| 18 May 2022 |
 lazygecko | do your shaders support additional texture materials? | 22:52:52 |
 zesterer#3131 | No more than vanilla OpenMW does, it would need code changes in the engine | 22:53:15 |
| lazygecko | the difference between terrain and object speculars is still a bit confusing to me. the way i understand it, for terrain speculars you have both the diffuse and specular baked together into the same file (with the diffuse in the rgb and the spec in the alpha) and then that _diffusespec suffix completely supercedes the original diffuse texture? | 22:55:02 |
 Quote#8404 | Again, yes, conceptually. But we're not simulating individual light rays, and the net effect is an accumulation that need not be a precise representation of only one or the other. Quoted by AnyOldName3#5223 from #shaders | 22:55:11 |
 AnyOldName3#5223 | So we use approximations of what happens when the things happen to multiple rays which have been derived from Maxwell's equations (or what was only a few instructions on the machine Bui Tuong Phong had access to when doing his PhD), and no approximations have been discovered that are any good at doing both at once. | 22:55:13 |
| AnyOldName3#5223 | Doing one extra thing and a million after reaching the surface is close enough to the same that they can all share one approximation, but just getting reflected immediately isn't. | 22:56:20 |
| AnyOldName3#5223 | Well, there's the option of summing an approximation of one and an approximation of the other, but then you are using a decently precise representation of them separately as an intermediate step. | 22:57:55 |
 Abdu | Backlighting when. | 23:08:55 |
| Abdu | We have post processing now... | 23:09:10 |
| AnyOldName3#5223 | That was kind of never the problem. The real issue is identifying what to apply it to, and that's not really something Morrowind's data contains. | 23:30:58 |
 CMAugust#6575 | No consensus on adding niextradata, no consensus on a simple suffix, no consensus on extending omgame format… guess we’re just stuck then. | 23:33:34 |
| Abdu | (I mean they pretty much agreed on the later nif support). | 23:34:22 |
| AnyOldName3#5223 | Once we figure out how to deal with later Nifs, we'll be in a better position to think about how we might be able to hack equivalent features into Morrowind Nifs. If we try something hacky based solely on what we want for Morrowind, we might end up adding stuff that causes problems when adding similar stuff for the later games. | 23:38:12 |
 Dexter | Later games are becoming outdated faster though... | 23:39:31 |
| Dexter | How about non-hacky? | 23:40:42 |
| Abdu | Define non-hacky if it isn't later nif support like. | 23:43:22 |
| Dexter | Modern | 23:43:45 |
| Abdu | I thought the way the define stuff is goid | 23:43:49 |
| Abdu | * I thought the way the define stuff is good. | 23:43:56 |
| Abdu | Well modern, as in how is stuff defined there? | 23:44:38 |
| Dexter | They are defined in Bethesda's way of hacky. | 23:45:50 |
| Abdu | * Well modern, as in how is stuff defined in modern? | 23:46:31 |
| Abdu | * Well modern, as in how is stuff defined in modern? | 23:46:41 |
| Dexter | And it would be a waste of time to recreate their weird stuff. We should lift them up to our higher standard. | 23:46:54 |
| AnyOldName3#5223 | The combine-enum-bits-to-enable-features idea is fine, really. | 23:49:16 |
| CMAugust#6575 | How does that work, again? | 23:49:49 |
| CMAugust#6575 | Is it something OpenMW can allow in the near future without much argument? | 23:50:22 |
| Abdu | * Well modern, how stuff is defined in modern? | 23:51:30 |
| lazygecko | isn't bethesda's "pbr" implementation in fallout 4 really weird and ghetto? like they used one of the rgb channels in the diffuse to define the roughness or something | 23:52:24 |
| CMAugust#6575 | I thought that was just being economical. Who wants to bind more textures than you have to? | 23:54:43 |
