I’m thinking to start a subscription for UE4 from january but I’d like to ask some questions first.
Some things about myself:
I finished my Game Development studies last year and while I do know my programming for the most part what I love doing is 3d. I have an idea for a game since forever and that’s something I want to start working on. I understand how hard it is to make a game work well even as simple as mario or tetris (did both of em during my studies) so I don’t really believe I’ll ever manage to finish what I have in mind. The main reason I want to start doing it is because I learn from the process and because even an intro level with 15 or so minutes of gameplay will look good on my portfolio. Chances are that I will continue working on that as a personal project but it’s not something I plan on releasing or selling at any point due to the massive work needed and the fact that I am alone in this (for now at least).
I worked a bit on Unity in the past for a year or so building and refining the cameras and/or character controllers and made a basic landscape for it but I decided to switch to UE4 for several reasons. Without getting into any details about the game I’d just like to know some basic stuff about UE4. What I want is (for the graphics at least) to achieve the graphical fidelity of games like Thief or Tomb Raider (mind you, just the looks. I hated both games with passion) and while I’m pretty sure I can do it with enough time and effort, I wanted to ask some things so I plan it better.
What I’d like to know is if there are any caps on details for models/textures (of course there are but what are those?). For example, draw calls, areound what number should I keep my cap at?
Around what number should I limit my polycount for a scene?
Texture size? Clamped textures or not? Things like that.
I never used Unreal or UDK before and I don’t know anything about stuff like that as every engine uses it’s own way. For example, making models for the Dark Mod which is based on Doom 3 engine I had to limit my mesh to around 1k polys and rely heavily on 2k textures/normal maps, else the frames would take a huge drop when playing a big scene. In other cases the problem was on texture size while even 10k poly models were no problem. I know it’s a trial and error procedure to find stuff like that but some general guidelines would help me here.
The way I want stuff to work is load what the camera sees at full resolution and for the rest of the level load what’s needed to make the game feel alive. For example if the character leaves a room where there’s another person inside walking around, while I don’t want the full room with every detail loaded, I still want it to feel alive so when the character re-enters the room the other person will still be walking around instead of being spawned in the same place that he was when the character left. I also plan on using tesselation heavily for some high details and physX for destruction physics etc. Both of them (tesselation and physX) greatly increase the polycount of the affected model so I’d like to know if it’s doable and if it’s worth it or not.
I’m not sure if my questions make any sense at all as I can’t say my English is good but if you do not understand something please ask and I’ll do my best to make you people understand.