I've had these sitting on my HDD for quite a while. They were an offshoot from some early textures test for Doom 3 (way back in the alpha days in fact.. phew..!!) and as they only needed a little bit more work to Quake3-ify them I thought I'd get to it and release them. So here they are. (Ivy texture from a photograph kindly provided by the sock meister. Big thx sock. everything else by lil' ol' me).
A little rule of thumb to go by, always make obvious repeating textures 2x width or even 4x total than what they will appear ingame. It removes the tiling effect a bit.
That looks really nice Kat!
It does in fact inspire me somewhat for a Q3 map. As I said in the screenshot thread, I feel like I want to map again but I lack the ideas. I might try to do a small medieval map with these textures. Only thing I'm afraid of is that these textures are hard to combine with other Q3 textures, but I'll have to figure something out for that then I guess.
Eraser wrote:... Only thing I'm afraid of is that these textures are hard to combine with other Q3 textures, but I'll have to figure something out for that then I guess.
Hold on for a sec I'll post some WIP shots of the Doom 3 set I've been building up. They can, with a little bit of tweeking, be turned into Quake 3 textures so it should broading the walls a bit.. brb...
Ok, some very quick conversions just to give you an idea of what I've got... these are actual size btw. I'd need to spend a bit of time on these to get them right as the conversion from D3 back to Q3 isn't always that straightforward, and as per usual there are several versions of certain ones - the piller has about 6 ot 7 different versions for 'outdoors'.
this'll need a bit of work, but gives you an idea, might even plonk some ivy around the top..
Hipshot wrote:On the other hand, the first I see now looks very good... might get reptable (spelling?) though...
You will get repeat which is why it's 1024 long to try and minimise that. The *only* way to stop repeat pattern is to have the ivy sitting at the same level along the whole length of the wall which isn't what I was after.
The Ivy itself is the only photosourcing in these textures btw, so short of taking a photograph of a wall and making that tile they won't look 'photoreal'.
Eraser wrote:Nice stuff Kat :icon14:
The theme is refreshing and doesn't seem to be done before that much.
And how different are Doom3 textures? The "base layer" is pretty much an ordinary texture (JPG or TGA), right? Or am I wrong?
Yeah, that's a pretty good approximation of what the D3 ones are. Basically they tend to just be flat colours so all the shadowing, the cracks, nicks and dinks you can see in Quake 3 textures isn't there; all that is handled by the bumpmaps. It creates some interesting effects when you see them in game.
The D3 texture set is actually part of a little pet project I'm working on for Doom 3 (not sure if you've seen this so I'll link to it here) so although these can be converted to Quake 3 people might prefer the D3 (engined game) version - given that there's very little in the way of custom textures yet and very little non-tech stuff around.
I know that you have your own webpage, but since when did you get your own forum?
You may want to paint a few extra shadows on some of those textures, I know that you're just layering on the normal maps on top of the diffuse, but there are some shadows that you still need to manually paint on. Like on the morter between those stones... you should paint a darker shadow underneath each rock, at the moment, they look a little too well lit. In Doom3 it should look fine since the normalmapping should generate those kinds of shadows for you... but for Q3, I think you really need to paint them in.
BTW, how did you generate the normalmaps for the D3 versions... did you use renderbumpflat or did you use the nVidia Photoshop normalmap plugin or some other trickery?
The best way to create normal maps is to make em from models, next to that, make em yourself in Pshop, normal maps that is just generated from nvfilter or ATis version, sucks...
Eraser wrote:And how different are Doom3 textures? The "base layer" is pretty much an ordinary texture (JPG or TGA), right? Or am I wrong?
Doom3 textures are made up of 3 texture-maps, diffuse, normal and specular.
Diffuse maps is just the colour of each pixel without any kind of lighting alterations. IOW, you don't paint on shadows or lighting of any kind for the most part.
Normal maps contain the actual bump surface normal information depicted as a range of colours. They look kind of weird and you can't just paint them... you usually need to generate one from a model or from a grayscale heightmap.
Specular maps are grayscale images that specifies how shiny or matte an object is.
Add them all together with a MTR (material - the new shader file) and you have a typical D3 texture.
obsidian wrote:I know that you have your own webpage, but since when did you get your own forum?
That's been there for yonks, but I think people get distracted by all the stuff that on the main site so they don't see or forget to visit it..!! (that might be a good thing though, hehe)
You may want to paint a few extra shadows on some of those textures...
Yeah I did think about that and tried it but the results weren't as good as I wanted so ditched them in favour of the 'flatter' RtCW style of wall. Also sock told me something, a long time ago now, about shadows, it's best too keep them to a minimum with certain things.
BTW, how did you generate the normalmaps for the D3 versions... did you use renderbumpflat or did you use the nVidia Photoshop normalmap plugin or some other trickery?
The real big dissadvantage of using an image over 3D is that photo apps don't have 'oversampling' like 3D renderers do which tightens up the texture (you don't get as much pixel based 'jaggies' with 3D renders); the closest thing to that would be to do the texture at least 2x the final size and then shrink it down before passing it through the normalmap plugins.
Last edited by Kat on Thu Aug 04, 2005 7:29 pm, edited 2 times in total.
obsidian is close, but not bang on, so i'll run over it all again.
A diffuse map is a texture that contains standard colour data only. The diffuse map contains only surface detail detail such as colour differences and texture influences. It doesn't contain any sort of shading for the most part or shadowing, however many people tend to do a slight bit of shading and shadowing on their diffuse maps as it cand add more detail to the overall texture. The diffuse map can be lower res than the specular and local map as it doesn't usaully contain as much vital data.
A local map is a RGB height based image. The general rule of thumb is to create a model and then generate them, but you CAN paint them if you know what your doing (I paint a lot of mine). Local maps tend to have outline detail only, i.e., anything that would change physical shape, however some people tend to add smaller details to the local map instead of the height map (more on that in a sec). The local map is the msot important part of your surface and can make or break the effect. It needs to be high res, no matter what, otherwise your pixels "merge" into each other when not given enough space and the effect can be messy or muddy (Take a look at the latest Quake 4 pics. Lower res textures dont work for the most part).
A height map is a greyscale or colour (colours not really needed) image that can add detail that the local map could not. Its usally done for smaller detail, that a local map couldn't gain without it being very high resolution. For example, a height map could be used ot generate stubble on a mans face, or a cut/scar. Its absed on colour = depth, and so black is zero bump, white is true bump, and inbetween... Well, you figure it out.
A specular map is the second most important part of your mesh. Its VITAL that this map be colour most of the time. Using a greyscale specular map will make things look plastic, and should only be used on, well, shiney things. It works the same way as a heigth map really. The brighter the pixel, the mroe it shines. You need to keep the colours dark for the most part however as you dont want things to overshine (Like CoD2 seems to be). Like I said, colour is vital to get things looking real. For example, a face would usally have a slight blue specualr map, with white for eyes, and maybe a red for the hair (if his hair is brown).
So basically your plain D3 texture lacks a lot of detail that you would expect from a Q3 texture which, for conversion to Q3, has to be added back into the texture?
Eraser wrote:So basically your plain D3 texture lacks a lot of detail that you would expect from a Q3 texture which, for conversion to Q3, has to be added back into the texture?
Yes. They will always look flat. You can get away with some kind of automated conversion to give it a little spec and depth, but TBH you will really need to re-pain them to look good.
Also, remember Q3 textures are low res by default and you dont want to be mixing different res textures or it will look inconsisten and hurt the eye.
@ Kaz
