yeah thats another case when rendering, some people maybe years ago used to use like 20 k sample for a simple scene still render like a kitchen for example, but not with the new tech improvements maybe we wont need to use that again, probably not go up more about to 3-5 k if necessary
yeah i would usually render a still image in 2-4 k if its for my portfolio showcase…
yeah but actually 100 k is pretty low poly these days
this is even twice than you say and still looks good the surface for animation or rendering mainly…
Not at all, when you’re talking about forms in sculpting. This is something you would evaluate by squinting, so details don’t come into it. Once these forms are pretty much nailed then you can more confidently move on to breakup and tertiary detailing, if realism is your goal.
That guy is able to add a nice amount of details to his I-rex at only 680K polygons using a stencil. But when I do the exact same thing at several millions poly my details are super blurry … How come ?
Am I missing something here? Folks are talking about millions of polys but OP is asking about a game not a Blender render. A full screen render in a game is already pushing the video card if it’s over 1-2mil triangles.
A model should not exceed 100k, even on today’s gpu because in a game there are lots of other things that are added dynamically to the scene. Overloading the scene causes the FPS to drop and cause lag. Depending on the engine, the model could be bigger, but intensive LOD will be applied hence model detail is lost anyway.
For a game model, you get the high detail from texture and normal map. You add more physical detail (more polys) only where the texture can’t do it. If you’re adding polys for scales and rough skin you’re most likely wasting your time.
Yes, yes, but the original question was about sculpting (i.e. creating the detail to transfer from). Although it’s also true that we kind of, well, derailed the topic. Somewhat. A little bit. Well, almost.
Yes, the whole beginning of the thread before the derail. I explained in great detail the workflow for unique 0-1 unwraps for large assets in a game engine.
As Stan says, but there’s no reason you can’t do this in Blender using the texture painting tools, and capture that to a normal map. The thing to remember with normal baking is that in no way alters the silhouette, it only captures shading information. A lot of people waste time on high poly sculpts when 90% of the work can be painted directly to the final in-game and stored directly in the normal map.