An example of Bethesda's beautiful normal map making from Doom Eternal. Say something nice about it.
>>1015695Oops, too small.
>>1015695Let's see yours. If you wanna get hired you're gonna have to learn to make those
>>1015695>>1015697What part am I looking at? How much of the screen does it take up?
>>1015695>>1015697How about you post a screenshot of the game model for context too?A 1K resolution for a health pack, which should take a high priority in the first place, seems fine. The details can be judged after looking at the geometry.As of right now, I think OP ist a faggot
>>1015695hm, seems really clever. its just one corner of the health pack with a bunch of small parts all tiling? to make the other faces?
>>1015700>should take a high priorityshouldn't take a high priority*My bad. I am a faggot too. A phonefag, to be more precise.
>>1015695Considering the health packs are basically unlit in game, why should I care?
>>1015701>>1015699Not OP, but here's the full texture.I'm not sure what OP's thread is about. Maybe he wants us to notice some of the the edges are aliased.
>>1015734Oh, well thats disappointing
>>1015727>>1015734>>1015727Hyperfocusing on (sometimes only perceived) non ideal practices in production is the traditional fastest way to feel smart in 3dcg. It's the whole logic behind constantly whining about topology. It's easy to pick up for newbies and instantly you can point to at an unretopod sculpt used for renders and feel good about yourself - an enlightened professional. OP just wanted to be a tiny bit more original
>>1015754I mean, OP hasn't argued why they think there's something wrong with that normal map
>>1015734My god, the projection on the buttons is slightly angled. This means the cage they baked from is incorrect. Its all over or something. IDK i haven't baked a normal map in a decade
>>1015695If it actually bothers you, just take a slight blur brush to it. In a 1024x1024 texture, this kind of issue on minor details will be virtually undetectable without close scrutiny. Nobody would notice this in game.>>1015734Yeah it looks perfectly fine, especially with the full context.
>>1015695Better than containment breaches.
>>1015695i looked at some of the ripped assets... they are badi was wondering when i played whats up with all the nromal pixelization, but its just all the reainbows, they compress like shit>>1015734>I'm not sure what OP's thread is about.me neitherbut i can right away say that if you gonna use weighted normals there is no excuse for how shit the mesh normals look for such a shit bakethis is how propely baked normal map looks >>1016015all sides average to single color, its important for in game compresion, those large color shift create those ugly ass giant squares on surfaces you see in games often, if you ever wondered what causes it, its this
>>1016232>all sides average to single color>those large color shift huh? In English, please?
Is this one of those threads where we pretend a ripped asset is the actual asset?
>>1016288most people who come here are /v/ tourists, unfortunately
>>1016251this shitrainbow bakes produce 2 problems, one is improper shading in most software, and two, awful artifactsironically last game i saw some really god awful normal artifacts is doom dark ages, some of environment textures look like liquid shit
>>1015695I think it was done via software.
>>1016455I like the middle one better, it has a cool look
>>1016455The normal artifacts are because the straps are floating geometry. Their baking software is not anti aliasing the floating geometry.
>>1016682I think what that guy is saying, is that some games and software compress normal maps in vram after they get loaded. So if you have these huge gradients covering a lot of space, and if the map gets compressed, you get really ugly "stepping" on these normal mapped surfaces. The solution to that is to have as little of these large gradients as you can by manipulating the normals in vertex data semi-manually to minimize large curves on the normal map. Though maybe the OP image isn't a very good example because because the gradient on the map there is cloth folding a bit, and depending on its polycount you may not be able to adjust the normals to follow the "contours" of the folds. The solution there, I guess, is to not compress that particular map to begin with.
>>1016634You're fucking insane if you think that looks good. it looks absolutely awful in motion when you do that.
