Creating A 3d Environment

As a concept or 3D environment artist, do you have to learn a lot about architecture and interior design, or do you just use images as references (because most games have buildings, furniture, etc.)?

I know some pretty talented concept artists, (who have worked on some high profile projects).They do not, to the best of my knowledge study architecture and interior design in the same way as someone working in those fields.But what they do is collect, observe and draw from vast amounts of reference. They will often approach a project by collecting hundreds of reference images. They tend to walk about with cameras and take snaps of buildings, rooms and environments which can be repurposed, often years later.It’s a good idea to draw buildings from life too. There’s something about drawing which forces your brain to ingest the forms.

What is the process to create 3D environments for the video game industry?

I hope you have the basic knowledge of the production pipeline.Here is the process:Concept and Layout phase: Initially the concept art is developed by the responsible people to get the look of the environment. Things like color theory and composition comes into place. Also the artists makes the list of assets to be used in the environment.Blocking: To get the idea of the environment scale in game, artists create a basic version of the level with some basics shapes. Softwares: Autodesk Maya or 3ds Max.Asset Creation: Based on the lists, the artists start to create the asset one by one in Maya/Max and Zbrush.Texturing: Paralelly textures are also created by seperate team of artists in softwares like Photoshop, Quixel Suite, Substance Designer and Painter.Baking: During the previous process, Artists create very low polygon models to give a better performance in game so eventually they fake the details by a process called Baking. It comes under the texturing process. Things like Normal maps etc. comes into consideration.Importing the assets to the Game Engine: So now the assets and their textures are created. Now they have to import everything to a game engine to use it. The type of game engine depends on the studio. Some of them are available online and some of them are proprietary. This is a time taking process because there's a lot of technical work here like connecting the assets to the textures, creating shaders and so on.Set Dressing: The assets has to be laid out based on the design of the level. Most of the time, it's not time consuming.Lighting: Most of the time or say everytime, the lighting has to be built in the engine only. Artists take reference from Concept Art.Finalisation and Testing Phase: By this time most of the environment work is done and is passed to testers who test it in which gameplay, simulation (if any) and animation comes into consideration. It's Art Director who approves it.I hope you get the idea behind the process. I would suggest you to watch some lecture/seminar videos on YouTube published by The Gnomon School to get a even clearer view. Will update with links and pictures.Cheers!

What is the best way to create environments (2D and 3D) for reinforcement learning?

There are many great tools which make this process fast and easy.In the past years we have been indexing a great diversity of environments and tools (120+ at the time of writing); you can find them easily by visiting RLenv.directory and selecting as tags, for example, “Framework” and “Engine”.(2D/3D) Unity ML agents frameworkPersonally I am a big fan of the Unity ML agents framework, it allows you to design environments in the Unity game engine (which is very straightforward to use) and it takes care of providing a Python interface.Do submit a pull request to the RLenv.directory Github repository once you create a new environment so other people can find it easily!Good luck!

What is the easiest way to create a 3D environment that you can walk around in, in a program or game?

Download unity, open it and create a plane, think of it as a floor, stretch it as much as you like. Add a floating 3d object, add controller to it and push play. Search the web during this process. Enhance the environment according to your imagination.Depending on your interest and level of your profession you may need a separate 3d modeling software(blender, it is free), a graphical design software(adobe photoshop or illustrator), sound editing software and other side tools. If you are a proggrammer you can find an artist to take care of the visual part of the development or use asset libraries which has ready to use assets, otherwise you may end up spending lots of time for even the tiniest enhancement. If you have a busy life, it gets really overwhelming without a project plan.

I have a virtual 3D environment in Maya. What do I need to do to make a Google Cardboard app with that virtual environment?

Easiest thing would be to import it into http://Unity3d.com and download the cardboard unity plugin (which will be natively integrated in a future Unity version.) That will allow you to easily output an Android APK suitable for Cardboard or GearVR, an iPhone app for Cardboard, or a PC executable suitable for an Oculus Rift.

Can any 3d video game environment be ported into vr with a minimum of additional effort?

No. At a minimum then entire UI has to be redone. Most games use screen space UI - that is a 2D image drawn on top of the game scene at the end. That is wonderful for normal displays. For VR - there are two screens. Where do you draw it? Left? Right? Both? - The answer is neither. Best practice right now is to convert to a world space UI.Input also needs to change. Using a keyboard and mouse is much harder when you can't see the keys.How the camera works also needs to change. General VR has the camera controlled by your head movement. Many video games have algorithms that control or bound the camera. Moving the camera for the player when you are in VR is a horrible experience. Think vomit land. FPSs could probably get away with no/minimal changes. Eye in the sky cameras (think civilization) would also be okay as long as you don't teleport the camera on events. There are more considerations but those are the ones that stop you from just dropping in a VR library and walking away.