1
\$\begingroup\$

Usually for the game objects a high poly model is prepared, to then transfer the details onto a low poly model, so that it can be rendered in realtime.

Is this workflow always used, with any object?

I imagine this is used for the polygonal models of the characters in general, but what about the rest?

I'm especially interested in the environment, like a snowy mountain for example. I do not think this high-poly to low poly detail baking workflow is appropriate for this case.

\$\endgroup\$
1
  • 1
    \$\begingroup\$ It seems like this question is encouraging discussion, which is considered off-topic here. Are you trying to apply this in one of your games? Then its up to you if you want to follow that technique or not. It's not a must, and not a "best" solution. If you change your question to something like "How do I do X?" or something more specific to what you are looking for, it would be more on-topic here. Until then, I flagged this question as off-topic. \$\endgroup\$
    – TomTsagk
    Feb 21 '19 at 11:27
1
\$\begingroup\$

Yes, there are usually multiple models of different detail for the same Objects. [in open world games for sure].

One Example is PUBG where - if your PC sucks - houses are at first only rendered as vaguely housy looking boxes, then replaced with the real full-detail object.

And that is where your assumption is wrong: usually there are no high-detail models just for development, these are all shipped to be rendered in live. The players PC will pick which "level" he can do and render that.

For the environment, that is up to the engine, some just don't render anything after a certain range (Planetside 2) and make a fog after that, others do in fact lower the poly count on distant objects.

\$\endgroup\$
0
\$\begingroup\$

While user125108 is correct that games will often have multiple levels of detail for environment assets and terrain, I think they may have overlooked a key part of your question:

Usually for the game objects a high poly model is prepared, to then transfer the details on a low poly model, so that it can be rendered in real time.

It sounds like you're talking about normal map baking, where an artist sculpts highly detailed geometry, then bakes the surface detail to a normal map that can give the appearance of similarly fine shading details even on a lower-poly model.

That particular workflow can be used across a wide range of game assets, including characters, props, and even whole chunks of buildings.

For extremely large models though - like the terrain mesh - this type of workflow is not typically used. Instead, we texture these assets with tiling albedo texture & normal maps, interpolating between multiple texture sets where we need a smooth change in the surface material. This lets us add detail to a practically unlimited amount of geographic area, with a modestly-sized amount of shared texture data.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.