Check it here: https://ldjam.com/events/ludum-dare/39/energy-scraps

I'd like to show our 3d-content creation process and highlight some tips along the way.

Our Entry: https://ldjam.com/events/ludum-dare/39/energy-scraps

So, this is what we've achieved during Jam:

Our plans were to make a relatively large location for player to explore, including both exterior and interior parts.

The first thing we've done was modular pieces of interior. It was really draft at start, but we managed to assemble a simple test scene, animate doors and pass all of this to programmer as fast as possible: I've created colliders and prefabs from this sketch models as well.

Meanwhile, I've updated models and animation:

When models were done, the only thing that is left was to update test scene. And all I had to do is to replace meshes and pick new colliders' mesh in prefabs in the engine. Door animation were replaced in the same fashion. By doing this we've saved a lot of time and the scene was refreshed automatically and remained playable:

In the final version interior part has evolved. The modularity of the blocks really helped us to reduce time needed to create location. All this work came down to copy-paste pieces and snap them to the grid in the engine:

The situation was different with outdoor part of location: it wasn't necessary to align objects to a grid anymore, but I had to keep number of unique models as little as possible and somehow fill the canyon. And we wanted that player had ability to jump around the canyon, so we needed a platform-model.

We got off with this most minimal set:

It's only two wall pieces: the platform and the column. To make the whole location out of them, they had to be severely scaled, non-uniformly as well. Appearance of the meshes was acceptable, but the difference in texture scale and offset was obvious. As a result, the view of the canyon was a mess.

We solved this problem with shader based on the world coordinates. Thus, all the pieces have the same color variation in vertical direction, and the overall canyon looks more consistent. An additional advantage: we didn't have to bother with UV-mapping.

Props and enemies that were done as well. It was just basic modelling.

To avoid making explosions or some kind of death animation I quickly made another shader, that we animate through the script. It also uses only the world coordinates, so it is suitable for all enemies.

By the end of the Jam we've managed to create a location of this scale:

It took us about 10-15 hours to model and texture everything. It should be noted that even in this short-term project about half of the models was re-modelled, some of them completely from scratch. The rest part of Jam I was building shaders, improving the visual appearance and creating location (here we worked together). Making final build and publishing also ate quite a large chunk of the last day.

It was a crazy little Jam, but we've managed to get quite a lot of experience out of it! Thanks everyone for your feedback!

This time I've done it alone. Here's what I've made: https://ldjam.com/events/ludum-dare/40/mining-empire

The more work you have, the worse it is!

You’re the only worker on a mining station. Station AI will give you more and more orders. How long will you tolerate her tyranny?