved to Complete: 26




The first and major thing I did this sprint was to block out the next section of the level. While I thought that this would be fairly simple to do because the space is not as populated, as it turns out making round things in Unity is hard. Move over, the perspective shift of going down into the hole made it very deceptive as to the scale this section needed to be in order to fit the space correctly. After a lot of trial and error making funky shapes in unity, I decided just to take it into Maya and build it out there. Making the various rounds was much easier and I had much more control over the vertices and scale. After a lot of fitting, importing, retrying, reimporting, etc, I finally found out that the helix that I thought took a quarter turn to move down to that level, actually took a half turn, doubling the height scaling I thought I was shooting for, so I continued to adjust and finally got something that really feels like the proper metro space.



After that mess I was able to start working on getting our new models and textures into scene and getting them ready for player interaction. The first step was assembling the train (which my modeler Travis did a wonderful job with), which mainly consisted of applying correct colliders. The complication with this prop is that the player will actually be able to be inside and outside the train, so I needed to make sure they didn't phase through the side but also be able to walk down the center isle. I ended up using something like 12 colliders in order to get the geometry to handle correctly.

After that I was able to complete a door opening animation. Right now it just triggers on enter to open, but I will need to be able to lock the doors so the player doesn't walk off the wrong side of the train later on.




Sprint 2 Review

The Team:

Travis Wilson: Modeler
Rein Tice: Modeler
Tucker Holmes: Programmer
Christian Smith: Lead / VFX Artist / Level Designer

Sprint Statistics

Tasks Assigned: 39
Tasks Moved to To Verify: 33
Tasks Moved to Complete: 33


My major focus this first sprint was to set up organization for the project as well as collect references for all the different parts we were going to cover. Given our modelling team, we're shooting for some ambitious goals, but I wanted to make sure they were still reasonable and could be scaled back if necessary. I started off with making sure our GitHub Repo was up and running. I originally assigned Tucker the job, but he was getting some strange errors on his end that I didn't get when I tried setting it up on my end so we ended up using mine. Then I made a Miro board so that we had a collaborative space to show off vision as well as scope for the project. Once that was done I started going through the level and making lists of all the different things the level had in it including props, modular pieces, animations, VFX, SFX, scripted events, and lighting. I took an obnoxious amount of screenshots in game, far too many to be useful, and ended up instead recording a walkthrough video and uploading it to YouTube for the team to make quick references.







I also finished up blocking out the first area of our level. This section is by far the most complex both in terms of number of objects as well as the overall layout. The next to sections are both fairly easy to block out as they are basically a big, round hallway with some doors in it, and the player only ever enters one quarter of them. Getting all the dimensions to feel right between the round hold and ceiling as well as all the building spots with each other. I'm pretty happy with how it turned out even though it's a fairly rough block out.



Comments