Search the Community

DeepO is talking about things which are bad and good game modelling practice. Things which improve performance. Basically, any game engine likes nice neat rows of triangles which it can render all fairly easily. It doesn't like to have them all bunched up unevenly; it doesn't read well. Also it can lead to clipping problems, that sort of glitch.

Looks good. I don't see any problems. Only challenges.

Well, I managed to get it into the editor, but as you can see there are a few problems...

LOL! I was just posting this problem. Bouding Boxes Fixing this issue should also fix the frob problems. I played around this for some coins and I noticed that this box determines when the frob triggers.

Yeah, it might be that the light is pretty harsh. I also tried having the books hilight at the .50 setting and it was quite visible. Could be a combination of the harshness of the light directly above the books and the frobbability problems I was experiencing. The only thing I found was that the books and coins didn't seem to frob when they were centered on the screen. Sometimes they would be off to the side a little bit. How is the middle of the object determined, is there some kind of box around each item that it has to be centered in?

Well, I tried creating one manually, but somehow ran into problems getting everything to be the correct size. I somehow have also lost the webpage tutorial about how to import them into the editor, so that has knocked me off schedule, somewhat.

I think the table is ok. It's the environment where the table doesn't really fit in, though. I don't know. The question is how this impacst performance. Of course the stuff should look good, but will there so many books at once that we will get problems of this? Of course you always have to keep the whole picture in mind. Teh game shouldn't start to get a slidshow when three Ai's are coming in the room. I noticed that you spelled the names differently. One time you wrote coffetab1 and the other one you wrote coffeetab1. After I realized that the table showed up in the editor as well. And of course you must set the game to darkmod as well, but I think you already did this.

I've got some serious problems to resize the table in blender - when I make it bigger parts of the model dissappear in the 3d-view and it doesn't look any bigger in the modelviewer

I just noted that the lightgem model is on the FTP server. This is good to know. Unfortunately it is in a non-ASCII format and thus we can not edit it. I don't want to install lightwave just to be able to to open the file and I wouldn't even know how to do this in ligthwave. Ignoring that fact that I don't want to install a pirated copy of Lightwave as Blender perfectly serves my purpose. As the lightgem is only a static mesh, I expect that I can easily import it into Blender. Of course to avoid such problems in the future you should always assign the material name right from the start. I think we should create a HUD directory in the models section where we can store the respective HUD models. So the lightgem would have a path like this in the material: //darkmod/models/hud/lightgem Also always provide a textfile with the authors name/email, what the file contains and which version this is. oofnish did that in this case, but we should make this as a rule. Since we will store other HUD elements as well the description file should be name <modelname>.txt --------------------------- Author: Name <email> Version: 0.1 Description: this is the most woot model for our mod. ---------------------------

I've uploaded three versions of the Lightgem model (.lwo files) to the FTP site. They are in /models/lightgem. The whatis.txt file gives a basic description of what each file is. The lightgem.lwo file is the high detail model that we probably don't want to use. 2.5 k polys The lightgemr2.lwo file is a medium detail model that renders nearly as well as the high detail model, at 1024 polys. Still probably too high, plus there are some non-planar quads that may cause problems. The lightgemr3.lwo file is a lower res model that looses some visual detail and has 770 polys. Still might be too high, and the nonplanars are still there... The lightgemr4.lwo file is a algorithmically reduced object with 692 polys. These are all tris. The mesh is quite messy in comparison, but the model doesn't render too badly. We may want to use this. The reason I provide all these models is so we can choose which model is the best size and complexity. Also if someone is better at reducing models than I am, they can give it a crack and perhaps come up with a better result. Material definitions are all procedural and stored within the lwo files. I also changed the point colors a bit. Here is a new sample:

I don't know what to do about the other problems, but I think we can avoid at least the model issue next time. Every stuff should go into CVS. Most programs support ASCII formats and this is what we should put in CVS. The textures and soundfiles can stay on the FTP server, as my machine and upload bandwidth is not exactly fast, but everything else should go into CVS. I think this is a better thing to do anway. And with everything I mean EVERYTHING. This includes material, def, map and any othe file that we need. As it is, already I have to change stuff that Renzatics also works on and there are already chances that we overwrite each others changes (custom.mtr for example).

I am pleased its not just me then ..I have over fifteen years of commercial 2d 3d (traditional ) experience (www.alandickinson.net)and have worked with most of the major companies within my field …try to get into the games industry and nothing ..not a sausage! I think one of my problems is that game companies like to put you into very defined boxes and I don’t really fit into any. I must say I am surprised you have not had any success getting an environment role. Renzatic is absolutely right in what he says …I have heard the same said so many times mods are a great way to build experience. Shame you are in Canada I have seen a lot of environment positions here in the uk.

I htink that loot mustn't really be immediately recognizable. If a valuable jug is lying in the sunken city for hundreds of years, it will not really look THAT valuable on first sight. So personally I prefer to have loot sticking out visually as it was done in T1/T2. You could identify any valuable glass immediatly because of the look of it. Some things were not so obvious. I had always problems with wine bottles but that was not a real problem.

ok, its in game... ...sort of. Its not without its problems though DeepO... first of all, its a little on the big side (about 3 zombies high, lol). Scaling is something that happens in Max, not in the Doom3 editor... If it helps you can export objects from Doom3 Radiant and load the into max as a scale reference. Second, its a non Doom standard grid size. Setting up your grid size in Max to match the grid size in doom will make it a hell of a lot easier to attach the lamp accurately to walls Doom uses grids based on 8s, does max default to grids based on 10s? Thirdly, I am having trouble getting textures onto models and ingame generally. (it requires manually hacking the model file and frankly it isn't fun). You're lamp is no exception I'll come back to it in a day or so when I have had a bit more practice.... Forthly, I don't use Max (do I look rich ) otherwise I'd have had a go at fixing this stuff for you. I use blender. It might be possible to share editable copies of the file in .3ds format... Fithly, Radiant has a small moan about defunct triangles. It might be worth checking the model for loose ends...

The overall shape is great, the proportions are perfect, it would be excellent as a low poly game model - but isn't this supposed to be the high res version to source the nomal maps from? You seem to be very tight-fisted with the polys, when you should be writing blank cheques. The main problems as I see it, is the geometric uniform nature of the organic shapes - legs and arms are not perfect cylinders, leather in real life is not flat and perfectly smooth. Eceyting need to be messed up a bt, like it's had a long hard life. And as you say youself, and I pointed out earlier, a lot of the detail still looks injection molded like it's all been beveled out. I have to say, and this is possibly my worst crit - it looks like you're trying to emulate the T3 Garrett model, which I considered risible when I saw in untextured in Lightwave.

Not necessarily. I'm sure the FM community at large will be more than happy to build their own custom content. Much better to provide a solid set of IK bones and walk (etc.) animations that can be customised than multiple variations on everything. It is, aftar all a (sort of) fairly streight forward proposition to plonk a crayman model and skin on top of a zombie bone-set and call it a beasticle. I hate to be the nay sayer (again) but we really should be concentrating on getting some basic characters in the Dooooom engine, getting them animated and ironing out all the problems as a priority. Hows about applying some thought to character animation and poses springheel and Darkness Falls. Something a bit like a story board (but for individual actions) that can be used a guide for key frames. How about giving some detail on the existing concepts. Or providing side and front views to make the modelling process that much easier.... Grrr. I really don't like making this point over and over because the concept art that you guys are producing is a treat. They really do look fantastic. But these are parallel tasks that can be taken up at pretty much any point. There are mission critical tasks that need addressing first. err, yeah - anyway. what the other guys have said. The grey and green colours are a bit alien-eseque. Are these things blind/poor sighted or do they have big reflective eyes for seeing in the dark? Big crayfish tenticles would be cool for feeling your way around in the gloom. Spiders. I want Spiders.

Please guys, stop with the praise already or I be unable to squeeze my swollen head though the door of the computer room, and you won't see any more work Having said that, this is the best model I've ever made (which isn't saying much since I've only made about 12) but I suppose I have Springheel to thank for the inspiring concept art.. I think the key to success on projects such as these is enthusiasm and a desire to learn new things and improve on whxt you previously thought you were capable of. Springheel - yes, it was a toss up between expensive satin shoes or sandals, but I can't see him wearing big armoured boots. This all looks fine and dandy as a static model, but I can see problems with animatng it. All that cloth around the arms and legs makes it very difficult. The animations would have to be very stilted. Unless it's possible to animate it in a 3d app with real cloth dynamics and then export it to the doom engine complete wth the new animatons. I dn't know yet if doom will only recognise vertices that have been animated by bones.

I don't see why anyone who wants to record some character voices can't have a shot. As long as they're all good quality there shouldn't be any problems.

Ok then, it's a date. I'm having problems with my internet connection ATM, and can only access from various friends computers, so don't despair if you don't hear from me in a while) I love your work on the main thief DeepOmega, very (n)ice. Having such great concept art really helps, huh? With Nightblade I'm just having to do all my own concepts, and they're quite derivitive and hackneyed for the most part.. If you could snag K^2 from the ion boards, it would be nice, he's knows a lot about scripting, writing shaders, plugins and all sorts of programing goodness. Pester him about it. He's also a great Thief fan. He and I are also involved in another mod for the forthcoming Halflife 2. I'm PM you his email Fingernail. edit: I have doom3 (shit game) but what other tools do I need? Have they released an SDK yet?

There shouldn't be any legal problems with using old paintings. Copywrites only cover works that are less than 75 years old (though Disney might change that), after that it becomes public domain. And I think Blackthief's photochop job would be a good template for a Hammerite Priest.

Well I have just been listening to the MP3's, and Darkness Falls has my vote to keep working on the ambience tracks for our mod That kind of thing sounds like it could be developed into the style from Thief we all know and love. I tried uploading them to the DarkMod FTP, but I had problems, I posted in the FTP thread. (I have cable access here at Uni)

I'm back from my trip to the castle with a bunch of photos. The 1st texture is already on the ftp: http://208.49.149.118/TheDarkMod/Textures/door2.jpg @renz I don't think we should downsize the textures, just that they fit easier. IMO there won't be any problems with huge high-res textures, cauze 1. our mod will take some time and people will have even better hardware than today 2. HD space also shouldn't be a problem - today almost everybody has more than 100gb HD.

Renz: Nice as your room is, and 1337 though your skillz be, there are some problems. Did you test the map in game mode? You can't have done, seeing as it is not buildable in it's current state. Even though I added an end wall, it still 'leaked', because the other walls were too thin. Make them thicker. And no gaps. Also, the scale is a bit off, as seen here: I know walk-in fireplaces might be considered a luxury. Anyway, otherwise it looks great. Put in a scientist for Great Scalability.

Short sword. No problems.

WOOHOO! I CAN POST AGAIN! lol, anyway...I have a few wallpaper textures of my own to contribute. None of them are quite as Thiefy as Blackthief's, but I'm sure they'll have their place in the game somewhere. This one turned out great ingame, but alas...it's sort of a funky hell of an ugly wallpaper. Could fit for bathrooms, though. This one is more intheme, but it looks more like a carpet than a wallpaper. They're not bad, but I like the ones posted above a helluva lot better. So far I have about 27 textures from various sources ingame with about 5 or 6 of them refusing to show up properly. Originally I thought tweaking the color levels confused the editor, but after some tweaked textures showed up I'm starting to believe it has more to do with either the directory structure or just plain randomness . Since some of the default Id textures refused to show up even before I started modding it's hard to say. Edit: Once again I have to bring up how much better it is to do normalmaps by hand... In the end it makes it far cleaner and looks a good deal better. It's not that hard to do either, I've never done any texturework before I started this project so I'm sure some of you photoshop pros won't have any problems at all.