My input:
1) do your textures for example double the resolution you think will be the final rezz in game, then downscale it before saving for game use. You can also sharpen the final "small" texture with "unsharp mask" method for crispier look:
http://www.google.fi/search?hl=fi&q=unsharp+mask&meta=lr%3Dlang_en%7Clang_fi
Also the blurriness can often come from MipMaps and how the app uses them, so even if your textures are HQ and crispy the app might use too soon lower level MipMaps and thus make the model look blurry.
Also keeping everything at constant scale will make things look better, ie do NOT have a table with 512 texture and a bottle on top of it having a 512 texture also - both models will look bad because they have so big (comparative) texture scale difference.
2) I do specular map after diffuse (your "normal" texture/color maps) maps are done and usually start by making it BW and trying out different contrast/brightness values. Often this process loses areas I want to be specular ot non-specular and then you have to draw this on top of the base spec map using what ever method is most handy.
3) I use .tgas at work and home because it's lossless. JPG is lossy so I wouldn't use that - it does not save your VRAM or anything, just makes the textures load bit faster. I rather take the better in-game quality instead of one-time load from disk speed-up.
Having your textures in .tga format is good for editing too; if you use .jpgs and edit it you'll lose details everytime you edit and save it because image gets re-packed and usually the whole image is affected (degenerated) a bit in result. Soon your whole texture is a one big mess... :)
At work we convert our (BIG) textures to .dds because it's a good way to save VRAM and still keep quite nice quality. I would use .dds's in Blitz3D too of course if it would support that but BRL is not going add it... maybe I should switch to A6 too? :)
1) do your textures for example double the resolution you think will be the final rezz in game, then downscale it before saving for game use. You can also sharpen the final "small" texture with "unsharp mask" method for crispier look:
http://www.google.fi/search?hl=fi&q=unsharp+mask&meta=lr%3Dlang_en%7Clang_fi
Also the blurriness can often come from MipMaps and how the app uses them, so even if your textures are HQ and crispy the app might use too soon lower level MipMaps and thus make the model look blurry.
Also keeping everything at constant scale will make things look better, ie do NOT have a table with 512 texture and a bottle on top of it having a 512 texture also - both models will look bad because they have so big (comparative) texture scale difference.
2) I do specular map after diffuse (your "normal" texture/color maps) maps are done and usually start by making it BW and trying out different contrast/brightness values. Often this process loses areas I want to be specular ot non-specular and then you have to draw this on top of the base spec map using what ever method is most handy.
3) I use .tgas at work and home because it's lossless. JPG is lossy so I wouldn't use that - it does not save your VRAM or anything, just makes the textures load bit faster. I rather take the better in-game quality instead of one-time load from disk speed-up.
Having your textures in .tga format is good for editing too; if you use .jpgs and edit it you'll lose details everytime you edit and save it because image gets re-packed and usually the whole image is affected (degenerated) a bit in result. Soon your whole texture is a one big mess... :)
At work we convert our (BIG) textures to .dds because it's a good way to save VRAM and still keep quite nice quality. I would use .dds's in Blitz3D too of course if it would support that but BRL is not going add it... maybe I should switch to A6 too? :)