Wait, you're telling me there's no actual person on this board who has admin powers? Are you joking? I thought Sixen et all were admins?
It's much easier to check initial posts and see which ones are legit and then approve them (which then approves the account) rather than having to clean up the constant mess left behind by bots that only serves to drive newcomers away.
They wanted feedback. I am giving feedback. In the modding world, "oh my god this is so amazing!!!!" is not actually considered feedback. I'm not trying to be negative, if that's what it comes off as.
Turn on moderator preview for new accounts. I've only said it like 3 times. Unless you can't actually do that. Then, I'm afraid, it's time to upgrade to forum software from the 20th century at long last.
Being an artist takes a lot of dedication, that's for sure. I've been modding for a very, very long time, and I'm still a long ways from trying to become an artist.
I can review these screenshots I suppose. If I did a real playthrough I would record it into a series of videos, because that's how I normally do that. But, yeah, kind of got a lot going on as it is.
1 - While the white tiles appearing as sort of plates or additions to the wall is kind of a neat idea, they lack the grunge texturing of the base textures. If possible, adding some dirt to them would go a long way to helping them fit in the scene. Additionally, consider breaking up repetitious segments with smaller details, perhaps some cables or small vents. Nothing that affects the collision, just something built into the existing framework. For the floor, I was exploring the idea of making custom metal decals to break up repeating paneling, but unfortunately I wasn't able to actually test the idea. I think it still has some merit in concept. Some decals of some nature to help break up the flooring will also help, but keep them fairly minimal as the color balance of the scene is already fairly noisy and too much more can easily hurt readability. You already have some decals for the door which is great.
2 - This scene is more indicative of what I was thinking for scene #1. The objects on the walls have some variation between them. Consider adopting a standard for the decals for the doors, as this door appears to be dramatically different in labelling compared to the one in #1. The pipes in the upper corner are blocking access to the gas tanks, which seems illogical if this is where they are accessed. If the gun turrets are insinuated as being placed as defenses prior to the events, considering encircling them in decals. This was something I did with gas geysers and similar structures in a space platform map for Apex A and I felt it made them appear more like they belonged. That is to say, make the turrets a part of the environment rather than existing solely as a gameplay element standing out of it.
3 - The gas tanks appear repetitous and monotonous. Unless it's insinuated that the facility is fully stocked, consider having some empty or drained cabinets. The left and right walls are comparatively bare, and I feel like something should be between the pipes to justify their entry and leaving the surface plating (which is otherwise engineering complication for the sake of complication and breaks realism). The lower level is pretty solid.
4 - I feel like the terrain texturing should be more detailed in that the cliff areas should have more weathering. The far right cliff I think symbolizes this well, but the sentiments are not fully shared with the left inner cliff. The most problematic part to me seems to be the back area in general, which has this feeling of being a bunch of random doodads taped together. This could be alleviated by working the rocks into more parts of the terrain so they feel more natural when viewed in this context, and consider that the burning charr doodads are in fact burning and are standing next to vegetation that otherwise may not survive. Using the grass more liberally to flesh out the valley will also make its appearance in the corner less unexpected. In general, I feel like the color palettes here kind of clash between the purple-orange and then the blue terrain, but you could, if you have a free texture entry, slip in something to help blend them together via the texture tiles.
5 - I feel like the compositing here is much more patient than #4 and as a result it appears much more natural. Most of my qualms about this are subjective, like how I really, really hate how heightmap mountains look like, but there's not much you can do about that. In the left far corner there is vegetation that I feel like could use some support from the ground textures. I would also consider placing very sparse amounts of creep about the zerg structures to help them fit a bit more, perhaps using some of the doodads in very small scale to give them more footing.
Hopefully that helps you with some alternate perspective!
Your fourth screenshot looks very strange. The lighting for the static mesh objects is very bright and oversaturated compared to the rest of the scene (e.g. the heightmap terrain). I find it hard on the eyes, even. It would also appear that, while you placed many objects and decals for doodads on opposing walls, the walls facing away from the camera very clearly have no detailing whatsoever, and break the scene considerably. You should add details there to complete the scene. Additionally, I would consider turning down the specularity for the water in that second image. A sheet of white is also not very easy on the eyes nor pleasant to look at (maybe this is due to your camera angle, though). A consideration is to also brush work some textured weathering into the shoreside to reflect hydro erosion. I also recommend updating Gna's carrier texture to fit more closely with the existing protoss textures. I have such an updated texture on hand, but I also updated it for my modified version of that mesh which uses player colors. If you are interested perhaps I can modify it for your needs.
I have too much on my plate to give you a playtest, as I would probably spend hours per mission nitpicking details and I have a lot of other stuff to cast at the moment. Perhaps in the late future.
Any of you guys know what are the biggest performance sinks for the existing Melee AI? We've stripped a lot out of it, but the performance is still mindblowingly wretched. It makes any hope of a project I might want to make in sc2 a pipedream at best. Even just two computer players in a 256x256 map can drag the fps down to 7 for my reasonably modern hardware that can otherwise handle hundreds of thousands to millions of polies in this engine without issue. Adding another 3 Ai players doesn't make a noticeable difference, manually painting pathing made no difference, dropping unit counts (which weren't high to begin with) makes no difference, stripping tactical AI makes little difference (some are very bad, like spore colony, but most are insignificant). There's some kind of overhead that really hurts performance.
I've spent five years off and on trying to solve issues with the AI in this game which have all but crippled any desire I had to get involved with it beyond proof of concepts. I would love to put my designs to the test in a large campaign project, but it starts and ends with the AI's performance. I've long given up hope to find answers because, like I said, no one seemed to know anything. But if you know, I'd love to put it to the test.
/edit
I see. Unsurprising lack of response. Just wanted to check. Good day.
Turn off Extra Scouts flag and that will stop their flailing around, I think. They have some other stuff associated with them but I think that's the big part for their scouting. It also eats a lot of CPU for Zerg.
Ravens are in the Terran tactical galaxy file, iirc.
I somehow doubt he intends to change all of the tilesets and units to fit with it. Though that would be something to behold if it was achieved. It would be much easier just to recreate the weapon effects using existing textures already available in the game.
Also, those images were for a fixed angle and functioned as overlays in sc1 with specific sorting priorities. Spawning them as a billboard in sc2 will look really weird.
I will just assume he has some alternative goal in mind for the textures than using them just to replace existing particles.
Starcraft 1 is weird because it uses a bunch of pcx files to translate those images into transparent versions. Those files also determine stuff like how cloaking works. The files used a special palette so they could be used by any of those conversions.
Although why you would want to use such low resolution and low color depth images in sc2 is beyond me.
Huh. I don't seem to recall that in the documentation. I thought I read that they are an alternative to the volume for something. I guess I should read again if we do stuff again.
Textured paneling would bring out the UV stretching a lot more, like it did in Sins. It's possible with Unreal's material editor I can create dynamic UV's, however, and place tiles independently of mapped co-ordinates. But that is way too advanced for me at the moment.
Ideally, when updating the meshes, I would model all panels instead.
I still don't think turrets will be manageable, so I won't really be pursuing this idea.
The lighting is rather dark. I should have modified the directional light a bit, or added a skylight, before finishing up.
I think you could get away with dependency on tiled textures if you A.) modelled a lot of detail and split the meshes for different materials, and B.) used a lot of effects to help cover stuff (smoke, steam, fire, movement of other ships, lighting, etc.)
0
Wait, you're telling me there's no actual person on this board who has admin powers? Are you joking? I thought Sixen et all were admins?
It's much easier to check initial posts and see which ones are legit and then approve them (which then approves the account) rather than having to clean up the constant mess left behind by bots that only serves to drive newcomers away.
0
They wanted feedback. I am giving feedback. In the modding world, "oh my god this is so amazing!!!!" is not actually considered feedback. I'm not trying to be negative, if that's what it comes off as.
0
@xcorbo: Go
Turn on moderator preview for new accounts. I've only said it like 3 times. Unless you can't actually do that. Then, I'm afraid, it's time to upgrade to forum software from the 20th century at long last.
0
Being an artist takes a lot of dedication, that's for sure. I've been modding for a very, very long time, and I'm still a long ways from trying to become an artist.
I can review these screenshots I suppose. If I did a real playthrough I would record it into a series of videos, because that's how I normally do that. But, yeah, kind of got a lot going on as it is.
Hopefully that helps you with some alternate perspective!
0
Your fourth screenshot looks very strange. The lighting for the static mesh objects is very bright and oversaturated compared to the rest of the scene (e.g. the heightmap terrain). I find it hard on the eyes, even. It would also appear that, while you placed many objects and decals for doodads on opposing walls, the walls facing away from the camera very clearly have no detailing whatsoever, and break the scene considerably. You should add details there to complete the scene. Additionally, I would consider turning down the specularity for the water in that second image. A sheet of white is also not very easy on the eyes nor pleasant to look at (maybe this is due to your camera angle, though). A consideration is to also brush work some textured weathering into the shoreside to reflect hydro erosion. I also recommend updating Gna's carrier texture to fit more closely with the existing protoss textures. I have such an updated texture on hand, but I also updated it for my modified version of that mesh which uses player colors. If you are interested perhaps I can modify it for your needs.
I have too much on my plate to give you a playtest, as I would probably spend hours per mission nitpicking details and I have a lot of other stuff to cast at the moment. Perhaps in the late future.
0
Any of you guys know what are the biggest performance sinks for the existing Melee AI? We've stripped a lot out of it, but the performance is still mindblowingly wretched. It makes any hope of a project I might want to make in sc2 a pipedream at best. Even just two computer players in a 256x256 map can drag the fps down to 7 for my reasonably modern hardware that can otherwise handle hundreds of thousands to millions of polies in this engine without issue. Adding another 3 Ai players doesn't make a noticeable difference, manually painting pathing made no difference, dropping unit counts (which weren't high to begin with) makes no difference, stripping tactical AI makes little difference (some are very bad, like spore colony, but most are insignificant). There's some kind of overhead that really hurts performance.
I've spent five years off and on trying to solve issues with the AI in this game which have all but crippled any desire I had to get involved with it beyond proof of concepts. I would love to put my designs to the test in a large campaign project, but it starts and ends with the AI's performance. I've long given up hope to find answers because, like I said, no one seemed to know anything. But if you know, I'd love to put it to the test.
/edit
I see. Unsurprising lack of response. Just wanted to check. Good day.
0
Turn off Extra Scouts flag and that will stop their flailing around, I think. They have some other stuff associated with them but I think that's the big part for their scouting. It also eats a lot of CPU for Zerg.
Ravens are in the Terran tactical galaxy file, iirc.
0
@TaylorMouse: Go
Does the crab pinch when you block with it?
0
I somehow doubt he intends to change all of the tilesets and units to fit with it. Though that would be something to behold if it was achieved. It would be much easier just to recreate the weapon effects using existing textures already available in the game.
Also, those images were for a fixed angle and functioned as overlays in sc1 with specific sorting priorities. Spawning them as a billboard in sc2 will look really weird.
I will just assume he has some alternative goal in mind for the textures than using them just to replace existing particles.
0
Starcraft 1 is weird because it uses a bunch of pcx files to translate those images into transparent versions. Those files also determine stuff like how cloaking works. The files used a special palette so they could be used by any of those conversions.
Although why you would want to use such low resolution and low color depth images in sc2 is beyond me.
0
Huh. I don't seem to recall that in the documentation. I thought I read that they are an alternative to the volume for something. I guess I should read again if we do stuff again.
0
Care to elaborate?
0
Textured paneling would bring out the UV stretching a lot more, like it did in Sins. It's possible with Unreal's material editor I can create dynamic UV's, however, and place tiles independently of mapped co-ordinates. But that is way too advanced for me at the moment.
Ideally, when updating the meshes, I would model all panels instead.
0
I believe ref_target and the target volume are for that function.
0
A few final pieces from the above experiments.
I still don't think turrets will be manageable, so I won't really be pursuing this idea.
The lighting is rather dark. I should have modified the directional light a bit, or added a skylight, before finishing up.
I think you could get away with dependency on tiled textures if you A.) modelled a lot of detail and split the meshes for different materials, and B.) used a lot of effects to help cover stuff (smoke, steam, fire, movement of other ships, lighting, etc.)