next-gen prop unwrap question

Doomathon

Hi all,

So I am making a shipping container that would be for a first-person/third-person shooter game, one that the player could enter to use for cover and I was wondering how to approach the unwrapping.

So I have the outside and all the parts that are associated with it, then I have the interior as a separate mesh. Should I combine the whole asset onto one texture sheet? Or would it be viable to have a separate sheet for the exterior and the interior, to get the most texel density? I'm just unsure if that is a common industry workflow approach to have multiple texture sheets for a single asset.

Thanks for your advice!!!

sprunghunt

More sheets mean more drawcalls. More drawcalls are worse than low texel density as drawcalls will hit the framerate very hard. Texel density isn't a performance problem in modern platforms.

It's probably better to just re-use the same texture on the inside and outside if you can.

Doomathon

Interesting, thanks for the information and I'll give that a shot! Just curious, what if I split the sizes? Like made the outside 2048 and the inside 1024? Does that size make a huge difference or is it the actual drawcall process that is taxing overall?

sprunghunt

Doomathon wrote: »

Interesting, thanks for the information and I'll give that a shot! Just curious, what if I split the sizes? Like made the outside 2048 and the inside 1024? Does that size make a huge difference or is it the actual drawcall process that is taxing overall?

It's the extra drawcall/texture. Although two drawcalls rather than one isn't too bad as you'll usually have hundreds of drawcalls per scene. The problem comes when you repeat the same object over and over again (in this case if you were making a whole container yard). This can cause a two drawcall container to turn 200 drawcalls into 400 drawcalls.

It's just good practice to minimize draw-calls.

Doomathon

Okay that makes sense, thanks for the help!